Fujitsu MPG3XXXAH E User Manual

MPG3xxxAH-E  
DISK DRIVES  
PRODUCT MANUAL  
C141-E116-01EN  
This page is intentionally left blank.  
MANUAL ORGANIZATION  
MPG3xxxAH-E  
DISK DRIVES  
PRODUCT  
MANUAL  
(C141-E116)  
<This manual>  
• DEVICE OVERVIEW  
• DEVICE CONFIGURATION  
• INSTALLATION CONDITIONS  
• THEORY OF DEVICE OPERATION  
• INTERFACE  
• OPERATIONS  
MPG3xxxAH-E  
DISK DRIVES  
MAINTENANCE  
MANUAL  
• MAINTENANCE AND DIAGNOSIS  
• REMOVAL AND REPLACEMENT PROCEDURE  
(C141-F050)  
C141-E116-01EN  
iii  
This page is intentionally left blank.  
PREFACE  
This manual describes the MPG3xxxAH-E series, a 3.5-inch hard disk drive with a BUILT-IN controller  
that is compatible with the ATA interface.  
This manual explains, in detail, how to incorporate the hard disk drives into user systems.  
This manual assumes that users have a basic knowledge of hard disk drives and their application in  
computer systems.  
This manual consists of the following six chapters:  
Chapter 1  
Chapter 2  
Chapter 3  
Chapter 4  
Chapter 5  
Chapter 6  
DEVICE OVERVIEW  
DEVICE CONFIGURATION  
INSTALLATION CONDITIONS  
THEORY OF DEVICE OPERATION  
INTERFACE  
OPERATIONS  
In this manual, disk drives may be referred to as drives or devices.  
C141-E116-01EN  
v
Conventions for Alert Messages  
This manual uses the following conventions to show the alert messages. An alert message consists of an  
alert signal and alert statements. The alert signal consists of an alert symbol and a signal word or just a  
signal word.  
The following are the alert signals and their meanings:  
This indicates a hazardous situation likely to result in serious personal  
injury if the user does not perform the procedure correctly.  
This indicates a hazardous situation could result in personal injury if the  
user does not perform the procedure correctly.  
This indicates a hazardous situation could result in minor or moderate  
personal injury if the user does not perform the procedure correctly. This  
alert signal also indicates that damages to the product or other property,  
may occur if the user does not perform the procedure correctly.  
This indicates information that could help the user use the product more  
efficiently.  
In the text, the alert signal is centered, followed below by the indented message. A wider line space  
precedes and follows the alert message to show where the alert message begins and ends. The following is  
an example:  
(Example)  
IMPORTANT  
HA (host adapter) consists of address decoder, driver, and receiver.  
ATA is an abbreviation of "AT attachment". The disk drive is  
conformed to the ATA-4 interface  
The main alert messages in the text are also listed in the “Important Alert Items.”  
vi  
C141-E116-01EN  
LIABILITY EXCEPTION  
"Disk drive defects" refers to defects that involve adjustment, repair, or replacement.  
Fujitsu is not liable for any other disk drive defects, such as those caused by user misoperation or  
mishandling, inappropriate operating environments, defects in the power supply or cable, problems of the  
host system, or other causes outside the disk drive.  
C141-E116-01EN  
vii  
This page is intentionally left blank.  
CONTENTS  
page  
DEVICE OVERVIEW..........................................................................................1 - 1  
CHAPTER 1  
1.1  
Features .................................................................................................................................1 - 1  
Functions and performance ...................................................................................................1 - 1  
Adaptability...........................................................................................................................1 - 2  
Interface.................................................................................................................................1 - 2  
Device Specifications............................................................................................................1 - 4  
Specifications summary.........................................................................................................1 - 4  
Model and product number ...................................................................................................1 - 5  
Power Requirements..............................................................................................................1 - 5  
Environmental Specifications................................................................................................1 - 8  
Acoustic Noise ......................................................................................................................1 - 8  
Shock and Vibration..............................................................................................................1 - 9  
Reliability..............................................................................................................................1 - 9  
Error Rate..............................................................................................................................1 - 10  
Media Defects .......................................................................................................................1 - 10  
1.1.1  
1.1.2  
1.1.3  
1.2  
1.2.1  
1.2.2  
1.3  
1.4  
1.5  
1.6  
1.7  
1.8  
1.9  
CHAPTER 2  
DEVICE CONFIGURATION .............................................................................2 - 1  
2.1  
Device Configuration ............................................................................................................2 - 1  
System Configuration............................................................................................................2 - 3  
ATA interface........................................................................................................................2 - 3  
1 drive connection.................................................................................................................2 - 3  
2 drives connection................................................................................................................2 - 3  
2.2  
2.2.1  
2.2.2  
2.2.3  
CHAPTER 3  
INSTALLATION CONDITIONS........................................................................3 - 1  
3.1  
Dimensions............................................................................................................................3 - 1  
Handling Cautions.................................................................................................................3 - 3  
General notes.........................................................................................................................3 - 3  
Installation.............................................................................................................................3 - 3  
Recommended equipments....................................................................................................3 - 3  
Mounting...............................................................................................................................3 - 4  
Cable Connections.................................................................................................................3 - 8  
Device connector...................................................................................................................3 - 8  
3.2  
3.2.1  
3.2.2  
3.2.3  
3.3  
3.4  
3.4.1  
C141-E116-01EN  
ix  
3.4.2  
3.4.3  
3.4.4  
3.4.5  
3.5  
Cable connector specifications..............................................................................................3 - 9  
Device connection.................................................................................................................3 - 9  
Power supply connector (CN1) .............................................................................................3 - 10  
System configuration for Ultra DMA....................................................................................3 - 10  
Jumper Settings .....................................................................................................................3 - 13  
Location of setting jumpers...................................................................................................3 - 13  
Factory default setting...........................................................................................................3 - 14  
Jumper configuration.............................................................................................................3 - 14  
3.5.1  
3.5.2  
3.5.3  
CHAPTER 4  
THEORY OF DEVICE OPERATION ...............................................................4 - 1  
4.1  
Outline...................................................................................................................................4 - 1  
Subassemblies .......................................................................................................................4 - 1  
Disk .......................................................................................................................................4 - 1  
Head ......................................................................................................................................4 - 2  
Spindle ..................................................................................................................................4 - 2  
Actuator.................................................................................................................................4 - 3  
Air filter.................................................................................................................................4 - 3  
Circuit Configuration.............................................................................................................4 - 4  
Power-on Sequence...............................................................................................................4 - 6  
Self-calibration......................................................................................................................4 - 8  
Self-calibration contents........................................................................................................4 - 8  
Execution timing of self-calibration ......................................................................................4 - 9  
Command processing during self-calibration ........................................................................4 - 9  
Read/write Circuit .................................................................................................................4 - 10  
Read/write preamplifier (PreAMP) .......................................................................................4 - 10  
Write circuit ..........................................................................................................................4 - 10  
Read circuit ...........................................................................................................................4 - 11  
Synthesizer circuit .................................................................................................................4 - 12  
Servo Control ........................................................................................................................4 - 13  
Servo control circuit..............................................................................................................4 - 13  
Data-surface servo format .....................................................................................................4 - 16  
Servo frame format................................................................................................................4 - 17  
Actuator motor control..........................................................................................................4 - 18  
Spindle motor control............................................................................................................4 - 19  
4.2  
4.2.1  
4.2.2  
4.2.3  
4.2.4  
4.2.5  
4.3  
4.4  
4.5  
4.5.1  
4.5.2  
4.5.3  
4.6  
4.6.1  
4.6.2  
4.6.3  
4.6.4  
4.7  
4.7.1  
4.7.2  
4.7.3  
4.7.4  
4.7.5  
x
C141-E116-01EN  
CHAPTER 5  
INTERFACE.........................................................................................................5 - 1  
5.1  
Physical Interface ..................................................................................................................5 - 2  
Interface signals.....................................................................................................................5 - 2  
Signal assignment on the connector.......................................................................................5 - 3  
Logical Interface ...................................................................................................................5 - 6  
I/O registers...........................................................................................................................5 - 6  
Command block registers......................................................................................................5 - 8  
Control block registers ..........................................................................................................5 - 13  
Host Commands ....................................................................................................................5 - 13  
Command code and parameters.............................................................................................5 - 14  
Command descriptions..........................................................................................................5 - 16  
Error posting..........................................................................................................................5 - 76  
Command Protocol................................................................................................................5 - 77  
Data transferring commands from device to host..................................................................5 - 77  
Data transferring commands from host to device..................................................................5 - 79  
Commands without data transfer...........................................................................................5 - 81  
Other commands....................................................................................................................5 - 82  
DMA data transfer commands...............................................................................................5 - 82  
Ultra DMA feature set...........................................................................................................5 - 84  
Overview...............................................................................................................................5 - 84  
Phases of operation................................................................................................................5 - 85  
Ultra DMA data in commands...............................................................................................5 - 85  
5.1.1  
5.1.2  
5.2  
5.2.1  
5.2.2  
5.2.3  
5.3  
5.3.1  
5.3.2  
5.3.3  
5.4  
5.4.1  
5.4.2  
5.4.3  
5.4.4  
5.4.5  
5.5  
5.5.1  
5.5.2  
5.5.3  
5.5.3.1 Initiating an Ultra DMA data in burst....................................................................................5 - 85  
5.5.3.2 The data in transfer................................................................................................................5 - 86  
5.5.3.3 Pausing an Ultra DMA data in burst .....................................................................................5 - 86  
5.5.3.4 Terminating an Ultra DMA data in burst ..............................................................................5 - 87  
5.5.4  
Ultra DMA data out commands.............................................................................................5 - 89  
5.5.4.1 Initiating an Ultra DMA data out burst..................................................................................5 - 89  
5.5.4.2 The data out transfer..............................................................................................................5 - 90  
5.5.4.3 Pausing an Ultra DMA data out burst ...................................................................................5 - 90  
5.5.4.4 Terminating an Ultra DMA data out burst ............................................................................5 - 91  
5.5.5  
5.5.6  
5.6  
Ultra DMA CRC rules...........................................................................................................5 - 93  
Series termination required for Ultra DMA...........................................................................5 - 94  
Timing...................................................................................................................................5 - 95  
PIO data transfer ...................................................................................................................5 - 95  
Multiword data transfer.........................................................................................................5 - 96  
5.6.1  
5.6.2  
C141-E116-01EN  
xi  
5.6.3  
Ultra DMA data transfer........................................................................................................5 - 97  
5.6.3.1 Initiating an Ultra DMA data in burst....................................................................................5 - 97  
5.6.3.2 Ultra DMA data burst timing requirements...........................................................................5 - 98  
5.6.3.3 Sustained Ultra DMA data in burst .......................................................................................5 - 101  
5.6.3.4 Host pausing an Ultra DMA data in burst .............................................................................5 - 102  
5.6.3.5 Device terminating an Ultra DMA data in burst....................................................................5 - 103  
5.6.3.6 Host terminating an Ultra DMA data in burst .......................................................................5 - 104  
5.6.3.7 Initiating an Ultra DMA data out burst..................................................................................5 - 105  
5.6.3.8 Sustained Ultra DMA data out burst .....................................................................................5 - 106  
5.6.3.9 Device pausing an Ultra DMA data out burst........................................................................5 - 107  
5.6.3.10 Host terminating an Ultra DMA data out burst .....................................................................5 - 108  
5.6.3.11 Device terminating an Ultra DMA data in burst....................................................................5 - 109  
5.6.4  
Power-on and reset................................................................................................................5 - 110  
CHAPTER 6  
OPERATIONS ......................................................................................................6 - 1  
6.1  
Device Response to the Reset................................................................................................6 - 1  
Response to power-on ...........................................................................................................6 - 2  
Response to hardware reset ...................................................................................................6 - 3  
Response to software reset ....................................................................................................6 - 4  
Response to diagnostic command..........................................................................................6 - 5  
Address Translation...............................................................................................................6 - 6  
Default parameters ................................................................................................................6 - 6  
Logical address......................................................................................................................6 - 7  
Power Save............................................................................................................................6 - 8  
Power save mode...................................................................................................................6 - 8  
Power commands ..................................................................................................................6 - 10  
Defect Management...............................................................................................................6 - 10  
Spare area..............................................................................................................................6 - 11  
Alternating defective sectors .................................................................................................6 - 11  
Read-Ahead Cache................................................................................................................6 - 13  
Data buffer configuration ......................................................................................................6 - 13  
Caching operation .................................................................................................................6 - 14  
Usage of read segment...........................................................................................................6 - 15  
Write Cache...........................................................................................................................6 - 20  
6.1.1  
6.1.2  
6.1.3  
6.1.4  
6.2  
6.2.1  
6.2.2  
6.3  
6.3.1  
6.3.2  
6.4  
6.4.1  
6.4.2  
6.5  
6.5.1  
6.5.2  
6.5.3  
6.6  
xii  
C141-E116-01EN  
FIGURES  
page  
1.1  
2.1  
2.2  
2.3  
3.1  
3.2  
3.3  
3.4  
3.5  
3.6  
3.7  
3.8  
3.9  
3.10  
3.11  
3.12  
Current fluctuation (Typ.) when power is turned on .............................................................1 - 7  
Disk drive outerview .............................................................................................................2 - 1  
1 drive system configuration .................................................................................................2 - 3  
2 drives configuration............................................................................................................2 - 3  
Dimensions............................................................................................................................3 - 2  
Handling cautions..................................................................................................................3 - 3  
Direction................................................................................................................................3 - 4  
Limitation of side-mounting..................................................................................................3 - 5  
Mounting frame structure......................................................................................................3 - 5  
Surface temperature measurement points..............................................................................3 - 6  
Service area...........................................................................................................................3 - 7  
Connector locations...............................................................................................................3 - 8  
Cable connections .................................................................................................................3 - 9  
Power supply connector pins (CN1)......................................................................................3 - 10  
Cable configuration...............................................................................................................3 - 11  
Cable type detection using CBLID- signal  
(Host sensing the condition of the CBLID- signal) ...............................................................3 - 12  
3.13  
Cable type detection using IDENTIFY DEVICE data  
(Device sensing the condition of the CBLID- signal)............................................................3 - 12  
3.14  
3.15  
3.16  
3.17  
3.18  
3.19  
4.1  
Jumper location .....................................................................................................................3 - 13  
Factory default setting...........................................................................................................3 - 14  
Jumper setting of master or slave device...............................................................................3 - 14  
Jumper setting of Cable Select ..............................................................................................3 - 15  
Example (1) of Cable Select..................................................................................................3 - 15  
Example (2) of Cable Select..................................................................................................3 - 15  
Head structure .......................................................................................................................4 - 2  
MPG3xxxAH-E Block diagram ............................................................................................4 - 5  
Power-on operation sequence................................................................................................4 - 7  
Block diagram of servo control circuit..................................................................................4 - 13  
Physical sector servo configuration on disk surface..............................................................4 - 15  
96 servo frames in each track ................................................................................................4 - 17  
Execution example of READ MULTIPLE command ...........................................................5 - 19  
Read Sector(s) command protocol ........................................................................................5 - 78  
4.2  
4.3  
4.4  
4.5  
4.6  
5.1  
5.2  
C141-E116-01EN  
xiii  
5.3  
Protocol for command abort..................................................................................................5 - 79  
WRITE SECTOR(S) command protocol ..............................................................................5 - 80  
Protocol for the command execution without data transfer ...................................................5 - 81  
Normal DMA data transfer....................................................................................................5 - 83  
Ultra DMA termination with pull-up or pull-down ...............................................................5 - 94  
PIO data transfer timing ........................................................................................................5 - 95  
Multiword DMA data transfer timing (mode 2) ....................................................................5 - 96  
Initiating an Ultra DMA data in burst....................................................................................5 - 97  
Sustained Ultra DMA data in burst .......................................................................................5 - 101  
Host pausing an Ultra DMA data in burst .............................................................................5 - 102  
Device terminating an Ultra DMA data in burst....................................................................5 - 103  
Host terminating an Ultra DMA data in burst .......................................................................5 - 104  
Initiating an Ultra DMA data out burst..................................................................................5 - 105  
Sustained Ultra DMA data out burst .....................................................................................5 - 106  
Device pausing an Ultra DMA data out burst........................................................................5 - 107  
Host terminating an Ultra DMA data out burst .....................................................................5 - 108  
Device terminating an Ultra DMA data out burst..................................................................5 - 109  
Power-on Reset Timing.........................................................................................................5 - 110  
Response to power-on ...........................................................................................................6 - 2  
Response to hardware reset ...................................................................................................6 - 3  
Response to software reset ....................................................................................................6 - 4  
Response to diagnostic command..........................................................................................6 - 5  
Address translation (example in CHS mode) ........................................................................6 - 7  
Address translation (example in LBA mode) ........................................................................6 - 8  
Sector slip processing............................................................................................................6 - 11  
Alternate cylinder assignment ...............................................................................................6 - 12  
Data buffer configuration ......................................................................................................6 - 13  
5.4  
5.5  
5.6  
5.7  
5.8  
5.9  
5.10  
5.11  
5.12  
5.13  
5.14  
5.15  
5.16  
5.17  
5.18  
5.19  
5.20  
6.1  
6.2  
6.3  
6.4  
6.5  
6.6  
6.7  
6.8  
6.9  
xiv  
C141-E116-01EN  
TABLES  
page  
1.1  
Specifications ........................................................................................................................1 - 4  
Model names and product numbers.......................................................................................1 - 5  
Current and power dissipation...............................................................................................1 - 6  
Environmental specifications ................................................................................................1 - 8  
Acoustic noise specification..................................................................................................1 - 8  
Shock and vibration specification..........................................................................................1 - 9  
Surface temperature measurement points and standard values..............................................3 - 6  
Cable connector specifications..............................................................................................3 - 9  
Transfer rate of each zone .....................................................................................................4 - 12  
Interface signals.....................................................................................................................5 - 2  
Signal assignment on the interface connector........................................................................5 - 3  
I/O registers...........................................................................................................................5 - 7  
Command code and parameters.............................................................................................5 - 14  
Information to be read by IDENTIFY DEVICE command...................................................5 - 29  
Features register values and settable modes ..........................................................................5 - 36  
Diagnostic code.....................................................................................................................5 - 40  
Features Register values (subcommands) and functions........................................................5 - 50  
Device attribute data structure...............................................................................................5 - 53  
Warranty failure threshold data structure ..............................................................................5 - 54  
Log directory data structure...................................................................................................5 - 59  
Error logging data structure...................................................................................................5 - 60  
Self Test log data structure....................................................................................................5 - 61  
Contents of security password...............................................................................................5 - 63  
Contents of SECURITY SET PASSWORD data..................................................................5 - 68  
1.2  
1.3  
1.4  
1.5  
1.6  
3.1  
3.2  
4.1  
5.1  
5.2  
5.3  
5.4  
5.5  
5.6  
5.7  
5.8  
5.9  
5.10  
5.11  
5.12  
5.13  
5.14  
5.15  
5.16  
Relationship between combination of Identifier and Security level,  
and operation of the lock function.........................................................................................5 - 68  
5.17  
5.18  
5.19  
5.20  
6.1  
Command code and parameters.............................................................................................5 - 76  
Recommended series termination for Ultra DMA.................................................................5 - 94  
Ultra DMA data burst timing requirements...........................................................................5 - 98  
Ultra DMA sender and recipient timing requirements...........................................................5 - 100  
Default parameters ................................................................................................................6 - 6  
C141-E116-01EN  
xv  
This page is intentionally left blank.  
CHAPTER 1  
DEVICE OVERVIEW  
1.1  
1.2  
1.3  
1.4  
1.5  
1.6  
1.7  
1.8  
1.9  
Features  
Device Specifications  
Power Requirements  
Environmental Specifications  
Acoustic Noise  
Shock and Vibration  
Reliability  
Error Rate  
Media Defects  
Overview and features are described in this chapter, and specifications and power requirement are  
described.  
The MPG3xxxAH-E series are a 3.5-inch hard disk drive with a built-in ATA controller. The disk drive is  
compact and reliable.  
1.1  
Features  
1.1.1  
(1)  
Functions and performance  
Compact  
The disk has 1 or 2 disks of 95 mm (3.5 inches) diameter, and its height is 26.1 mm (1 inch).  
(2)  
(3)  
Large capacity  
The disk drive can record up to 20.49 GB (formatted) on one disk using the 48/52 EEPR4ML  
recording method and 15 recording zone technology. The MPG3xxxAH-E series have a formatted  
capacity of 20.49 GB to 40.99 GB respectively.  
High-speed Transfer rate  
The disk drive has an internal data rate up to 60.7 MB/s. The disk drive supports an external data  
rate up to 16.7 MB/s (PIO mode 4, DMA mode 2), 66.6 MB/s (ultra DMA mode 4) or 100MB/s  
(ultra DMA mode 5).  
C141-E116-01EN  
1 - 1  
(4)  
Average positioning time  
Use of a rotary voice coil motor in the head positioning mechanism greatly increases the  
positioning speed. The average positioning time is 8.5 ms (at read).  
1.1.2  
Adaptability  
(1)  
Power save mode  
The power save mode feature for idle operation, stand by and sleep modes makes the disk drive  
ideal for applications where power consumption is a factor.  
(2)  
(3)  
Wide temperature range  
The disk drive can be used over a wide temperature range (5°C to 55°C).  
Low noise and vibration  
In Ready status, the noise of the disk drive is only about 3.3 bels (MPG3409AH-E, Typical Sound  
Power per ISO7779 and ISO9296).  
1.1.3  
Interface  
(1)  
Connection to interface  
With the built-in ATA interface controller, the disk drive can be connected to an ATA interface of  
a personal computer.  
(2)  
(3)  
2MB data buffer  
The disk drive uses a 2MB data buffer to transfer data between the host and the disk media.  
In combination with the read-ahead cache system described in item (3) and the write cache  
described in item (6), the buffer contributes to efficient I/O processing.  
Read-ahead cache system  
After the execution of a disk read command, the disk drive automatically reads the subsequent data  
block and writes it to the data buffer (read ahead operation). This cache system enables fast data  
access. The next disk read command would normally cause another disk access. But, if the read  
ahead data corresponds to the data requested by the next read command, the data in the buffer can  
be transferred instead.  
(4)  
Master/slave  
The disk drive can be connected to ATA interface as daisy chain configuration. Drive 0 is a  
master device, drive 1 is a slave device.  
1 - 2  
C141-E116-01EN  
(5)  
(6)  
Error correction and retry by ECC  
If a recoverable error occurs, the disk drive itself attempts error recovery. The 40 bytes ECC has  
improved buffer error correction for correctable data errors.  
Write cache  
When the disk drive receives a write command, the disk drive posts the command completion at  
completion of transferring data to the data buffer completion of writing to the disk media. This  
feature reduces the access time at writing.  
C141-E116-01EN  
1 - 3  
1.2  
Device Specifications  
1.2.1  
Specifications summary  
Table 1.1 shows the specifications of the disk drive.  
Table 1.1 Specifications  
MPG3204AH-E  
MPG3307AH-E  
MPG3409AH-E  
Formatted Capacity (*1)  
20.49 GB  
30.74 GB  
40.99 GB  
Number of Disks  
Number of Heads  
1
2
2
4
2
4
Number of Cylinders  
(User + Alternate & SA)  
30,784 + 735 + 30  
Bytes per Sector  
Recording Method  
Track Density  
512  
48/52 CC2EPRML  
33,000 TPI  
364,119 BPI  
7,200 rpm  
Bit Density  
491,112 BPI  
491,112 BPI  
Rotational Speed  
Average Latency  
Positioning time (Fast)  
4.17 ms  
Minimum  
Average  
Maximum  
(Read) 0.8 ms typical, (Write) 1.0 ms typical  
(Read) 8.5 ms typical, (Write) 9.5 ms typical  
(Read) 16 ms typical, (Write) 17 ms typical  
Positioning time (Slow)  
Minimum  
Average  
Maximum  
(Read) 0.8 ms typical, (Write) 1.0 ms typical  
(Read) 11 ms typical, (Write) 12 ms typical  
(Read) 20 ms typical, (Write) 21 ms typical  
Start/Stop time  
• Start (0 rpm to Drive Read)  
Typical: 8 sec.,  
Typical: 20 sec.,  
Maximum: 15 sec.  
Maximum: 30 sec.  
Stop (at Power Down)  
Interface  
ATA–5  
(Maximum Cable length: 0.46 m)  
Data Transfer Rate  
To/From Media  
38.0 to 60.7 MB/s  
27.9 to 49.7 MB/s  
38.0 to 60.7 MB/s  
16.7 MB/s Max. (burst PIO mode 4, burst DMA mode 2),  
66.6 MB/s Max. (burst ultra DMA mode 4)  
To/From Host  
100.0 MB/s Max. (burst ultra DMA mode 5)  
Data buffer  
2 MB  
Physical Dimensions  
(Height × Width × Depth)  
26.1 mm max. × 101.6 mm × 146.0 mm  
(1.0” max. × 4.0” × 5.7”)  
Weight  
600 g or less  
*1: Capacity under the LBA mode.  
Under the CHS mode (normal BIOS specification), formatted capacity, number of cylinders, number of heads, and number of  
sectors are as follows.  
1 - 4  
C141-E116-01EN  
Model  
Formatted Capacity  
20,496 MB  
No. of Cylinder  
16,383  
No. of Heads  
No. of Sectors  
MPG3204AH-E  
MPG3307AH-E  
MPG3409AH-E  
16  
16  
16  
63  
63  
63  
30,743 MB  
16,383  
40,992 MB  
16,383  
1.2.2  
Model and product number  
Table 1.2 lists the model names and product numbers.  
Table 1.2 Model names and product numbers  
Capacity  
(user area)  
Mounting  
Screw  
Model Name  
Order No.  
Remarks  
MPG3204AH-E  
MPG3307AH-E  
MPG3409AH-E  
20.49 GB  
30.74 GB  
40.99 GB  
No. 6-32UNC  
No. 6-32UNC  
No. 6-32UNC  
CA05762-B521  
CA05762-B544  
CA05762-B542  
1.3  
Power Requirements  
Input Voltage  
(1)  
+ 5 V ±5 %  
+ 12 V ±8 %  
(2)  
Ripple  
+12 V  
+5 V  
Maximum  
Frequency  
200 mV (peak to peak)  
DC to 1 MHz  
100 mV (peak to peak)  
DC to 1 MHz  
C141-E116-01EN  
1 - 5  
(3)  
Current Requirements and Power Dissipation  
Table 1.3 lists the current and power dissipation.  
Table 1.3 Current and power dissipation  
Typical RMS current (*1) [mA]  
+12 V  
Mode of Operation  
Typical Power (*2) [watts]  
MPG  
+5 V  
Model  
MPG  
MPG  
MPG  
All  
MPG  
MPG  
3204AH-E  
3307AH-E  
3409AH-E  
Models  
3204AH-E  
3307AH-E  
3409AH-E  
Spin up  
1950 peak  
350  
1950 peak  
400  
1950 peak  
400  
800 peak  
490  
27.4  
27.4  
27.4  
7.3  
Idle (Ready) (*3)  
R/W (On Track) (*4)  
Seek (Random) (*5)  
Standby  
6.7  
7.3  
370  
450  
450  
600  
7.4  
8.4  
8.4  
480  
540  
540  
570  
8.6  
9.3  
9.3  
18  
18  
18  
120  
0.8  
0.8  
0.8  
Sleep  
18  
18  
18  
120  
0.8  
0.8  
0.8  
Energy efficiency (rank) (*6) [watt / GB]  
0.327 (A)  
0.237 (B)  
0.178 (B)  
*1 Current is typical rms except for spin up.  
*2 Power requirements reflect nominal values for +12V and +5V power.  
*3 Idle mode is in effect when the drive is not reading, writing, seeking, or executing any commands. A portion of the R/W circuitry is powered down,  
the spindle motor is up to speed and the Drive ready condition exists.  
*4 R/W mode is defined as 50% read operations and 50% write operations on a single physical track.  
*5 Seek/W/R mode is defined as 33% seek operations, 33% write operations, and 33% read operations.  
*6 Energy efficiency based on the Law concerning the Rational Use of Energy indicates the value obtained by dividing power consumption by the storage  
capacity. (Japan only)  
1 - 6  
C141-E116-01EN  
(4)  
Current fluctuation (Typ.) when power is turned on  
Note:  
Maximum current is 1.95A.  
Figure 1.1 Current fluctuation (Typ.) when power is turned on  
(5)  
Power on/off sequence  
The voltage detector circuit monitors +5 V and +12 V. The circuit does not allow a write signal if  
either voltage is abnormal. This prevents data from being destroyed and eliminates the need to be  
concerned with the power on/off sequence.  
C141-E116-01EN  
1 - 7  
1.4  
Environmental Specifications  
Table 1.4 lists the environmental specifications.  
Table 1.4 Environmental specifications  
Temperature  
• Operating  
5°C to 55°C (ambient)  
5°C to 60°C (disk enclosure surface)  
• Non-operating  
–40°C to 60°C  
• Thermal Gradient  
20°C/hour or less  
Humidity  
• Operating  
• Non-operating  
• Maximum Wet Bulb  
8% to 80%RH (Non-condensing)  
5% to 85%RH (Non-condensing)  
29°C  
Altitude (relative to sea level)  
• Operating  
–60 to 3,000 m (–200 to 10,000 ft)  
–60 to 12,000 m (–200 to 40,000 ft)  
• Non-operating  
1.5  
Acoustic Noise  
Table 1.5 lists the acoustic noise specification.  
Table 1.5 Acoustic noise specification  
MPG3307AH-E  
MPG3409AH-E  
MPG3204AH-E  
Model  
Sound Power  
per ISO 7779 and ISO9296  
(Typical at 1m)  
Idle mode (DRIVE READY)  
Seek mode (Random)  
3.6 bels  
4.0 bels  
31 dBA  
35 dBA  
3.3 bels  
3.8 bels  
Sound Pressure  
(Typical at 1m)  
Idle mode (DRIVE READY)  
Seek mode (Random)  
28 dBA  
33 dBA  
1 - 8  
C141-E116-01EN  
1.6  
Shock and Vibration  
Table 1.6 lists the shock and vibration specification.  
Table 1.6 Shock and vibration specification  
Vibration (swept sine, one octave per minute)  
• Operating  
4.9 m/s2 (0.5G0-P); 5 to 300 Hz  
(without non-recovered errors)  
• Non-operating  
39.2 m/s2 (4.0G0-P); 5 to 400 Hz (no damage)  
Shock (half-sine pulse, Operating)  
• 2 ms duration  
392 m/s2 (40G0-P) (without non-recovered error)  
2940 m/s2 (300G0-P) (Typical, no damage)  
Shock (half-sine pulse, Non-operating)  
• 2 ms duration  
1.7  
Reliability  
(1)  
Mean time between failures (MTBF)  
The mean time between failures (MTBF) is 500,000 POH (power on hours) or more (operation: 24  
hours/day, 7 days/week).  
This does not include failures occurring during the first three months after installation.  
MTBF is defined as follows:  
Total operation time in all fields  
MTBF=  
(H)  
number of device failure in all fields  
"Disk drive defects" refers to defects that involve repair, readjustment, or replacement. Disk drive  
defects do not include failures caused by external factors, such as damage caused by handling,  
inappropriate operating environments, defects in the power supply host system, or interface cable.  
(2)  
(3)  
Mean time to repair (MTTR)  
The mean time to repair (MTTR) is 30 minutes or less, if repaired by a specialist maintenance staff  
member.  
CSS cycle  
The number of CSS must be less than 50,000.  
C141-E116-01EN  
1 - 9  
(4)  
Service life  
In situations where management and handling are correct, the disk drive requires no overhaul for  
five years when the DE surface temperature is less than 48°C. When the DE surface temperature  
exceeds 48°C, the disk drives requires no overhaul for five years or 20,000 hours of operation,  
whichever occurs first. Refer to item (3) in Subsection 3.3 for the measurement point of the DE  
surface temperature.  
(5)  
1.8  
(1)  
Data assurance in the event of power failure  
Except for the data block being written to, the data on the disk media is assured in the event of any  
power supply abnormalities. This does not include power supply abnormalities during disk media  
initialization (formatting) or processing of defects (alternative block assignment).  
Error Rate  
Known defects, for which alternative blocks can be assigned, are not included in the error rate  
count below. It is assumed that the data blocks to be accessed are evenly distributed on the disk  
media.  
Unrecoverable read error  
Read errors that cannot be recovered by read retries without user's retry and ECC corrections shall  
occur no more than 10 times when reading data of 1015 bits. Read retries are executed according to  
the disk drive's error recovery procedure, and include read retries accompanying head offset  
operations.  
(2)  
Positioning error  
Positioning (seek) errors that can be recovered by one retry shall occur no more than 10 times in  
107 seek operations.  
1.9  
Media Defects  
Defective sectors are replaced with alternates when the disk is formatted prior to shipment from the  
factory (low level format). Thus, the host sees a defect-free device.  
Alternate sectors are automatically accessed by the disk drive. The user need not be concerned  
with access to alternate sectors.  
Chapter 6 describes the low level format at shipping.  
1 - 10  
C141-E116-01EN  
CHAPTER 2  
DEVICE CONFIGURATION  
2.1  
2.2  
Device Configuration  
System Configuration  
2.1  
Device Configuration  
Figure 2.1 shows the disk drive. The disk drive consists of a disk enclosure (DE), read/write  
preamplifier, and controller PCA. The disk enclosure contains the disk media, heads, spindle  
motors actuators, and a circulating air filter.  
Figure 2.1 Disk drive outerview  
C141-E116-01EN  
2 - 1  
(1)  
Disk  
The outer diameter of the disk is 95 mm. The inner diameter is 25 mm. The number of disks used  
varies with the model, as described below. The disks are rated at over 50,000 start/stop  
operations.  
MPG3204AH-E: 1 disks  
MPG3307AH-E: 2 disks  
MPG3409AH-E: 2 disks  
(2)  
Head  
The heads are of the contact start/stop (CSS) type. The head touches the disk surface while the  
disk is not rotating and automatically lifts when the disk starts.  
(3)  
(4)  
Spindle motor  
The disks are rotated by a direct drive Hall-less DC motor.  
Actuator  
The actuator uses a revolving voice coil motor (VCM) structure which consumes low power and  
generates very little heat. The head assembly at the tip of the actuator arm is controlled and  
positioned by feedback of the servo information read by the read/write head. If the power is not on  
or if the spindle motor is stopped, the head assembly stays in the specific CSS zone on the disk and  
is fixed by a mechanical lock.  
(5)  
Air circulation system  
The disk enclosure (DE) is sealed to prevent dust and dirt from entering. The disk enclosure  
features a closed loop air circulation system that relies on the blower effect of the rotating disk.  
This system continuously circulates the air through the recirculation filter to maintain the  
cleanliness of the air in the disk enclosure.  
(6)  
(7)  
Read/write circuit  
The read/write circuit uses a LSI chip for the read/write preamplifier. It improves data reliability  
by preventing errors caused by external noise.  
Controller circuit  
The controller circuit consists of a LSI chip to improve reliability. The high-speed microprocessor  
unit (MPU) achieves a high-performance AT controller.  
2 - 2  
C141-E116-01EN  
2.2  
System Configuration  
ATA interface  
2.2.1  
Figures 2.2 and 2.3 show the ATA interface system configuration. The drive has a 40-pin PC AT  
interface connector and supports the PIO transfer till 16.7 MB/s (PIO mode 4), the DMA transfer  
till 16.7 MB/s (Multiword DMA mode 2), the ultra DMA transfer till 66.6 MB/s (Ultra DMA  
mode 4), and the ultra DMA transfer till 100 MB/s (Ultra mode 5).  
2.2.2  
1 drive connection  
HA  
Host  
Disk drive  
(Host adaptor)  
ATA interface  
AT bus  
(Host interface)  
Figure 2.2 1 drive system configuration  
2.2.3  
2 drives connection  
HA  
Host  
Disk drive #0  
(Host adaptor)  
AT bus  
(Host interface)  
Disk drive #1  
ATA interface  
Note:  
When the drive that is not conformed to ATA is connected to the disk drive is above  
configuration, the operation is not guaranteed.  
Figure 2.3 2 drives configuration  
C141-E116-01EN  
2 - 3  
IMPORTANT  
HA (host adapter) consists of address decoder, driver, and receiver.  
ATA is an abbreviation of "AT attachment". The disk drive is  
conformed to the ATA-5 interface.  
At high speed data transfer (PIO mode 3, mode 4, DMA mode 2, ultra  
DMA mode 4 or ultra DMA mode 5), occurrence of ringing or  
crosstalk of the signal lines (AT bus) between the HA and the disk drive  
may be a great cause of the obstruction of system reliability. Thus, it is  
necessary that the capacitance of the signal lines including the HA and  
cable does not exceed the ATA-3, ATA-4 and ATA-5 standard, and the  
cable length between the HA and the disk drive should be as short as  
possible.  
2 - 4  
C141-E116-01EN  
CHAPTER 3  
INSTALLATION CONDITIONS  
3.1  
3.2  
3.3  
3.4  
3.5  
Dimensions  
Handling Cautions  
Mounting  
Cable Connections  
Jumper Settings  
3.1  
Dimensions  
Figure 3.1 illustrates the dimensions of the disk drive and positions of the mounting screw holes.  
All dimensions are in mm.  
C141-E116-01EN  
3 - 1  
Figure 3.1 Dimensions  
3 - 2  
C141-E116-01EN  
3.2  
Handling Cautions  
Please keep the following cautions, and handle the HDD under the safety environment.  
3.2.1  
General notes  
ESD mat  
Shock absorbing mat  
Wrist strap  
Place the shock absorbing mat on the  
operation table, and place ESD mat on it.  
Use the Wrist strap.  
Do not hit HDD each other.  
Do not stack when carrying.  
Do not place HDD vertically  
to avoid falling down.  
Do not drop.  
Figure 3.2 Handling cautions  
3.2.2  
Installation  
(1)  
Please use the driver of a low impact when you use an electric driver.  
HDD is occasionally damaged by the impact of the driver.  
(2)  
Please observe the tightening torque of the screw strictly.  
6-32UNC ······· Max. 0.59 N·m (6 Kg·cm)  
3.2.3  
Recommended equipments  
Contents  
Model  
JX-1200-3056-8  
76000DES (ASK7876)  
SS-3000  
Maker  
SUMITOMO 3M  
COMKYLE  
HIOS  
ESD  
Wrist strap  
ESD mat  
Shock  
Low shock driver  
C141-E116-01EN  
3 - 3  
3.3  
Mounting  
(1)  
Direction  
Figure 3.3 illustrates normal direction for the disk drive. The disk drives can be mounted in any  
direction.  
Horizontal mounting with the PCB facing down  
Figure 3.3 Direction  
(2)  
Frame  
The disk enclosure (DE) body is connected to signal ground (SG) and the mounting frame is also  
connected to signal ground. These are electrically shorted.  
Note:  
Use No.6-32UNC screw for the mounting screw and the screw length should satisfy the  
specification in Figure 3.5.  
(3)  
Limitation of side-mounting  
When the disk drive is mounted using the screw holes on both side of the disk drive, use two screw  
holes shown in Figure 3.4.  
Do not use the center hole. For screw length, see Figure 3.5.  
3 - 4  
C141-E116-01EN  
Use these screw  
holes  
Do not use this screw holes  
Figure 3.4 Limitation of side-mounting  
Side surface  
mounting  
2.5  
2.5  
2.5  
Bottom surface mounting  
DE  
DE  
2
PCA  
B
A
Frame of system  
cabinet  
Frame of system  
cabinet  
4.5 or  
less  
Screw  
Screw  
5.0 or less  
Details of A  
Details of B  
Figure 3.5 Mounting frame structure  
C141-E116-01EN  
3 - 5  
(4)  
Ambient temperature  
The temperature conditions for a disk drive mounted in a cabinet refer to the ambient temperature  
at a point 3 cm from the disk drive. Pay attention to the air flow to prevent the DE surface  
temperature from exceeding 60°C.  
Provide air circulation in the cabinet such that the PCA side, in particular, receives sufficient  
cooling. To check the cooling efficiency, measure the surface temperatures of the DE. Regardless  
of the ambient temperature, this surface temperature must meet the standards listed in Table 3.1.  
Figure 3.6 shows the temperature measurement point.  
1
Figure 3.6 Surface temperature measurement points  
Table 3.1 Surface temperature measurement points and standard values  
No.  
1
Measurement point  
Temperature  
60°C max  
DE cover  
3 - 6  
C141-E116-01EN  
(5)  
Service area  
Figure 3.7 shows how the drive must be accessed (service areas) during and after installation.  
- Mounting screw hole  
[Q side]  
- Mounting screw hole  
[P side]  
- Cable connection  
- Mode setting switches  
[R side]  
- Mounting screw hole  
Figure 3.7 Service area  
(6)  
External magnetic fields  
Avoid mounting the disk drive near strong magnetic sources such as loud speakers. Ensure that the  
disk drive is not affected by external magnetic fields.  
C141-E116-01EN  
3 - 7  
3.4  
Cable Connections  
Device connector  
3.4.1  
The disk drive has the connectors and terminals listed below for connecting external devices.  
Figure 3.8 shows the locations of these connectors and terminals.  
Power supply connector (CN1)  
ATA interface connector (CN1)  
Power supply  
connector (CN1)  
Mode  
Setting  
Pins  
ATA  
interface  
connector  
Figure 3.8 Connector locations  
3 - 8  
C141-E116-01EN  
3.4.2  
Cable connector specifications  
Table 3.2 lists the recommended specifications for the cable connectors for Host system that do not  
support Ultra DMA modes greater than mode 2.  
For Host system that support Ultra DMA modes greater than mode 2, the 80-conductor cable  
assemblies shall be used. The 80-conductor cable assemblies are manufactured by AMP or 3M.  
Table 3.2 Cable connector specifications  
Name  
Cable socket  
Model  
Manufacturer  
Fujitsu  
FCN-707B040-AU/B  
(closed-end type)  
ATA interface cable  
(40-pin, CN1)  
Cable socket  
FCN-707B040-AU/O  
Fujitsu  
(through-end type)  
Cable socket housing  
Contact  
1-480424-0  
60617-4  
AMP  
AMP  
Power supply cable  
(CN1)  
Note :  
The cable of twisted pairs and neighboring line separated individually is not allowed to use for  
the host interface cable. It is because that the location of signal lines in these cables is not  
fixed, and so the problem on the crosstalk among signal lines may occur.  
It is recommended to use the ribbon cable for ATA interface that cable length is less than 46  
cm (18 inch) and cable capacitance is less than 35 pico farad. Also it is recommended to use  
AWG18 power supply cable.  
3.4.3  
Device connection  
Figure 3.9 shows how to connect the devices.  
ATA interface cable  
Power supply cable  
Disk Drive #0  
DC  
power supply  
Host system  
Disk Drive #1  
Figure 3.9 Cable connections  
C141-E116-01EN  
3 - 9  
3.4.4  
Power supply connector (CN1)  
Figure 3.10 shows the pin assignment of the power supply connector (CN1).  
1
2
3
4
+12VDC  
+12V RETURN  
1
2
3
4
+5V RETURN  
+5VDC  
(Viewed from cable side)  
Figure 3.10 Power supply connector pins (CN1)  
3.4.5  
System configuration for Ultra DMA  
Host system that support Ultra DMA transfer modes greater than mode 2 shall not share I/O ports.  
They shall provide separate drivers and separate receivers for each cable.  
a) The 80-conductor cable assemblies shall be used for systems operating at Ultra DMA modes  
greater than 2. The 80-coductor cable assemblies may be used in place of 40-conductor cable  
assemblies to improve signal quality for data transfer modes that do not require an 80-  
conductor cable assembly. And the 80-conductor cable assembly shall meet the following  
specifications.  
1) The assembly utilizes a fine pitch cable to double the number of conductors available to  
the 40-pin connector. The grounds assigned by the interface are commoned with the  
additional 40 conductors to provide a ground between each signal line and provide the  
effect of a common ground plane.  
2) The cable assembly may contain up to 3 connectors which shall be uniquely colored as  
follows. All connectors shall have position 20 blocked.  
The System Board Connector shall have a Blue base and Black retainer. Pin 34  
(PDIAG-: CBLID-) shall be connected to ground and shall not be wired to the cable  
assembly.  
Connector Device “0” shall have a Black base and Black retainer.  
Connector Device “1” shall have a Gray base and Black retainer. Pin 28 (CSEL)  
shall not be connected to the cable (contact 28 may be removed to meet this  
requirement).  
The cable assembly may be printed with connector identifiers.  
3) Typical cable characteristics are shown as follows.  
Cable: AWG 30 (pitch: 0.635 mm)  
Single Ended impedance: typical 80 .  
Cable capacitance: typical 57 pF/m  
4) The dimensions are shown in Figure 3.11.  
3 - 10  
C141-E116-01EN  
254.0 to 457.2 mm  
(10 to 18 inch)  
127.0 to 304.8 mm  
(5 to 12 inch)  
101.6 to 152.4 mm  
(4 to 6 inch)  
Pin 40 (Ground)  
Pin 34  
open  
Pin 34 contact  
(PDIAG-:CBLID- signal)  
Pin 30 (Ground)  
Symbolizes Pin 34  
Conductor being cut  
Pin 26 (Ground)  
Pin 24 (Ground)  
Pin 22 (Ground)  
Pin 19 (Ground)  
Position 1  
Pin 2 (Ground)  
System Board  
Connector  
Connector 1  
Connector 2  
Figure 3.11 Cable configuration  
b) Host system that do support Ultra DMA modes greater than mode 2 shall either connect  
directly to the device without using a cable assembly, or determine the cable assembly type.  
Determining the cable assembly type may be done either by the host sensing the condition of  
the PDIAG-:CBLID- signal (see Figure 3.12), or by relying on information from the device  
(see Figure 3.13). Hosts that rely on information from the device shall have a 0.047 µF  
capacitor connected from the PDIAG-:CBLID- signal to ground. The tolerance on this  
capacitor shall be 20% or less.  
C141-E116-01EN  
3 - 11  
Host detected CBLID- above VIH  
PDIAG-: CBLID- conductor  
Host detected CBLID- below VIL  
open  
PDIAG-: CBLID- conductor  
Host  
Device 1  
Device 0  
Host  
Device 1  
Device 0  
with 40-conductor cable  
with 80-conductor cable  
Figure 3.12 Cable type detection using CBLID- signal  
(Host sensing the condition of the CBLID- signal)  
IDENTIFY DEVICE information  
word 93 bit13:0  
IDENTIFY DEVICE information  
word 93 bit13:1  
Device detected CBLID- below VIL  
Device detected CBLID- above VIH  
open  
PDIAG-:CBLID- conductor  
PDIAG-:CBLID- conductor  
0.047 µF  
±10% or  
±20%  
0.047 µF  
±10% or  
±20%  
Host  
Device 1  
Device 0  
Host  
Device 1  
Device 0  
with 40-conductor cable  
with 80-conductor cable  
Figure 3.13 Cable type detection using IDENTIFY DEVICE data  
(Device sensing the condition of the CBLID- signal)  
3 - 12  
C141-E116-01EN  
3.5  
Jumper Settings  
3.5.1  
Location of setting jumpers  
Figure 3.14 shows the location of the jumpers to select drive configuration and functions.  
DC Power Connector  
Interface Connector  
40  
2
1
1
Figure 3.14 Jumper location  
C141-E116-01EN  
3 - 13  
3.5.2  
Factory default setting  
Figure 3.15 shows the default setting position at the factory. (Master device setting)  
DC Power Connector  
Interface Connector  
Figure 3.15 Factory default setting  
3.5.3  
Jumper configuration  
Device type  
Master device (device #0) or slave device (device #1) is selected.  
(1)  
2
4
6 8  
2
4
6 8  
= shorted  
1
3
5 7  
9
1
3
5 7  
9
(a) Master device  
(b) Slave device  
Figure 3.16 Jumper setting of master or slave device  
Note:  
When the device type is set by the jumper on the device, the device should not be configured  
for cable selection.  
(2)  
Cable Select (CSEL)  
In Cable Select mode, the device can be configured either master device or slave device. For use of  
Cable Select function, Unique interface cable is needed.  
3 - 14  
C141-E116-01EN  
2
1
4
3
6 8  
5 7  
9
CSEL connected to the interface cable selection  
can be done by the special interface cable.  
Figure 3.17 Jumper setting of Cable Select  
Figures 3.18 and 3.19 show examples of cable selection using unique interface cables.  
By connecting the CSEL of the master device to the CSEL Line (conductor) of the cable and  
connecting it to ground further, the CSEL is set to low level. The device is identified as a master  
device. At this time, the CSEL of the slave device does not have a conductor. Thus, since the  
slave device is not connected to the CSEL conductor, the CSEL is set to high level. The device is  
identified as a slave device.  
CSEL conductor  
Open  
GND  
Host system  
Master device  
Slave device  
Figure 3.18 Example (1) of Cable Select  
CSEL conductor  
Open  
GND  
Host system  
Slave device  
Master device  
Figure 3.19 Example (2) of Cable Select  
C141-E116-01EN  
3 - 15  
(3)  
Special jumper settings  
(a) 2.1 GB clip (Limit capacity to 2.1 GB) / 33.8 GB clip (Limit capacity to 33.8 GB)  
If the drive cannot be recognized by system with legacy BIOS’s which do not allow single  
volume sizes greater than approximately 2.1 GB, the following jumper settings should be  
applied.  
This jumper settings is also used as the 33.8 GB clip for MPG3409AH-E. (MPG3409AH-E  
does not have the 2.1 GB clip feature.)  
2
4
6 8  
2
4
6 8  
2
4
6 8  
1
3
5 7  
9
1
3
5 7  
9
1
3
5 7  
9
Master Device  
Slave Device  
Cable Select  
Model  
No. of cylinders  
4,092  
No. of heads  
No. of sectors  
Capacity  
2.1 GB  
2.1 GB  
33.8 GB  
MPG3204AH-E  
MPG3307AH-E  
MPG3409AH-E  
16  
16  
16  
63  
63  
63  
4,092  
16,383  
(b) Slave present  
If the slave device does not use the Device Active/Slave Present (DASP–) signal to indicate its  
presence, the device is configured as a Master with slave present when the following jumper  
settings is applied.  
2
4
6 8  
1
3
5 7  
9
Slave present  
NOTE:  
The following Jumper Plug is the recommended specification for jumper setting on this device.  
Parts name  
Parts number  
IMAS-9251H-GF  
206-A-BLK  
Manufacturer  
Remarks  
2.54 mm Pitch  
! 0.64 mm  
IRISO ELECTRONICS CO., LTD  
OUPIIN ENTERPRISE CO., LTD  
Jumper Plug  
3 - 16  
C141-E116-01EN  
CHAPTER 4  
THEORY OF DEVICE OPERATION  
4.1  
4.2  
4.3  
4.4  
4.5  
4.6  
4.7  
Outline  
Subassemblies  
Circuit Configuration  
Power-on sequence  
Self-calibration  
Read/write Circuit  
Servo Control  
This chapter explains basic design concepts of the disk drive. Also, this chapter explains subassemblies of  
the disk drive, each sequence, servo control, and electrical circuit blocks.  
4.1  
Outline  
This chapter consists of two parts. First part (Section 4.2) explains mechanical assemblies of the  
disk drive. Second part (Sections 4.3 through 4.7) explains a servo information recorded in the  
disk drive and drive control method.  
4.2  
Subassemblies  
The disk drive consists of a disk enclosure (DE) and printed circuit assembly (PCA).  
The DE contains all movable parts in the disk drive, including the disk, spindle, actuator,  
read/write head, and air filter. For details, see Subsections 4.2.1 to 4.2.5.  
The PCA contains the control circuits for the disk drive. The disk drive has one PCA. For details,  
see Sections 4.3.  
4.2.1  
Disk  
The DE contains the disks with an outer diameter of 95 mm. The MPG3204AH-E has 1 disk, and  
MPG3307AH-E and MPG3409AH-E have 2 disk.  
The head contacts the disk each time the disk rotation stops; the life of the disk is 50,000 contacts  
or more.  
Servo data is recorded on each cylinder (total 96). Servo data written at factory is read out by the  
read/write head. For servo data, see Section 4.7.  
C141-E116-01EN  
4 - 1  
4.2.2  
Head  
Figure 4.1 shows the read/write head structures. The Numerals 0 to 3 indicate read/write heads.  
These heads are raised from the disk surface as the spindle motor approaches the rated rotation  
speed.  
MPG3204AH-E  
Spindle  
Actuator  
1
0
MPG3307AH-E / MPG3409AH-E  
Spindle Actuator  
3
2
1
0
Figure 4.1 Head structure  
4.2.3  
Spindle  
The spindle consists of a disk stack assembly and spindle motor. The disk stack assembly is  
activated by the direct drive sensor-less DC spindle motor, which has a speed of 7,200 rpm. The  
spindle is controlled with detecting a PHASE signal generated by counter electromotive voltage of  
the spindle motor at starting. After that, the rotational speed is kept with detecting a servo  
information.  
4 - 2  
C141-E116-01EN  
4.2.4  
4.2.5  
Actuator  
The actuator consists of a voice coil motor (VCM) and a head carriage. The VCM moves the head  
carriage along the inner or outer edge of the disk. The head carriage position is controlled by  
feeding back the difference of the target position that is detected and reproduced from the servo  
information read by the read/write head.  
Air filter  
There are two types of air filters: a breather filter and a circulation filter.  
The breather filter makes an air in and out of the DE to prevent unnecessary pressure around the  
spindle when the disk starts or stops rotating. When disk drives are transported under conditions  
where the air pressure changes a lot, filtered air is circulated in the DE.  
The circulation filter cleans out dust and dirt from inside the DE. The disk drive cycles air  
continuously through the circulation filter through an enclosed loop air cycle system operated by a  
blower on the rotating disk.  
C141-E116-01EN  
4 - 3  
4.3  
Circuit Configuration  
Figure 4.2 shows the disk drive circuit configuration.  
(1)  
Read/write circuit  
The read/write circuit consists of two LSIs; read/write preamplifier (PreAMP) and read channel  
(RDC).  
The PreAMP consists of the write current switch circuit, that flows the write current to the head  
coil, and the voltage amplifier circuit, that amplitudes the read output from the head.  
The RDC is the read demodulation circuit using the Extended Partial Response Class 4 (EPR4),  
and contains the Viterbi detector, programmable filter, adaptable transversal filter, times base  
generator, and data separator circuits. The RDC also contains the 48/52 group coded recording  
(GCR) encoder and decoder and servo demodulation circuit.  
(2)  
Servo circuit  
The position and speed of the voice coil motor are controlled by 2 closed-loop servo using the  
servo information recorded on the data surface. The servo information is an analog signal  
converted to digital for processing by a MPU and then reconverted to an analog signal for control  
of the voice coil motor.  
(3)  
(4)  
Spindle motor driver circuit  
The circuit measures the interval of a PHASE signal generated by counter-electromotive voltage of  
a motor, or servo mark at the MPU and controls the motor speed comparing target speed.  
Controller circuit  
Major functions are listed below.  
Data buffer management  
ATA interface control and data transfer control  
Sector format control  
Defect management  
ECC control  
Error recovery and self-diagnosis  
4 - 4  
C141-E116-01EN  
ATA Interface  
PCA  
HDC & MCU & RDC  
CL-SH8671 (Himalaya-2)  
Data Buffer  
SDRAM  
MCU  
ARM7TDMI  
HDC  
CL-SH7660  
Flash ROM  
FROM  
RDC  
Number-B  
SVC  
Resonator  
20MHz  
HA13627  
DE  
Temp.  
Sensor  
SP Motor  
Media  
VCM  
R/W Pre-Amp  
HEAD  
SR1756 or  
M61850  
Figure 4.2 MPG3xxxAH Block diagram  
C141-E116-01EN  
4 - 5  
4.4  
Power-on Sequence  
Figure 4.3 describes the operation sequence of the disk drive at power-on. The outline is described  
below.  
a) After the power is turned on, the disk drive executes the MPU bus test, internal register  
read/write test, and work RAM read/write test. When the self-diagnosis terminates  
successfully, the disk drive starts the spindle motor.  
b) The disk drive executes self-diagnosis (data buffer read/write test) after enabling response to  
the ATA bus.  
c) After confirming that the spindle motor has reached rated speed, the disk drive releases the  
heads from the actuator magnet lock mechanism by applying current to the VCM. This  
unlocks the heads which are parked at the inner circumference of the disks.  
d) The disk drive positions the heads onto the SA area and reads out the system information.  
e) The disk drive executes self-seek-calibration. This collects data for VCM torque and  
mechanical external forces applied to the actuator, and updates the calibrating value.  
f) The drive becomes ready. The host can issue commands.  
4 - 6  
C141-E116-01EN  
Power on  
Start  
Self-diagnosis 1  
• MPU bus test  
• Inner register  
write/read test  
• Work RAM write/read  
test  
a)  
The spindle motor starts.  
Self-diagnosis 2  
• Data buffer write/read  
test  
b)  
c)  
Confirming spindle motor  
speed  
Release heads from  
actuator lock  
d)  
Initial on-track and read  
out of system information  
e)  
f)  
Execute self-calibration  
Drive ready state  
(command waiting state)  
End  
Figure 4.3 Power-on operation sequence  
C141-E116-01EN  
4 - 7  
4.5  
Self-calibration  
The disk drive occasionally performs self-calibration in order to sense and calibrate mechanical  
external forces on the actuator, and VCM torque. This enables precise seek and read/write  
operations.  
4.5.1  
Self-calibration contents  
(1)  
Sensing and compensating for external forces  
The actuator suffers from torque due to the FPC forces and winds accompanying disk revolution.  
The torque vary with the disk drive and the cylinder where the head is positioned. To execute  
stable fast seek operations, external forces are occasionally sensed.  
The firmware of the drive measures and stores the force (value of the actuator motor drive current)  
that balances the torque for stopping head stably. This includes the current offset in the power  
amplifier circuit and DAC system.  
The forces are compensated by adding the measured value to the specified current value to the  
power amplifier. This makes the stable servo control.  
To compensate torque varying by the cylinder, the disk is divided into 28 areas from the innermost  
to the outermost circumference and the compensating value is measured at the measuring cylinder  
on each area at factory calibration. The measured values are stored in the SA cylinder. In the self-  
calibration, the compensating value is updated using the value in the SA cylinder.  
(2)  
Compensating open loop gain  
Torque constant value of the VCM has a dispersion for each drive, and varies depending on the  
cylinder that the head is positioned. To realize the high speed seek operation, the value that  
compensates torque constant value change and loop gain change of the whole servo system due to  
temperature change is measured and stored.  
For sensing, the firmware mixes the disturbance signal to the position signal at the state that the  
head is positioned to any cylinder. The firmware calculates the loop gain from the position signal  
and stores the compensation value against to the target gain as ratio.  
For compensating, the direction current value to the power amplifier is multiplied by the  
compensation value. By this compensation, loop gain becomes constant value and the stable servo  
control is realized.  
To compensate torque constant value change depending on cylinder, whole cylinders from most  
inner to most outer cylinder are divided into 14 partitions at calibration in the factory, and the  
compensation data is measured for representative cylinder of each partition. This measured value  
is stored in the SA area. The compensation value at self-calibration is calculated using the value in  
the SA area.  
4 - 8  
C141-E116-01EN  
4.5.2  
Execution timing of self-calibration  
Self-calibration is executed when:  
The power is turned on.  
The self-calibration execution timechart of the disk drive specifies self-calibration.  
The disk drive performs self-calibration according to the timechart based on the time elapsed from  
power-on. After power-on, self-calibration is performed about every 30 minutes, when the host  
command is not issued for 15 seconds.  
4.5.3  
Command processing during self-calibration  
If the disk drive receives a command execution request from the host while executing self-  
calibration according to the timechart, the disk drive terminates self-calibration and starts  
executing the command precedingly. In other words, if a disk read or write service is necessary,  
the disk drive positions the head to the track requested by the host, reads or writes data. Then  
restarts calibration if the host command is not issued for 15 seconds.  
This enables the host to execute the command without waiting for a long time, even when the disk  
drive is performing self-calibration. Only the first command execution wait time is about  
maximum 100 ms.  
C141-E116-01EN  
4 - 9  
4.6  
Read/write Circuit  
The read/write circuit consists of the read/write preamplifier (PreAMP), the write circuit, the read  
circuit, and the synthesizer in the read channel (RDC).  
4.6.1  
Read/write preamplifier (PreAMP)  
One PreAMP is mounted on the FPC. The PreAMP consists of a 4-channel read preamplifier and  
a write current switch and senses a write error. Each channel is connected to each data head. The  
head IC switches the heads by the serial port (SDEN, SCLK, SDATA). The IC generates a write  
error sense signal (WUS) when a write error occurs due to head short-circuit or head  
disconnection.  
4.6.2  
Write circuit  
The write data is output from the hard disk controller (HDC) with the NRZ data format, and sent to  
the encoder circuit in the RDC with synchronizing with the write clock. The NRZ write data is  
converted from 48-bits data to 52-bits data by the encoder circuit then sent to the PreAMP, and the  
data is written onto the media.  
(1)  
(2)  
48/52 GCR  
The disk drive converts data using the 48/52 group coded recording (GCR) algorithm.  
Write precompensation  
Write precompensation compensates, during a write process, for write non-linearity generated at  
reading.  
4 - 10  
C141-E116-01EN  
4.6.3  
Read circuit  
The head read signal from the PreAMP is regulated by the variable gain amplifier (VGA) circuit.  
Then the output is converted into the sampled read data pulse by the programmable filter circuit  
and the adaptive equalizer circuit. This clock signal is converted into the NRZ data by the 48/52  
GCR decoder circuit based on the read data maximum-likelihood-detected by the Viterbi detection  
circuit, then is sent to the HDC.  
(1)  
(2)  
VGA circuit  
The VGA circuit automatically regulates the output amplitude to a constant value even when the  
input amplitude level fluctuates. The VGA output is maintained at a constant level even when the  
head output fluctuates due to the head characteristics or outer/inner head positions.  
Programmable filter  
The programmable filter circuit has a low-pass filter function that eliminates unnecessary high  
frequency noise component and a high frequency boost-up function that equalizes the waveform of  
the read signal.  
Cut-off frequency of the low-pass filter and boost-up gain are controlled from each DAC circuit in  
read channel. The MPU optimizes the cut-off frequency and boost-up gain according to the  
transfer frequency of each zone.  
(3)  
FIR (Digital Finite Impulse Response Equalization Filter) adaptation circuit  
The FIR provides support for changing equalization needs from head to head and zone to zone.  
The FIR is a specialized digital filter with ten independently controlled coefficients.  
(4)  
(5)  
Viterbi detection circuit  
The Viterbi detection circuit demodulates data according to the survivor path sequence.  
Data separator circuit  
The data separator circuit generates clocks in synchronization with the output of the adaptive  
equalizer circuit. To write data, the VFO circuit generates clocks in synchronization with the clock  
signals from a synthesizer.  
(6)  
48/52 GCR decoder  
This circuit converts the 52-bits read data into the 48-bits NRZ data.  
C141-E116-01EN  
4 - 11  
4.6.4  
Synthesizer circuit  
The drive uses constant density recording to increase total capacity. This is different from the  
conventional method of recording data with a fixed data transfer rate at all data area. In the  
constant density recording method, data area is divided into zones by radius and the data transfer  
rate is set so that the recording density of the inner cylinder of each zone is nearly constant. The  
drive divides data area into 15 zones to set the data transfer rate. Table 4.1 describes the data  
transfer rate and recording density (BPI) of each zone.  
Table 4.1 Transfer rate of each zone  
MPG3307AH-E  
Zone  
0
1
2
3
4
5
6
7
Cylinder  
0
to  
2208  
to  
4224  
to  
7232  
to  
9824  
to  
13120 15616 17088  
to to to  
2207  
4223  
7231  
9823  
13119 15615 17087 19359  
Transfer rate  
[MB/s]  
49.73  
48.87  
47.22  
45.79  
43.61  
41.88  
41.01  
39.28  
Zone  
8
9
10  
11  
12  
13  
14  
Cylinder  
19360 22048 23840 25248 26592 28768 30176  
to to to to to to to  
22047 23839 25247 26591 28767 30175 30783  
Transfer rate  
[MB/s]  
37.22  
35.77  
34.62  
33.45  
31.43  
30.12  
27.92  
MPG3204AH-E / MPG3409AH-E  
Zone  
Cylinder  
0
1
2
3
4
5
6
7
0
to  
2784  
to  
5568  
to  
8352  
to  
11264 13440 16096 17984  
to to to to  
2783  
5567  
8351  
11263 13439 16095 17983 19839  
Transfer rate  
[MB/s]  
60.71  
60.71  
60.71  
60.71  
59.22  
56.86  
54.90  
52.94  
Zone  
8
9
10  
11  
12  
13  
14  
Cylinder  
19840 22176 23584 25088 26720 28576 29952  
to to to to to to to  
22175 23583 25087 26719 28575 29951 30783  
Transfer rate  
[MB/s]  
50.35  
48.82  
47.06  
45.10  
42.75  
40.94  
38.04  
The MPU transfers the data transfer rate setup data to the RDC that includes the time base  
generator circuit to change the data transfer rate.  
4 - 12  
C141-E116-01EN  
4.7  
Servo Control  
The actuator motor and the spindle motor are submitted to servo control. The actuator motor is  
controlled for moving and positioning the head to the track containing the desired data. To turn  
the disk at a constant velocity, the actuator motor is controlled according to the servo data that is  
written on the data side beforehand.  
4.7.1  
Servo control circuit  
Figure 4.4 is the block diagram of the servo control circuit. The following describes the functions  
of the blocks:  
(1)  
MPU  
SVC  
(5)  
(2)  
(3)  
(4)  
DAC  
VCM current  
Servo  
DSP  
unit  
P.  
Amp.  
ADC  
burst  
Head  
capture  
CSR  
Position Sense  
VCM  
(6)  
(7)  
Driver  
Spindle  
motor  
control  
Spindle  
motor  
CSR: Current Sense Resistor  
VCM: Voice Coil Motor  
Figure 4.4 Block diagram of servo control circuit  
(1)  
Microprocessor unit (MPU)  
The MPU includes DSP unit, etc., and the MPU starts the spindle motor, moves the heads to the  
reference cylinders, seeks the specified cylinder, and executes calibration according to the internal  
operations of the MPU.  
The major internal operations are listed below.  
a. Spindle motor start  
Starts the spindle motor and accelerates it to normal speed when power is applied.  
C141-E116-01EN  
4 - 13  
b. Move head to reference cylinder  
Drives the VCM to position the head at the any cylinder in the data area. The logical initial  
cylinder is at the outermost circumference (cylinder 0).  
c. Seek to specified cylinder  
Drives the VCM to position the head to the specified cylinder.  
d. Calibration  
Senses and stores the thermal offset between heads and the mechanical forces on the actuator,  
and stores the calibration value.  
4 - 14  
C141-E116-01EN  
Servo frame  
(96 servo frames per  
revolution)  
Figure 4.5 Physical sector servo configuration on disk surface  
Servo burst capture circuit  
(2)  
(3)  
The four servo signals can be synchronously detected by the DEMOD signal (internal), full-wave  
rectified integrated.  
A/D converter (ADC)  
The A/D converter (ADC) receives the servo signals are integrated, converts them to digital, and  
transfers the digital signal to the DSP unit.  
C141-E116-01EN  
4 - 15  
(4)  
D/A converter (DAC)  
The D/A converter (DAC) converts the VCM drive current value (digital value) calculated by the  
DSP unit into analog values and transfers them to the power amplifier.  
(5)  
(6)  
Power amplifier  
The power amplifier feeds currents, corresponding to the DAC output signal voltage to the VCM.  
Spindle motor control circuit  
The spindle motor control circuit controls the sensor-less spindle motor. This circuit detects  
number of revolution of the motor by the interrupt generated periodically, compares with the target  
revolution speed, then flows the current into the motor coil according to the differentiation  
(aberration).  
(7)  
Driver circuit  
The driver circuit is a power amplitude circuit that receives signals from the spindle motor control  
circuit and feeds currents to the spindle motor.  
(8)  
VCM current sense resistor (CSR)  
This resistor controls current at the power amplifier by converting the VCM current into voltage  
and feeding back.  
4.7.2  
(1)  
Data-surface servo format  
Figure 4.5 describes the physical layout of the servo frame. The three areas indicated by (1) to (3)  
in Figure 4.6 are described below.  
Inner guard band  
The head is in contact with the disk in this space when the spindle starts turning or stops, and the  
rotational speed of the spindle can be controlled on this cylinder area for head moving.  
(2)  
(3)  
Data area  
This area is used as the user data area and SA area.  
Outer guard band  
This area is located at outer position of the user data area, and the rotational speed of the spindle  
can be controlled on this cylinder area for head moving.  
4 - 16  
C141-E116-01EN  
4.7.3  
Servo frame format  
As the servo information, the drive uses the two-phase servo generated from the gray code and Pos  
A to D. This servo information is used for positioning operation of radius direction and position  
detection of circumstance direction.  
The servo frame consists of 5 blocks; write/read recovery, servo mark, gray code, Pos A to D and  
PAD. Figure 4.6 shows the servo frame format.  
0.17 µs  
0.53 µs 0.74 µs 0.56 µs 0.56 µs  
0.16 µs  
0.80 µs  
0.72 µs  
1.81 µs  
0.56 µs  
Pos  
A
Pos  
B
Pos  
D
Pos  
C
PAD  
SCD  
PA  
ASM  
SSM  
R/W Recovery Field  
6.63 µs  
Servo  
Frame  
Servo  
Frame  
DATA  
DATA  
86.81 µs  
Figure 4.6 96 servo frames in each track  
C141-E116-01EN  
4 - 17  
(1)  
(2)  
Write/read recovery  
This area is used to absorb the write/read transient and to stabilize the AGC.  
Servo mark (ASM, SSM)  
This area generates a timing for demodulating the gray code and position-demodulating Pos A to D  
by detecting the servo mark.  
(3)  
(4)  
(5)  
Preamble  
This area is used to synchronize with the PLL, which is used to search the SSM by detecting the  
ASM.  
Gray code (including index bit) (SCD)  
This area is used as cylinder address. The data in this area is converted into the binary data by the  
gray code demodulation circuit.  
Pos A, Pos B, Pos C, Pos D  
This area is used as position signals between tracks, and the device control at on-track so that Pos  
A level equals to Pos B level.  
(6)  
PAD  
This area is used as a gap between servo and data.  
4.7.4  
Actuator motor control  
The voice coil motor (VCM) is controlled by feeding back the servo data recorded on the data  
surface. The MPU fetches the position sense data on the servo frame at a constant interval of  
sampling time, executes calculation, and updates the VCM drive current.  
The servo control of the actuator includes the operation to move the head to the reference cylinder,  
the seek operation to move the head to the target cylinder to read or write data, and the track-  
following operation to position the head onto the target track.  
(1)  
Operation to move the head to the reference cylinder  
The MPU moves the head to the reference cylinder when the power is turned. The reference  
cylinder is in the data area.  
When power is applied the heads are moved from the inner circumference shunt zone to the normal  
servo data zone in the following sequence:  
a) Micro current is fed to the VCM to press the head against the inner circumference.  
4 - 18  
C141-E116-01EN  
b) A current is fed to the VCM to move the head toward the outer circumference.  
c) When the servo mark is detected the head is moved slowly toward the outer circumference at a  
constant speed.  
d) If the head is stopped at the reference cylinder from there. Track following control starts.  
Seek operation  
(2)  
Upon a data read/write request from the host, the MPU confirms the necessity of access to the disk.  
If a read or instruction is issued, the MPU seeks the desired track.  
The MPU feeds the VCM current via the D/A converter and power amplifier to move the head.  
The MPU calculates the difference (speed error) between the specified target position and the  
current position for each sampling timing during head moving. The MPU then feeds the VCM  
drive current by setting the calculated result into the D/A converter. The calculation is digitally  
executed by the firmware. When the head arrives at the target cylinder, the track is followed.  
(3)  
Track following operation  
Except during head movement to the reference cylinder and seek operation under the spindle  
rotates in steady speed, the MPU does track following control. To position the head at the center  
of a track, the DSP drives the VCM by feeding micro current. For each sampling time, the VCM  
drive current is determined by filtering the position difference between the target position and the  
position clarified by the detected position sense data. The filtering includes servo compensation.  
These are digitally controlled by the firmware.  
4.7.5  
Spindle motor control  
Hall-less three-phase eight-pole motor is used for the spindle motor, and the 3-phase full/half-wave  
analog current control circuit is used as the spindle motor driver (called SVC hereafter). The  
firmware operates on the MPU manufactured by Fujitsu. The spindle motor is controlled by  
sending several signals from the MPU to the SVC. There are three modes for the spindle control;  
start mode, acceleration mode, and stable rotation mode.  
(1)  
Start mode  
When power is supplied, the spindle motor is started in the following sequence:  
a) After the power is turned on, the MPU sends a signal to the SVC to charge the change pump  
capacitor of the SVC. The charged amount defines the current that flows in the spindle motor.  
b) When the charge pump capacitor is charged enough, the MPU sets the SVC to the motor start  
mode. Then, a starting current flows into the spindle motor.  
c) The SVC generates a phase switching signal by itself, and changes the phase of the current  
flowed in the motor in the order of (V-phase to U-phase), (W-phase to U-phase), (W-phase to  
V-phase), (U-phase to V-phase), (U-phase to W-phase), and (V-phase to W-phase) (after that,  
repeating this order).  
C141-E116-01EN  
4 - 19  
d) During phase switching, the spindle motor starts rotating in low speed, and generates a counter  
electromotive force. The SVC detects this counter electromotive force and reports to the MPU  
using a PHASE signal for speed detection.  
e) The MPU is waiting for a PHASE signal. When no phase signal is sent for a specific period,  
the MPU resets the SVC and starts from the beginning. When a PHASE signal is sent, the  
SVC enters the acceleration mode.  
(2)  
Acceleration mode  
In this mode, the MPU stops to send the phase switching signal to the SVC. The SVC starts a  
phase switching by itself based on the counter electromotive force. Then, rotation of the spindle  
motor accelerates. The MPU calculates a rotational speed of the spindle motor based on the  
PHASE signal from the SVC, and accelerates till the rotational speed reaches 7,200 rpm. When  
the rotational speed reaches 7,200 rpm, the SVC enters the stable rotation mode.  
(3)  
Stable rotation mode  
The MPU calculates a time for one revolution of the spindle motor based on the PHASE signal  
from the SVC. The MPU takes a difference between the current time and a time for one revolution  
at 7,200 rpm that the MPU already recognized. Then, the MPU keeps the rotational speed to 7,200  
rpm by charging or discharging the charge pump for the different time. For example, when the  
actual rotational speed is 7,400 rpm, the time for one revolution is 8.108 ms. And, the time for one  
revolution at 7,200 rpm is 8.333 ms. Therefore, the MPU discharges the charge pump for 0.225  
ms × k (k: constant value). This makes the flowed current into the motor lower and the rotational  
speed down. When the actual rotational speed is later than 7,200 rpm, the MPU charges the pump  
the other way. This control (charging/discharging) is performed every 1/4 revolution.  
4 - 20  
C141-E116-01EN  
CHAPTER 5  
INTERFACE  
5.1  
5.2  
5.3  
5.4  
5.5  
5.6  
Physical Interface  
Logical Interface  
Host Commands  
Command Protocol  
Ultra DMA feature set  
Timing  
C141-E116-01EN  
5 - 1  
5.1  
Physical Interface  
5.1.1  
Interface signals  
Table 5.1 shows the interface signals.  
Table 5.1 Interface signals  
Description  
Host  
Dir  
Dev  
Acrorym  
CSEL  
Cable select  
see note  
Chip select 0  
Chip select 1  
Data bus bit 0  
Data bus bit 1  
Data bus bit 2  
Data bus bit 3  
Data bus bit 4  
Data bus bit 5  
Data bus bit 6  
Data bus bit 7  
Data bus bit 8  
Data bus bit 9  
Data bus bit 10  
Data bus bit 11  
Data bus bit 12  
Data bus bit 13  
Data bus bit 14  
Data bus bit 15  
CS0–  
CS1–  
DD0  
DD1  
DD2  
DD3  
DD4  
DD5  
DD6  
DD7  
DD8  
DD9  
DD10  
DD11  
DD12  
DD13  
DD14  
DD15  
Device active or slave present  
Device address bit 0  
see note  
DASP–  
DA0  
Device address bit 1  
DA1  
Device address bit 2  
DA2  
DMA acknowledge  
DMACK–  
DMARQ  
INTRQ  
DIOR–  
HDMARDY–  
HSTROBE  
IORDY  
DDMARDY–  
DSTROBE  
DIOW–  
STOP  
DMA request  
Interrupt request  
I/O read  
DMA ready during Ultra DMA data in bursts  
Data strobe during Ultra DMA data out bursts  
I/O ready  
DMA ready during Ultra DMA data out bursts  
Data strobe during Ultra DMA data in bursts  
I/O write  
Stop during Ultra DMA data bursts  
Passed diagnostics  
see note  
PDIAG–  
CBLID–  
RESET–  
Cable type detection  
Reset  
Note : See signal descriptions  
5 - 2  
C141-E116-01EN  
5.1.2  
Signal assignment on the connector  
Table 5.2 shows the signal assignment on the interface connector.  
Table 5.2 Signal assignment on the interface connector  
Pin No.  
Signal  
Pin No.  
Signal  
1
3
5
7
RESET–  
DATA7  
DATA6  
DATA5  
DATA4  
DATA3  
DATA2  
DATA1  
DATA0  
GND  
2
4
6
8
GND  
DATA8  
DATA9  
DATA10  
DATA11  
DATA12  
DATA13  
DATA14  
DATA15  
(KEY)  
GND  
GND  
GND  
CSEL  
GND  
reserved  
PDIAG–, CBLID–  
DA2  
CS1–  
GND  
9
10  
12  
14  
16  
18  
20  
22  
24  
26  
28  
30  
32  
34  
36  
38  
40  
11  
13  
15  
17  
19  
21  
23  
25  
27  
29  
31  
33  
35  
37  
39  
DMARQ  
DIOW–, STOP  
DIOR–, HDMARDY–, HSTROBE  
IORDY, DDMARDY–, DSTROBE  
DMACK–  
INTRQ  
DA1  
DA0  
CS0–  
DASP–  
[signal]  
[I/O]  
[Description]  
RESET–  
I
Reset signal from the host. This signal is low active and is asserted  
for a minimum of 25 µs during power on. The device has a 10 kΩ  
pull-up resistor on this signal.  
DATA 0-15  
I/O  
I
Sixteen-bit bi-directional data bus between the host and the device.  
These signals are used for data transfer  
DIOW–, STOP  
DIOW– is the strobe signal asserted by the host to write device  
registers or the data port.  
DIOW– shall be negated by the host prior to initiation of an Ultra  
DMA burst. STOP shall be negated by the host before data is  
transferred in an Ultra DMA burst. Assertion of STOP by the host  
during an Ultra DMA burst signals the termination of the Ultra DMA  
burst.  
C141-E116-01EN  
5 - 3  
[signal]  
DIOR–  
[I/O]  
I
[Description]  
DIOR– is the strobe signal asserted by the host to read device  
registers or the data port.  
HDMARDY–  
I
HDMARDY– is a flow control signal for Ultra DMA data in bursts.  
This signal is asserted by the host to indicate to the device that the  
host is ready to receive Ultra DMA data in bursts.  
The host may negate HDMARDY- to pause an Ultra DMA data in  
burst.  
HSTROBE  
INTRQ  
I
HSTROBE is the data out strobe signal from the host for an Ultra  
DMA data out burst. Both the rising and falling edge of HSTROBE  
latch the data from DATA 0-15 into the device. The host may stop  
generating HSTROBE edges to pause an Ultra DMA data out burst.  
O
Interrupt signal to the host.  
This signal is negated in the following cases:  
– assertion of RESET– signal  
– Reset by SRST of the Device Control register  
– Write to the command register by the host  
– Read of the status register by the host  
– Completion of sector data transfer  
(without reading the Status register)  
When the device is not selected or interrupt is disabled, the INTRQ  
Signal shall be in a high impedance state.  
CS0–  
I
I
I
Chip select signal decoded from the host address bus. This signal is  
used by the host to select the command block registers.  
CS1–  
Chip select signal decoded from the host address bus. This signal is  
used by the host to select the control block registers.  
DA 0-2  
Binary decoded address signals asserted by the host to access task  
file registers.  
KEY  
Key pin for prevention of erroneous connector insertion  
PIDAG–  
I/O  
This signal is an input mode for the master device and an output  
mode for the slave device in a daisy chain configuration. This signal  
indicates that the slave device has been completed self diagnostics.  
This signal is pulled up to +5 V through 10 kresistor at each device.  
CBLID–  
DASP–  
I/O  
I/O  
This signal is used to detect the cable type (80 or 40-conductor  
cable) installed in the system. This signal is pulled up to +5 V  
through 10 kresistor at each device.  
This is a time-multiplexed signal that indicates that the device is  
active and a slave device is present.  
This signal is pulled up to +5 V through 10 kresistor at each device.  
5 - 4  
C141-E116-01EN  
[signal]  
IORDY  
[I/O]  
O
[Description]  
This signal is negated to extend the host transfer cycle of any host register  
access (Read or Write) when the device is not ready to respond to a data  
transfer request.  
DDMARDY–  
DSTROBE  
CSEL  
O
O
I
DDMARDY– is a flow control signal for Ultra DMA data out bursts.  
This signal is asserted by the device to indicate to the host that the device  
is ready to receive Ultra DMA data out bursts. The device may negate  
DDMARDY– to pause an Ultra DMA data out burst.  
DSTROBE is the data in strobe signal from the device for an Ultra  
DMA data in burst. Both the rising and falling edge of DSTROBE  
latch the data from DATA 0-15 into the host. The device may stop  
generating DSTROBE edges to pause an Ultra DMA data in burst.  
This signal to configure the device as a master or a slave device.  
When CSEL signal is grounded, the IDD is a master device.  
When CSEL signal is open, the IDD is a slave device.  
This signal is pulled up with 10 kresistor.  
DMACK–  
DMARQ  
I
The host system asserts this signal as a response that the host system  
receive data or to indicate that data is valid.  
O
This signal is used for DMA transfer between the host system and the  
device. The device asserts this signal when the device completes the  
preparation of DMA data transfer to the host system (at reading) or  
from the host system (at writing).  
The direction of data transfer is controlled by the IOR- and IOW-  
signals. In other word, the device negates the DMARQ signal after  
the host system asserts the DMACK– signal. When there is another  
data to be transferred, the device asserts the DMARQ signal again.  
When the DMA data transfer is performed, IOCW16–, CS0– and  
CS1- signals are not asserted. The DMA data transfer is a 16-bit data  
transfer. The device has a 10 kpull-down resistor on this signal.  
GND  
Grounded  
Note:  
"I" indicates input signal from the host to the device.  
"O" indicates output signal from the device to the host.  
"I/O" indicates common output or bi-directional signal between the host and the device.  
C141-E116-01EN  
5 - 5  
5.2  
Logical Interface  
The device can operate for command execution in either address-specified mode; cylinder-head-  
sector (CHS) or Logical block address (LBA) mode. The IDENTIFY DEVICE information  
indicates whether the device supports the LBA mode. When the host system specifies the LBA  
mode by setting bit 6 in the Device/Head register to 1, HS3 to HS0 bits of the Device/Head  
register indicates the head No. under the LBA mode, and all bits of the Cylinder High, Cylinder  
Low, and Sector Number registers are LBA bits.  
The sector No. under the LBA mode proceeds in the ascending order with the start point of LBA0  
(defined as follows).  
LBA0 = [Cylinder 0, Head 0, Sector 1]  
Even if the host system changes the assignment of the CHS mode by the INITIALIZE DEVICE  
PARAMETER command, the sector LBA address is not changed.  
LBA = [((Cylinder No.) × (Number of head) + (Head No.)) × (Number of sector/track)]  
+ (Sector No.) – 1  
5.2.1  
I/O registers  
Communication between the host system and the device is done through input-output (I/O)  
registers of the device.  
These I/O registers can be selected by the coded signals, CS0–, CS1–, and DA0 to DA2 from the  
host system. Table 5.3 shows the coding address and the function of I/O registers.  
5 - 6  
C141-E116-01EN  
Table 5.3 I/O registers  
I/O registers  
Host I/O  
address  
CS0–  
CS1–  
DA2  
DA1  
DA0  
Read operation  
Write operation  
Command block registers  
1
1
1
1
1
1
1
1
1
0
0
0
0
0
0
0
0
1
0
0
0
0
1
1
1
1
X
0
0
1
1
0
0
1
1
0
1
0
1
0
1
0
1
Data  
Data  
X'1F0'  
X'1F1'  
X'1F2'  
X'1F3'  
X'1F4'  
X'1F5'  
X'1F6'  
X'1F7'  
Error Register  
Sector Count  
Sector Number  
Cylinder Low  
Cylinder High  
Device/Head  
Status  
Features  
Sector Count  
Sector Number  
Cylinder Low  
Cylinder High  
Device/Head  
Command  
(Invalid)  
X
X
(Invalid)  
Control block registers  
0
0
1
1
1
1
1
1
0
1
Alternate Status  
Device Control  
X'3F6'  
X'3F7'  
Notes:  
1. The Data register for read or write operation can be accessed by 16 bit data bus (DATA0  
to DATA15).  
2. The registers for read or write operation other than the Data registers can be accessed by  
8 bit data bus (DATA0 to DATA7).  
3. When reading the Drive Address register, bit 7 is high-impedance state.  
4. The LBA mode is specified, the Device/Head, Cylinder High, Cylinder Low, and Sector  
Number registers indicate LBA bits 27 to 24, 23 to 16, 15 to 8, and 7 to 0.  
C141-E116-01EN  
5 - 7  
5.2.2  
Command block registers  
(1)  
Data register (X'1F0')  
The Data register is a 16-bit register for data block transfer between the device and the host  
system. Data transfer mode is PIO or LBA mode.  
(2)  
Error register (X'1F1')  
The Error register indicates the status of the command executed by the device. The contents of  
this register are valid when the ERR bit of the Status register is 1.  
This register contains a diagnostic code after power is turned on, a reset, or the EXECUTIVE  
DEVICE DIAGNOSTIC command is executed.  
[Status at the completion of command execution other than diagnostic command]  
Bit 7  
Bit 6  
UNC  
Bit 5  
X
Bit 4  
Bit 3  
X
Bit 2  
Bit 1  
Bit 0  
ICRC  
IDNF  
ABRT TK0NF AMNF  
X: Unused  
- Bit 7:  
Interface CRC error (ICRC). This bit indicates that an interface CRC error has  
occurred during an Ultra DMA data transfer. The content of this bit is not applicable  
for Multiword DMA transfers.  
- Bit 6:  
Uncorrectable Data Error (UNC). This bit indicates that an uncorrectable data error  
has been encountered.  
- Bit 5:  
- Bit 4:  
Unused  
ID Not Found (IDNF). This bit indicates an error except for, uncorrectable error and  
SB not found, and Aborted Command.  
- Bit 3:  
- Bit 2:  
Unused  
Aborted Command (ABRT). This bit indicates that the requested command was  
aborted due to a device status error (e.g. Not Ready, Write Fault) or the command  
code was invalid.  
- Bit 1:  
- Bit 0:  
Track 0 Not Found (TK0NF). This bit indicates that track 0 was not found during  
RECALIBRATE command execution.  
Address Mark Not Found. This bit indicates that a SB not found error has been  
encountered.  
5 - 8  
C141-E116-01EN  
[Diagnostic code]  
X'01': No Error Detected.  
X'02': HDC Register Compare Error  
X'03': Data Buffer Compare Error.  
X'05': ROM Sum Check Error.  
X'80': Device 1 (slave device) Failed.  
Error register of the master device is valid under two devices (master and slave)  
configuration. If the slave device fails, the master device posts X’80’ OR (the diagnostic  
code) with its own status (X'01' to X'05').  
However, when the host system selects the slave device, the diagnostic code of the slave  
device is posted.  
(3)  
(4)  
Features register (X'1F1')  
The Features register provides specific feature to a command. For instance, it is used with SET  
FEATURES command to enable or disable caching.  
Sector Count register (X'1F2')  
The Sector Count register indicates the number of sectors of data to be transferred in a read or  
write operation between the host system and the device. When the value in this register is X'00',  
the sector count is 256.  
When this register indicates X'00' at the completion of the command execution, this indicates that  
the command is completed successfully. If the command is not completed successfully, this  
register indicates the number of sectors to be transferred to complete the request from the host  
system. That is, this register indicates the number of remaining sectors that the data has not been  
transferred due to the error.  
The contents of this register has other definition for the following commands; INITIALIZE  
DEVICE PARAMETERS, FORMAT TRACK, SET FEATURES, IDLE, STANDBY and SET  
MULTIPLE MODE.  
(5)  
Sector Number register (X'1F3')  
The contents of this register indicates the starting sector number for the subsequent command.  
The sector number should be between X'01' and [the number of sectors per track defined by  
INITIALIZE DEVICE PARAMETERS command.  
Under the LBA mode, this register indicates LBA bits 7 to 0.  
C141-E116-01EN  
5 - 9  
(6)  
(7)  
(8)  
Cylinder Low register (X'1F4')  
The contents of this register indicates low-order 8 bits of the starting cylinder address for any disk-  
access.  
At the end of a command, the contents of this register are updated to the current cylinder number.  
Under the LBA mode, this register indicates LBA bits 15 to 8.  
Cylinder High register (X'1F5')  
The contents of this register indicates high-order 8 bits of the disk-access start cylinder address.  
At the end of a command, the contents of this register are updated to the current cylinder number.  
The high-order 8 bits of the cylinder address are set to the Cylinder High register.  
Under the LBA mode, this register indicates LBA bits 23 to 16.  
Device/Head register (X'1F6')  
The contents of this register indicate the device and the head number.  
When executing INITIALIZE DEVICE PARAMETERS command, the contents of this register  
defines "the number of heads minus 1".  
Bit 7  
X
Bit 6  
L
Bit 5  
X
Bit 4  
DEV  
Bit 3  
HS3  
Bit 2  
HS2  
Bit 1  
HS1  
Bit 0  
HS0  
- Bit 7:  
- Bit 6:  
- Bit 5:  
- Bit 4:  
- Bit 3:  
- Bit 2:  
- Bit 1:  
- Bit 0:  
Unused  
L. 0 for CHS mode and 1 for LBA mode.  
Unused  
DEV bit. 0 for the master device and 1 for the slave device.  
HS3 CHS mode head address 3 (23). LBA bit 27.  
HS2 CHS mode head address 3 (22). LBA bit 26.  
HS1 CHS mode head address 3 (21). LBA bit 25.  
HS0 CHS mode head address 3 (20). LBA bit 24.  
5 - 10  
C141-E116-01EN  
(9)  
Status register (X'1F7')  
The contents of this register indicate the status of the device. The contents of this register are  
updated at the completion of each command. When the BSY bit is cleared, other bits in this  
register should be validated within 400 ns. When the BSY bit is 1, other bits of this register are  
invalid. When the host system reads this register while an interrupt is pending, it is considered to  
be the Interrupt Acknowledge (the host system acknowledges the interrupt). Any pending interrupt  
is cleared (negating INTRQ signal) whenever this register is read.  
Bit 7  
BSY  
Bit 6  
Bit 5  
DF  
Bit 4  
DSC  
Bit 3  
DRQ  
Bit 2  
0
Bit 1  
0
Bit 0  
ERR  
DRDY  
- Bit 7:  
Busy (BSY) bit. This bit is set whenever the Command register is accessed. Then  
this bit is cleared when the command is completed. However, even if a command is  
being executed, this bit is 0 while data transfer is being requested (DRQ bit =  
1).When BSY bit is 1, the host system should not write the command block registers.  
If the host system reads any command block register when BSY bit is 1, the contents  
of the Status register are posted. This bit is set by the device under following  
conditions:  
(a) Within 400 ns after RESET- is negated or SRST is set in the Device Control  
register, the BSY bit is set. The BSY bit is cleared, when the reset process is  
completed.  
The BSY bit is set for no longer than 15 seconds after the IDD accepts reset.  
(b) Within 400 ns from the host system starts writing to the Command register.  
(c) Within 5 µs following transfer of 512 bytes data during execution of the READ  
SECTOR(S), WRITE SECTOR(S), FORMAT TRACK, or WRITE BUFFER  
command.  
Within 5 µs following transfer of 512 bytes of data and the appropriate number  
of ECC bytes during execution of READ LONG or WRITE LONG command.  
- Bit 6:  
Device Ready (DRDY) bit. This bit indicates that the device is capable to respond to  
a command.  
The IDD checks its status when it receives a command. If an error is detected (not  
ready state), the IDD clears this bit to 0. This is cleared to 0 at power-on and it is  
cleared until the rotational speed of the spindle motor reaches the steady speed.  
- Bit 5:  
- Bit 4:  
The Device Write Fault (DF) bit. This bit indicates that a device fault (write fault)  
condition has been detected.  
If a write fault is detected during command execution, this bit is latched and retained  
until the device accepts the next command or reset.  
Device Seek Complete (DSC) bit. This bit indicates that the device heads are  
positioned over a track.  
In the IDD, this bit is always set to 1 after the spin-up control is completed.  
C141-E116-01EN  
5 - 11  
- Bit 3:  
Data Request (DRQ) bit. This bit indicates that the device is ready to transfer data of  
word unit or byte unit between the host system and the device.  
- Bit 2:  
- Bit 1:  
- Bit 0:  
Always 0.  
Always 0.  
Error (ERR) bit. This bit indicates that an error was detected while the previous  
command was being executed. The Error register indicates the additional  
information of the cause for the error.  
(10)  
Command register (X'1F7')  
The Command register contains a command code being sent to the device. After this register is  
written, the command execution starts immediately.  
Table 5.3 lists the executable commands and their command codes. This table also lists the  
necessary parameters for each command which are written to certain registers before the Command  
register is written.  
5 - 12  
C141-E116-01EN  
5.2.3  
Control block registers  
(1)  
Alternate Status register (X'3F6')  
The Alternate Status register contains the same information as the Status register of the command  
block register.  
The only difference from the Status register is that a read of this register does not imply Interrupt  
Acknowledge and INTRQ signal is not reset.  
Bit 7  
BSY  
Bit 6  
Bit 5  
DF  
Bit 4  
DSC  
Bit 3  
DRQ  
Bit 2  
0
Bit 1  
0
Bit 0  
ERR  
DRDY  
(2)  
Device Control register (X'3F6')  
The Device Control register contains device interrupt and software reset.  
Bit 7  
X
Bit 6  
X
Bit 5  
X
Bit 4  
X
Bit 3  
X
Bit 2  
Bit 1  
nIEN  
Bit 0  
0
SRST  
- Bit 2:  
SRST is the host software reset bit. When this bit is set, the device is held reset state.  
When two devices are daisy chained on the interface, setting this bit resets both  
devices simultaneously.  
The slave device is not required to execute the DASP- handshake.  
- Bit 1:  
nIEN bit enables an interrupt (INTRQ signal) from the device to the host. When this  
bit is 0 and the device is selected, an interruption (INTRQ signal) can be enabled  
through a tri-state buffer. When this bit is 1 or the device is not selected, the INTRQ  
signal is in the high-impedance state.  
5.3  
Host Commands  
The host system issues a command to the device by writing necessary parameters in related  
registers in the command block and writing a command code in the Command register.  
The device can accept the command when the BSY bit is 0 (the device is not in the busy status).  
The host system can halt the uncompleted command execution only at execution of hardware or  
software reset.  
When the BSY bit is 1 or the DRQ bit is 1 (the device is requesting the data transfer) and the host  
system writes to the command register, the correct device operation is not guaranteed.  
C141-E116-01EN  
5 - 13  
5.3.1  
Command code and parameters  
Table 5.4 lists the supported commands, command code and the registers that needed parameters  
are written.  
Table 5.4 Command code and parameters (1 of 2)  
Command code (Bit)  
Parameters used  
Command name  
7
0
1
1
0
1
1
0
0
0
0
1
1
1
1
1
1
0
0
0
1
1
6
0
1
1
1
1
1
0
0
0
1
0
1
1
1
1
0
1
0
0
1
1
5
1
0
0
0
0
0
1
1
0
1
0
1
1
1
0
0
0
1
1
1
1
4
0
0
0
0
0
0
1
1
1
1
1
0
0
0
0
1
1
0
1
0
0
3
0
0
1
0
0
1
1
0
X
X
0
1
1
1
0
0
0
0
0
0
1
2
0
1
0
0
1
0
1
0
X
X
0
1
1
1
1
0
0
0
0
1
0
1
0
0
0
0
0
1
0
0
X
X
0
0
1
1
1
0
0
1
1
0
0
0
R
0
FR SC SN CY DH  
READ SECTOR(S)  
READ MULTIPLE  
READ DMA  
N
N
N
N
N
N
N
N
N
N
N
N
N
Y
N
N
N
N
N
N
N
Y
Y
Y
Y
Y
Y
Y
Y
N
N
Y
N
N
N*  
Y
N
N
Y
Y
N
N
Y
Y
Y
Y
Y
Y
Y
Y
N
Y
N
N
N
N
N
N
Y*  
Y
Y
N
N
Y
Y
Y
Y
Y
Y
Y
Y
N
Y
N
N
N
N
N
N
Y
Y
Y
N
N
Y
Y
Y
Y
Y
Y
Y
Y
D
Y
Y
D
D
D
D
D*  
Y
Y
Y
D
D
R
R
1
READ VERIFY SECTOR(S)  
WRITE MULTIPLE  
WRITE DMA  
R
0
WRITE VERIFY  
WRITE SECTOR(S)  
RECALIBRATE  
R
X
X
1
SEEK  
INITIALIZE DEVICE DIAGNOSTIC  
IDENTIFY DEVICE  
IDENTIFY DEVICE DMA  
SET FEATURES  
0
0
1
SET MULTIPLE MODE  
EXECUTE DEVICE DIAGNOSTIC  
FORMAT TRACK  
READ LONG  
0
0
0
R
R
0
WRITE LONG  
READ BUFFER  
WRITE BUFFER  
0
IDLE  
1
1
0
1
0
1
1
0
0
0
1
0
1
1
1
1
N
N
N
Y
N
Y
N
N
N
N
N
N
D
D
D
IDLE IMMEDIATE  
STANDBY  
1
1
0
1
0
1
1
0
0
0
1
0
0
0
1
1
1
1
0
1
0
1
1
0
0
0
1
0
1
1
0
0
5 - 14  
C141-E116-01EN  
Table 5.4 Command code and parameters (2 of 2)  
Command code (Bit)  
Parameters used  
Command name  
7
6
5
4
3
2
1
0
FR SC SN CY DH  
STANDBY IMMEDIATE  
1
1
0
1
0
1
1
0
0
0
1
0
0
0
0
0
N
N
N
N
N
N
N
N
N
N
N
N
D
D
D
SLEEP  
1
1
0
1
0
1
1
0
1
0
0
1
0
1
1
0
CHECK POWER MODE  
1
1
0
1
0
1
1
0
1
0
0
1
0
0
0
1
SMART  
1
1
1
1
1
1
1
1
1
1
0
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
0
1
1
1
1
1
1
1
1
0
0
0
0
0
0
0
0
1
1
0
1
1
0
1
1
0
0
0
0
0
1
1
1
0
0
0
1
0
0
0
1
0
1
0
1
1
0
1
0
Y
N
N
N
N
N
N
N
N
N
Y
N
N
N
N
N
N
N
Y
N
Y
N
N
N
N
N
N
N
Y
N
Y
N
N
N
N
N
N
N
Y
N
D
D
D
D
D
D
D
D
Y
D
FLUSH CACHE  
SECURITY DISABLE PASSWORD  
SECURITY ERASE PREPARE  
SECURITY ERASE UNIT  
SECURITY FREEZE LOCK  
SECURITY SET PASSWORD  
SECURITY UNLOCK  
SET MAX ADDRESS  
READ NATIVE MAX ADDRESS  
Notes:  
FR : Features Register  
CY: Cylinder Registers  
SC : Sector Count Register  
SN : Sector Number Register  
DH : Drive/Head Register  
R: R = 0 or 1  
Y: Necessary to set parameters  
Y*: Necessary to set parameters under the LBA mode.  
N: Necessary to set parameters (The parameter is ignored if it is set.)  
N*: May set parameters  
D: The device parameter is valid, and the head parameter is ignored.  
D*: The command is addressed to the master device, but both the master device and the slave  
device execute it.  
X: Do not care  
C141-E116-01EN  
5 - 15  
5.3.2  
Command descriptions  
The contents of the I/O registers to be necessary for issuing a command and the example indication  
of the I/O registers at command completion are shown as following in this subsection.  
Example: READ SECTOR(S)  
At command issuance (I/O registers setting contents)  
Bit  
7
0
6
0
5
1
4
0
3
0
2
0
1
0
0
0
1F7H(CM)  
1F6H(DH)  
L
DV  
Head No. / LBA [MSB]  
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(FR)  
Start cylinder address [MSB] / LBA  
Start cylinder address [LSB] / LBA  
Start sector No.  
Transfer sector count  
xx  
/ LBA [LSB]  
At command completion (I/O registers contents to be read)  
Bit  
7
6
5
4
3
2
1
0
1F7H(ST)  
1F6H(DH)  
Error information  
DV End Head No. / LBA [MSB]  
L
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(ER)  
End cylinder address [MSB] / LBA  
End cylinder address [LSB] / LBA  
End sector No.  
/ LBA [LSB]  
X‘00’  
Error information  
CM: Command register  
FR: Features register  
DH: Device/Head register  
CH: Cylinder High register  
CL: Cylinder Low register  
SN: Sector Number register  
SC: Sector Count register  
ST: Status register  
ER: Error register  
L: LBA (logical block address) setting bit  
DV: Device address. bit  
x, xx: Do not care (no necessary to set)  
5 - 16  
C141-E116-01EN  
Note:  
1. When the L bit is specified to 1, the lower 4 bits of the DH register and all bits of the CH,  
CL and SN registers indicate the LBA bits (bits of the DH register are the MSB (most  
significant bit) and bits of the SN register are the LSB (least significant bit).  
2. At error occurrence, the SC register indicates the remaining sector count of data transfer.  
3. In the table indicating I/O registers contents in this subsection, bit indication is omitted.  
(1)  
READ SECTOR(S) (X'20' or X'21')  
This command reads data of sectors specified in the Sector Count register from the address specified in  
the Device/Head, Cylinder High, Cylinder Low and Sector Number registers. Number of sectors can be  
specified to 256 sectors in maximum. To specify 256 sectors reading, '00' is specified. For the DRQ,  
INTRQ, and BSY protocols related to data transfer, see Subsection 5.4.1.  
If the head is not on the track specified by the host, the device performs an implied seek. After the  
head reaches to the specified track, the device reads the target sector.  
The DRQ bit of the Status register is always set prior to the data transfer regardless of an error  
condition.  
Upon the completion of the command execution, command block registers contain the cylinder,  
head, and sector addresses (in the CHS mode) or logical block address (in the LBA mode) of the  
last sector read.  
If an error occurs in a sector, the read operation is terminated at the sector where the error occurred.  
Command block registers contain the cylinder, the head, and the sector addresses of the sector (in  
the CHS mode) or the logical block address (in the LBA mode) where the error occurred, and  
remaining number of sectors of which data was not transferred.  
At command issuance (I/O registers setting contents)  
1F7H(CM)  
1F6H(DH)  
0
0
1
0
0
0
0
R
L
DV  
Start head No. /LBA [MSB]  
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(FR)  
Start cylinder No. [MSB] / LBA  
Start cylinder No. [LSB] / LBA  
Start sector No.  
/ LBA [LSB]  
Transfer sector count  
xx  
R = 0 or 1  
C141-E116-01EN  
5 - 17  
At command completion (I/O registers contents to be read)  
1F7H(ST) Status information  
DV End head No. /LBA [MSB]  
1F6H(DH)  
L
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(ER)  
End cylinder No. [MSB] / LBA  
End cylinder No. [LSB] / LBA  
End sector No.  
/ LBA [LSB]  
00 (*1)  
Error information  
*1 If the command is terminated due to an error, the remaining number of  
sectors of which data was not transferred is set in this register.  
(2)  
READ MULTIPLE (X'C4')  
This command operates similarly to the READ SECTOR(S) command. The device does not  
generate an interrupt (assertion of the INTRQ signal) on each every sector. An interrupt is  
generated after the transfer of a block of sectors for which the number is specified by the SET  
MULTIPLE MODE command.  
The implementation of the READ MULTIPLE command is identical to that of the READ  
SECTOR(S) command except that the number of sectors is specified by the SET MULTIPLE  
MODE command are transferred without intervening interrupts. In the READ MULTIPLE  
command operation, the DRQ bit of the Status register is set only at the start of the data block, and  
is not set on each sector.  
The number of sectors (block count) to be transferred without interruption is specified by the SET  
MULTIPLE MODE command. The SET MULTIPLE MODE command should be executed prior  
to the READ MULTIPLE command.  
When the READ MULTIPLE command is issued, the Sector Count register contains the number of  
sectors requested (not a number of the block count or a number of sectors in a block).  
Upon receipt of this command, the device executes this command even if the value of the Sector Count  
register is less than the defined block count (the value of the Sector Count should not be 0).  
If the number of requested sectors is not divided evenly (having the same number of sectors [block  
count]), as many full blocks as possible are transferred, then a final partial block is transferred.  
The number of sectors in the partial block to be transferred is n where n = remainder of ("number  
of sectors"/"block count").  
If the READ MULTIPLE command is issued before the SET MULTIPLE MODE command is  
executed or when the READ MULTIPLE command is disabled, the device rejects the READ  
MULTIPLE command with an ABORTED COMMAND error.  
If an error occurs, reading sector is stopped at the sector where the error occurred. Command  
block registers contain the cylinder, the head, the sector addresses (in the CHS mode) or the logical  
block address (in the LBA mode) of the sector where the error occurred, and remaining number of  
sectors that had not transferred after the sector where the error occurred.  
An interrupt is generated when the DRQ bit is set at the beginning of each block or a partial block.  
5 - 18  
C141-E116-01EN  
Figure 5.1 shows an example of the execution of the READ MULTIPLE command.  
Block count specified by SET MULTIPLE MODE command = 4 (number of sectors in a  
block)  
READ MULTIPLE command specifies;  
Number of requested sectors = 9 (Sector Count register = 9)  
Number of sectors in incomplete block = remainder of 9/4 =1  
Command Issue  
Parameter  
Write  
Status read  
Status read  
Status read  
~
BSY  
DRDY  
INTRQ  
DRQ  
5
6
7
8
9
1
2
3
4
Sector  
transferred  
Partial  
block  
Block  
Block  
Figure 5.1 Execution example of READ MULTIPLE command  
At command issuance (I/O registers setting contents)  
1F7H(CM)  
1
1
0
0
0
1
0
0
1F6H(DH)  
L
DV  
Start head No. /LBA [MSB]  
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(FR)  
Start cylinder No. [MSB] / LBA  
Start cylinder No. [LSB] / LBA  
Start sector No.  
Transfer sector count  
xx  
/ LBA [LSB]  
At command completion (I/O registers contents to be read)  
1F7H(ST) Status information  
DV End head No. /LBA [MSB]  
1F6H(DH)  
L
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(ER)  
End cylinder No. [MSB] / LBA  
End cylinder No. [LSB] / LBA  
End sector No.  
/ LBA [LSB]  
00H (*1)  
Error information  
*1 If the command is terminated due to an error, the remaining number of sectors for which data  
was not transferred is set in this register.  
C141-E116-01EN  
5 - 19  
(3)  
READ DMA (X'C8' or X'C9')  
This command operates similarly to the READ SECTOR(S) command except for following events.  
The data transfer starts at the timing of DMARQ signal assertion.  
The device controls the assertion or negation timing of the DMARQ signal.  
The device posts a status as the result of command execution only once at completion of the  
data transfer.  
When an error, such as an unrecoverable medium error, that the command execution cannot be continued  
is detected, the data transfer is stopped without transferring data of sectors after the erred sector. The  
device generates an interrupt using the INTRQ signal and posts a status to the host system. The format of  
the error information is the same as the READ SECTOR(S) command.  
In LBA mode  
The logical block address is specified using the start head No., start cylinder No., and first sector  
No. fields. At command completion, the logical block address of the last sector and remaining  
number of sectors of which data was not transferred, like in the CHS mode, are set.  
The host system can select the DMA transfer mode by using the SET FEATURES command.  
1) Multiword DMA transfer mode 2:  
Sets the FR register = X'03' and SC register = X'22' by the SET FEATURES command  
2) Ultra DMA transfer mode 2:  
Sets the FR register = X'03' and SC register = X'42' by the SET FEATURES command  
At command issuance (I/O registers setting contents)  
1F7H(CM)  
1F6H(DH)  
1
1
0
0
1
0
0
R
L
DV  
Start head No. /LBA [MSB]  
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(FR)  
Start cylinder No. [MSB] / LBA  
Start cylinder No. [LSB] / LBA  
Start sector No.  
/ LBA [LSB]  
Transfer sector count  
xx  
R = 0 or 1  
5 - 20  
C141-E116-01EN  
At command completion (I/O registers contents to be read)  
1F7H(ST) Status information  
DV End head No. /LBA [MSB]  
1F6H(DH)  
L
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(ER)  
End cylinder No. [MSB] / LBA  
End cylinder No. [LSB] / LBA  
End sector No.  
/ LBA [LSB]  
00 (*1)  
Error information  
*1 If the command is terminated due to an error, the remaining number of  
sectors of which data was not transferred is set in this register.  
(4)  
READ VERIFY SECTOR(S) (X'40' or X'41')  
This command operates similarly to the READ SECTOR(S) command except that the data is not  
transferred to the host system.  
After all requested sectors are verified, the device clears the BSY bit of the Status register and  
generates an interrupt. Upon the completion of the command execution, the command block  
registers contain the cylinder, head, and sector number of the last sector verified.  
If an error occurs, the verify operation is terminated at the sector where the error occurred. The  
command block registers contain the cylinder, the head, and the sector addresses (in the CHS  
mode) or the logical block address (in the LBA mode) of the sector where the error occurred. The  
Sector Count register indicates the number of sectors that have not been verified.  
At command issuance (I/O registers setting contents)  
1F7H(CM)  
1F6H(DH)  
0
1
0
0
0
0
0
R
L
DV  
Start head No. /LBA [MSB]  
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(FR)  
Start cylinder No. [MSB] / LBA  
Start cylinder No. [LSB] / LBA  
Start sector No.  
/ LBA [LSB]  
Transfer sector count  
xx  
R = 0 or 1  
C141-E116-01EN  
5 - 21  
At command completion (I/O registers contents to be read)  
1F7H(ST) Status information  
DV End head No. /LBA [MSB]  
1F6H(DH)  
L
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(ER)  
End cylinder No. [MSB] / LBA  
End cylinder No. [LSB] / LBA  
End sector No.  
/ LBA [LSB]  
00 (*1)  
Error information  
*1 If the command is terminated due to an error, the remaining number of  
sectors of which data was not transferred is set in this register.  
(5)  
WRITE SECTOR(S) (X'30' or X'31')  
This command writes data of sectors from the address specified in the Device/Head, Cylinder  
High, Cylinder Low, and Sector Number registers to the address specified in the Sector Count  
register. Number of sectors can be specified to 256 sectors in maximum. Data transfer begins at  
the sector specified in the Sector Number register. For the DRQ, INTRQ, and BSY protocols  
related to data transfer, see Subsection 5.4.2.  
If the head is not on the track specified by the host, the device performs an implied seek. After the  
head reaches to the specified track, the device writes the target sector.  
The data stored in the buffer, and CRC code and ECC bytes are written to the data field of the  
corresponding sector(s). Upon the completion of the command execution, the command block  
registers contain the cylinder, head, and sector addresses of the last sector written.  
If an error occurs during multiple sector write operation, the write operation is terminated at the  
sector where the error occurred. Command block registers contain the cylinder, the head, the  
sector addresses (in the CHS mode) or the logical block address (in the LBA mode) of the sector  
where the error occurred. Then the host can read the command block registers to determine what  
error has occurred and on which sector the error has occurred.  
At command issuance (I/O registers setting contents)  
1F7H(CM)  
1F6H(DH)  
0
0
1
1
0
0
0
R
L
DV  
Start head No. /LBA [MSB]  
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(FR)  
Start cylinder No. [MSB] / LBA  
Start cylinder No. [LSB] / LBA  
Start sector No.  
/ LBA [LSB]  
Transfer sector count  
xx  
R = 0 or 1  
5 - 22  
C141-E116-01EN  
At command completion (I/O registers contents to be read)  
1F7H(ST) Status information  
DV End head No. /LBA [MSB]  
1F6H(DH)  
L
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(ER)  
End cylinder No. [MSB] / LBA  
End cylinder No. [LSB] / LBA  
End sector No.  
/ LBA [LSB]  
00 (*1)  
Error information  
*1 If the command is terminated due to an error, the remaining number of  
sectors of which data was not transferred is set in this register.  
(6)  
WRITE MULTIPLE (X'C5')  
This command is similar to the WRITE SECTOR(S) command. The device does not generate  
interrupts (assertion of the INTRQ signal) on each sector but on the transfer of a block which  
contains the number of sectors for which the number is defined by the SET MULTIPLE MODE  
command.  
The implementation of the WRITE MULTIPLE command is identical to that of the WRITE  
SECTOR(S) command except that the number of sectors is specified by the SET MULTIPLE  
MODE command are transferred without intervening interrupts. In the WRITE MULTIPLE  
command operation, the DRQ bit of the Status register is required to set only at the start of the data  
block, not on each sector.  
The number of sectors (block count) to be transferred without interruption is specified by the SET  
MULTIPLE MODE command. The SET MULTIPLE MODE command should be executed prior  
to the WRITE MULTIPLE command.  
When the WRITE MULTIPLE command is issued, the Sector Count register contains the number of  
sectors requested (not a number of the block count or a number of sectors in a block).  
Upon receipt of this command, the device executes this command even if the value of the Sector Count  
register is less than the defined block count the value of the Sector Count should not be 0).  
If the number of requested sectors is not divided evenly (having the same number of sectors [block  
count]), as many full blocks as possible are transferred, then a final partial block is transferred. The  
number of sectors in the partial block to be transferred is n where n = remainder of ("number of  
sectors"/"block count").  
If the WRITE MULTIPLE command is issued before the SET MULTIPLE MODE command is  
executed or when WRITE MULTIPLE command is disabled, the device rejects the WRITE  
MULTIPLE command with an ABORTED COMMAND error.  
Disk errors encountered during execution of the WRITE MULTIPLE command are posted after  
attempting to write the block or the partial block that was transferred. Write operation ends at the sector  
where the error was encountered even if the sector is in the middle of a block. If an error occurs, the  
subsequent block shall not be transferred. Interrupts are generated when the DRQ bit of the Status  
register is set at the beginning of each block or partial block.  
C141-E116-01EN  
5 - 23  
The contents of the command block registers related to addresses after the transfer of a data block  
containing an erred sector are undefined. To obtain a valid error information, the host should retry  
data transfer as an individual requests.  
At command issuance (I/O registers setting contents)  
1F7H(CM)  
1F6H(DH)  
1
1
0
0
0
1
0
1
L
DV  
Start head No. /LBA [MSB]  
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(FR)  
Start cylinder No. [MSB] / LBA  
Start cylinder No. [LSB] / LBA  
Start sector No.  
Transfer sector count  
xx  
/ LBA [LSB]  
At command completion (I/O registers contents to be read)  
1F7H(ST) Status information  
DV End head No. /LBA [MSB]  
1F6H(DH)  
L
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(ER)  
End cylinder No. [MSB] / LBA  
End cylinder No. [LSB] / LBA  
End sector No.  
/ LBA [LSB]  
00H  
Error information  
Note:  
When the command terminates due to error, only the DV bit and the error information field are  
valid.  
(7)  
WRITE DMA (X'CA' or X'CB')  
This command operates similarly to the WRITE SECTOR(S) command except for following  
events.  
The data transfer starts at the timing of DMARQ signal assertion.  
The device controls the assertion or negation timing of the DMARQ signal.  
The device posts a status as the result of command execution only once at completion of the  
data transfer.  
When an error, such as an unrecoverable medium error, that the command execution cannot be  
continued is detected, the data transfer is stopped without transferring data of sectors after the  
erred sector. The device generates an interrupt using the INTRQ signal and posts a status to the  
host system. The format of the error information is the same as the WRITE SECTOR(S)  
command.  
A host system can be select the following transfer mode using the SET FEATURES command.  
5 - 24  
C141-E116-01EN  
1) Multiword DMA transfer mode 2:  
Sets the FR register = X'03' and SC register = X'22' by the SET FEATURES command  
2) Ultra DMA transfer mode 2:  
Sets the FR register = X'03' and SC register = X'42' by the SET FEATURES command  
At command issuance (I/O registers setting contents)  
1F7H(CM)  
1F6H(DH)  
1
1
0
0
1
0
1
R
L
DV  
Start head No. /LBA [MSB]  
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(FR)  
Start cylinder No. [MSB] / LBA  
Start cylinder No. [LSB] / LBA  
Start sector No.  
Transfer sector count  
xx  
/ LBA [LSB]  
R = 0 or 1  
At command completion (I/O registers contents to be read)  
1F7H(ST) Status information  
DV End head No. /LBA [MSB]  
1F6H(DH)  
L
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(ER)  
End cylinder No. [MSB] / LBA  
End cylinder No. [LSB] / LBA  
End sector No.  
/ LBA [LSB]  
00 (*1)  
Error information  
*1 If the command is terminated due to an error, the remaining number of  
sectors of which data was not transferred is set in this register.  
(8)  
WRITE VERIFY (X'3C')  
This command operates similarly to the WRITE SECTOR(S) command except that the device verifies  
each sector immediately after being written. The verify operation is a read and check for data errors  
without data transfer. Any error that is detected during the verify operation is posted.  
At command issuance (I/O registers setting contents)  
1F7H(CM)  
1F6H(DH)  
0
0
1
1
1
1
0
0
L
DV  
Start head No. /LBA [MSB]  
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(FR)  
Start cylinder No. [MSB] / LBA  
Start cylinder No. [LSB] / LBA  
Start sector No.  
/ LBA [LSB]  
Transfer sector count  
xx  
C141-E116-01EN  
5 - 25  
At command completion (I/O registers contents to be read)  
1F7H(ST) Status information  
DV End head No. /LBA [MSB]  
1F6H(DH)  
L
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(ER)  
End cylinder No. [MSB] / LBA  
End cylinder No. [LSB] / LBA  
End sector No.  
/ LBA [LSB]  
00 (*1)  
Error information  
*1 If the command is terminated due to an error, the remaining number of  
sectors of which data was not transferred is set in this register.  
(9)  
RECALIBRATE (X'1x', x: X'0' to X'F')  
This command performs the calibration. Upon receipt of this command, the device sets BSY bit of  
the Status register and performs a calibration. When the device completes the calibration, the  
device updates the Status register, clears the BSY bit, and generates an interrupt.  
This command can be issued in the LBA mode.  
At command issuance (I/O registers setting contents)  
1F7H(CM)  
1F6H(DH)  
0
0
0
1
x
x
x
x
DV  
xx  
×
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(FR)  
xx  
xx  
xx  
xx  
xx  
At command completion (I/O registers contents to be read)  
1F7H(ST)  
Status information  
DV  
1F6H(DH)  
xx  
×
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(ER)  
xx  
xx  
xx  
xx  
Error information  
5 - 26  
C141-E116-01EN  
(10)  
SEEK (X'7x', x : X'0' to X'F')  
This command performs a seek operation to the track and selects the head specified in the  
command block registers. After completing the seek operation, the device clears the BSY bit in  
the Status register and generates an interrupt.  
The IDD always sets the DSC bit (Drive Seek Complete status) of the Status register to 1.  
In the LBA mode, this command performs the seek operation to the cylinder and head position in  
which the sector is specified with the logical block address.  
At command issuance (I/O registers setting contents)  
1F7H(CM)  
1F6H(DH)  
0
1
1
1
x
x
x
x
L
DV  
Head No. /LBA [MSB]  
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(FR)  
Cylinder No. [MSB] / LBA  
Cylinder No. [LSB] / LBA  
Sector No.  
/ LBA [LSB]  
xx  
xx  
At command completion (I/O registers contents to be read)  
1F7H(ST) Status information  
DV Head No. /LBA [MSB]  
1F6H(DH)  
L
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(ER)  
Cylinder No. [MSB] / LBA  
Cylinder No. [LSB] / LBA  
Sector No.  
/ LBA [LSB]  
xx  
Error information  
C141-E116-01EN  
5 - 27  
(11)  
INITIALIZE DEVICE PARAMETERS (X'91')  
The host system can set the number of sectors per track and the maximum head number (maximum  
head number is "number of heads minus 1") per cylinder with this command. Upon receipt of this  
command, the device sets the BSY bit of Status register and saves the parameters. Then the device  
clears the BSY bit and generates an interrupt.  
When the SC register is specified to X'00', an ABORTED COMMAND error is posted. Other than  
X'00' is specified, this command terminates normally.  
The parameters set by this command are retained even after reset or power save operation  
regardless of the setting of disabling the reverting to default setting.  
In LBA mode  
The device ignores the L bit specification and operates with the CHS mode specification. An accessible  
area of this command within head moving in the LBA mode is always within a default area. It is  
recommended that the host system refers the addressable user sectors (total number of sectors) in word  
60 to 61 of the parameter information by the IDENTIFY DEVICE command.  
At command issuance (I/O registers setting contents)  
1F7H(CM)  
1F6H(DH)  
1
0
0
1
0
0
0
1
DV  
Max. head No.  
×
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(FR)  
xx  
xx  
xx  
Number of sectors/track  
xx  
At command completion (I/O registers contents to be read)  
1F7H(ST) Status information  
DV Max. head No.  
1F6H(DH)  
×
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(ER)  
xx  
xx  
xx  
xx  
Error Information  
(12)  
IDENTIFY DEVICE (X'EC')  
The host system issues the IDENTIFY DEVICE command to read parameter information (512  
bytes) from the device. Upon receipt of this command, the drive sets the BSY bit of Status register  
and sets required parameter information in the sector buffer. The device then sets the DRQ bit of  
the Status register, and generates an interrupt. After that, the host system reads the information out  
of the sector buffer. Table 5.5 shows the arrangements and values of the parameter words and the  
meaning in the buffer.  
5 - 28  
C141-E116-01EN  
At command issuance (I/O registers setting contents)  
1F7H(CM)  
1F6H(DH)  
1
1
1
0
1
1
0
0
DV  
xx  
×
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(FR)  
xx  
xx  
xx  
xx  
xx  
At command completion (I/O registers contents to be read)  
1F7H(ST)  
Status information  
DV  
1F6H(DH)  
xx  
×
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(ER)  
xx  
xx  
xx  
xx  
Error information  
Table 5.5 Information to be read by IDENTIFY DEVICE command (1 of 6)  
Word  
Value  
Description  
0
1
X‘045A’  
*2  
General Configuration *1  
Number of cylinders  
Reserved  
2
X‘0000’  
*3  
3
Number of Heads  
Retired  
4
X‘0000’  
X‘0000’  
X‘003F’  
5
Retired  
6
Number of sectors per track  
7-9  
10-19  
20  
21  
22  
23-26  
27-46  
47  
48  
49  
50  
51  
52  
53  
X‘000000000000’ Retired  
Serial number (ASCII code) *4  
X‘0003’  
X‘1000’  
X‘0004’  
Retired  
Buffer size in 512 byte increments  
Number of ECC bytes transferred at READ LONG or WRITE LONG command  
Firmware revision (ASCII code) *5  
Model number (ASCII code) *6  
X‘8010’  
X‘0000’  
X‘2B00’  
X‘4000’  
X‘0200’  
X‘0200’  
X‘0007’  
Maximum number of sectors per interrupt on READ/WRITE MULTIPLE command  
Reserved  
Capabilities *7  
Reserved  
PIO data transfer mode *8  
Retired  
Enable/disable setting of words 54-58, 64-70 and 88 *9  
C141-E116-01EN  
5 - 29  
Table 5.5 Information to be read by IDENTIFY DEVICE command (2 of 6)  
Word  
Value  
Description  
54  
55  
(Variable)  
(Variable)  
(Variable)  
(Variable)  
*10  
Number of current Cylinders  
Number of current Head  
56  
Number of current sectors per track  
Total number of current sectors  
57-58  
59  
Transfer sector count currently set by READ/WRITE MULTIPLE command  
Total number of user addressable sectors (LBA mode only)  
Retired  
60-61  
62  
*11  
X‘0000’  
X‘xx07’  
X‘0003’  
X‘0078’  
X‘0078’  
X‘00F0’  
X‘0078’  
X‘00’  
63  
Multiword DMA transfer mode *12  
64  
Advance PIO transfer mode support status *13  
Minimum multiword DMA transfer cycle time per word : 120 [ns]  
Manufacturer's recommended DMA transfer cycle time : 120 [ns]  
Minimum PIO transfer cycle time without flow control : 120 [ns]  
Minimum PIO transfer cycle time with IORDY flow control : 120 [ns]  
Reserved  
65  
66  
67  
68  
69-79  
80  
X‘003E’  
X‘0015’  
X‘346B’  
X‘4108’  
X‘4000’  
X‘xxxx’  
X‘xxxx’  
X‘4000’  
X‘xx3F’  
X‘00xx’  
X‘0000’  
X‘00xx’  
X‘0000’  
X‘xxxx’  
X‘00xx’  
X‘00’  
Major version number *14  
81  
Minor version number ATA/ATAPI5 X3T13 1321D Support of rev 1  
Support of command sets *15  
82  
83  
Support of command sets *16  
84  
Support of command set/feature extension (fixed)  
Enable/disable Command set/feature enabled. *17  
Enable/disable Command set/feature enabled. *18  
Default of command set/feature (fixed)  
Ultra DMA modes *19  
85  
86  
87  
88  
89  
Time required for security erase unit completion *20  
Time required for Enhanced security erase completion  
Current advanced power management value  
Reserved  
90  
91  
92  
93  
CBLID detection results *21  
94  
Automatic Acoustic Management (Slow Seek mode) *22  
Reserved  
95-127  
128  
129-255  
X‘xx’  
Security Status  
X‘00’  
Reserved  
5 - 30  
C141-E116-01EN  
Table 5.5 Information to be read by IDENTIFY DEVICE command (3 of 6)  
*1 Word 0: General configuration  
Bit 15: 0 = ATA device  
0
Bit 14-8: Vendor specific  
Bit 7: 1 = Removable media device  
Bit 6: 1 = not removable controller and/or device  
Bit 5-1: Vendor specific  
0
1
Bit 0: Reserved  
0
*2 Number of Cylinders, *3 Number of Heads,  
*11 Total number of user addressable sectors (LBA mode only.)  
MPG3204AH-E  
MPG3307AH-E  
X‘3FFF’  
MPG3409AH-E  
X‘4C5ABC0’  
*2  
*3  
X‘10’  
*11  
X‘262D5E01’  
X‘3943CE0’  
*4 Word 10-19: Serial number; ASCII code (20 characters, right-justified)  
*5 Word 23-26: Firmware revision; ASCII code (8 characters, Left-justified)  
*6 Word 27-46: Model number;  
ASCII code (40 characters, Left-justified), remainder filled with blank code (X'20')  
One of the following model numbers;  
Fujitsu MPG3204AH-E  
Fujitsu MPG3307AH-E  
Fujitsu MPG3409AH-E  
*7 Word 49: Capabilities  
Bit 15-14:  
Bit 13:  
Bit 12:  
Bit 11:  
Bit 10:  
Bit 9:  
Reserved  
Standby timer value 0 = Standby timer values shall be managed by the device  
Reserved  
IORDY support  
1=Supported  
IORDY inhibition 0=Disable inhibition  
LBA support  
DMA support  
Vendor specific  
1=Supported  
1=Supported  
Bit 8:  
Bit 7-0:  
*8 Word 51: PIO data transfer mode  
Bit 15-8:  
Bit 7-0:  
PIO data transfer mode  
Vendor specific  
X'02'=PIO mode 2  
*9 Word 53: Enable/disable setting of word 54-58, 64-70 and 88  
Bit 15-3:  
Bit 2:  
Reserved  
Enable/disable setting of word 88  
1=Enable  
Bit 1:  
Bit 0:  
Enable/disable setting of word 64-70 1=Enable  
Enable/disable setting of word 54-58 1=Enable  
*10 Word 59: Transfer sector count currently set by READ/WRITE MULTIPLE command  
Bit 15-9:  
Bit 8:  
Reserved  
Multiple sector transfer  
1=Enable  
Bit 7-0:  
Transfer sector count currently set by READ/WRITE MULTIPLE without  
interrupt supports 2, 4, 8 and 16 sectors.  
C141-E116-01EN  
5 - 31  
Table 5.5  
Information to be read by IDENTIFY DEVICE command (4 of 6)  
*12 Word 63: Multiword DMA transfer mode  
Bit 15-11:  
Bit 10:  
Reserved  
1 = Multiword DMA mode 2 is selected  
0 = Multiword DMA mode 2 is not selected  
1 = Multiword DMA mode 1 is selected  
0 = Multiword DMA mode 1 is not selected  
1 = Multiword DMA mode 0 is selected  
0 = Multiword DMA mode 0 is not selected  
Reserved  
1 = Multiword DMA mode 2 and below are supported  
1 = Multiword DMA mode 1 and below are supported  
1 = Multiword DMA mode 0 is supported  
Bit 9:  
Bit 8:  
Bit 7-3:  
Bit 2:  
Bit 1:  
Bit 0:  
*13 Word 64: Advance PIO transfer mode support status  
Bit 15-8:  
Bit 7-0:  
Reserved  
Advance PIO transfer mode  
Bit 1=1 Mode 4  
Bit 0=1 Mode 3  
*14 Word 80: Major version number  
Bit 15-5:  
Bit 5:  
Bit 4:  
Bit 3:  
Bit 2:  
Bit 1:  
Bit 0:  
Reserved  
ATA-5 Supported=1  
ATA-4 Supported=1  
ATA-3 Supported=1  
ATA-2 Supported=1  
ATA-1 Supported=1  
Undefined  
*15 Word 82: Support of command sets  
Bit 15:  
Bit 14:  
Bit 13:  
Bit 12:  
Bit 11:  
Bit 10:  
Bit 9:  
Bit 8:  
Bit 7:  
Bit 6:  
Bit 5:  
Bit 4:  
Bit 3:  
Bit 2:  
Bit 1:  
Bit 0:  
Reserved  
NOP command supported = 0  
Read Buffer command supported = 1  
Write Buffer command supported = 1  
Write Verify command supported (Old Spec.) = 0  
Host Protected Area feature command supported = 1  
Device Reset command supported = 0  
SERVICE Interrupt supported = 0  
Release Interrupt supported = 0  
Lock Ahead supported = 1  
Write-cache supported = 1  
Packet command feature set supported = 0  
Power Management feature set supported=1  
Removable feature set supported=0  
Security feature set supported=1  
SMART feature set supported=1  
*16 Word 83: Support of command sets  
Bit 15:  
Bit 14:  
Bit 13-10:  
Bit 9:  
Bit 8:  
Bit 7:  
0
1
Reserved  
Automatic Acoustic Management supported=1  
SET MAX security extension supported=1  
Reserved  
Bit 6:  
Bit 5:  
SET FEATURES subcommand required to spinup after power-up supported=0  
Power-Up In Standby feature set supported=0  
5 - 32  
C141-E116-01EN  
Table 5.5 Information to be read by IDENTIFY DEVICE command (5 of 6)  
Bit 4:  
Bit 3:  
Bit 2:  
Bit 1:  
Bit 0:  
Removable Media Status Notification feature set supported = 0  
Advanced Power Management feature set supported = 1  
CFA feature set supported = 0  
READ/WRITE DMA QUEUED supported = 0  
DOWNLOAD MICROCODE command supported = 0  
*17 Word 85: Enable/disable Command set/feature enabled  
Bit 15:  
Bit 14:  
Bit 13:  
Bit 12:  
Bit 11:  
Bit 10:  
Bit 9:  
Bit 8:  
Bit 7:  
Bit 6:  
Bit 5:  
Bit 4:  
Bit 3:  
Bit 2:  
Bit 1:  
Bit 0:  
Reserved  
NOP command enabled = 0  
READ BUFFER command enabled  
WRITE BUFFER command enabled  
Reserved  
Host Protected Area feature set enabled  
DEVICE RESET command enabled = 0  
SERVICE interrupt enabled = 0  
Release interrupt enabled = 0  
Look-ahead enabled  
Write cache enabled  
PACKET Command feature set enabled = 0  
Power Management feature set enabled  
Removable Media feature set enabled = 0  
Security Mode feature set enabled  
SMART feature set enabled  
*18 Word 86: Command set/feature enabled  
Bit 15-10:  
Bit 9:  
Bit 8:  
Bit 7:  
Bit 6:  
Bit 5:  
Bit 4:  
Bit 3:  
Bit 2:  
Bit 1:  
Bit 0:  
Reserved  
1 = Automatic Acoustic Management feature set enabled  
1 = SET MAX security extension enabled by SET MAX SET PASSWORD  
Reserved  
0 = SET FEATURES subcommand required to spin-up after power-up  
0 = Power-Up In Standby feature set enabled  
0 = Removable Media Status Notification feature set enabled  
1 = Advanced Power Management feature set enabled  
0 = CFA feature set enabled  
0 = READ/WRITE DMA QUEUED command supported  
0 = DOWNLOAD MICROCODE command supported  
*19 Word 88: Ultra DMA modes  
Bit 15-14:  
Bit 13:  
Reserved  
1 = Ultra DMA mode 5 is selected  
0 = Ultra DMA mode 5 is not selected  
1 = Ultra DMA mode 4 is selected  
0 = Ultra DMA mode 4 is not selected  
1 = Ultra DMA mode 3 is selected  
0 = Ultra DMA mode 3 is not selected  
1 = Ultra DMA mode 2 is selected  
0 = Ultra DMA mode 2 is not selected  
1 = Ultra DMA mode 1 is selected  
0 = Ultra DMA mode 1 is not selected  
1 = Ultra DMA mode 0 is selected  
0 = Ultra DMA mode 0 is not selected  
Reserved  
1 = Ultra DMA mode 5 and below are supported  
1 = Ultra DMA mode 4 and below are supported  
1 = Ultra DMA mode 3 and below are supported  
1 = Ultra DMA mode 2 and below are supported  
1 = Ultra DMA mode 1 and below are supported  
1 = Ultra DMA mode 0 is supported  
Bit 12:  
Bit 11:  
Bit 10:  
Bit 9:  
Bit 8:  
Bit 7-6:  
Bit 5:  
Bit 4:  
Bit 3:  
Bit 2:  
Bit 1:  
Bit 0:  
C141-E116-01EN  
5 - 33  
Table 5.5 Information to be read by IDENTIFY DEVICE command (6 of 6)  
*20 Word 89: Time required for SECURITY ERASE UNIT command to complete.  
MPG3204AH-E = 0004H: 8 minutes  
MPG3307AH-E = 0006H: 12minutes  
MPG3409AH-E = 0008H: 16 minutes  
*21 Word 93: Hardware reset result. The contents of bits 12-0 of this word shall change only during the  
execution of a hardware reset.  
Bit 15:  
Bit 14:  
Bit 13:  
0
1
1 = device detected CBLID- above VIH (80-conductor cable)  
0 = device detected CBLID- below VIL (40-conductor cable)  
Device 1 hardware reset result. Device 0 shall clear these bits to zero.  
Device 1 shall set these bits as follows:  
Bit 12: 0 = Reserved  
Bit 12-8:  
Bit 11: 0 = Device 1 did not assert PDIAG-  
1 = Device 1 asserted PDIAG-  
Bit 10-9: These bits indicate how Device 1 determined the device number.  
00 = Reserved  
01 = a jumper was used  
10 = the CSEL signal was used  
11 = some other method was used or the method is unknown  
Bit 8:  
1
Bit 7-0:  
Device 0 hardware reset result. Device 1 shall clear these bits to zero.  
Device 0 shall set these bits as follows:  
Bit 7:  
Bit 6:  
0
0 = Device 0 does not respond when Device 1 is selected  
1 = Device 0 responds when Device 1 is selected  
0 = Device 0 did not detect the assertion of DASP-  
1 = Device 0 detected the assertion of DASP-  
0 = Device 0 did not detect the assertion of PDIAG-  
1 = Device 0 detected the assertion of PDIAG-  
0 = Device 0 failed diagnostics  
Bit 5:  
Bit 4:  
Bit 3:  
1 = Device 0 passed diagnostics  
Bit 2-1: These bits indicate how Device 0 determined the device number.  
00 = Reserved  
01 = a jumper was used  
10 = the CSEL signal was used  
11 = some other method was used or the method is unknown  
Bit 0:  
1
*22 Word 94: Automatic Acoustic Management  
Bit 15-8:  
Bit 7-0:  
0
Current automatic acoustic management value  
5 - 34  
C141-E116-01EN  
(13)  
IDENTIFY DEVICE DMA (X'EE')  
When this command is not used to transfer data to the host in DMA mode, this command functions  
in the same way as the Identify Device command.  
At command issuance (I/O registers setting contents)  
1F7H(CM)  
1F6H(DH)  
1
1
1
0
1
1
1
0
DV  
xx  
×
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(FR)  
xx  
xx  
xx  
xx  
xx  
At command completion (I/O registers contents to be read)  
1F7H(ST)  
Status information  
DV  
1F6H(DH)  
xx  
×
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(ER)  
xx  
xx  
xx  
xx  
Error information  
(14)  
SET FEATURES (X'EF')  
The host system issues the SET FEATURES command to set parameters in the Features register  
for the purpose of changing the device features to be executed. For the transfer mode (Feature  
register = 03), detail setting can be done using the Sector Count register.  
Upon receipt of this command, the device sets the BSY bit of the Status register and saves the  
parameters in the Features register. Then, the device clears the BSY bit, and generates an  
interrupt.  
If the value in the Features register is not supported or it is invalid, the device posts an ABORTED  
COMMAND error.  
Table 5.6 lists the available values and operational modes that may be set in the Features register.  
C141-E116-01EN  
5 - 35  
Table 5.6 Features register values and settable modes  
Features Register  
X‘02’  
Drive operation mode  
Enables the write cache function.  
X‘03’  
Specifies the transfer mode. Supports PIO mode 4, single word DMA mode 2,  
and multiword DMA mode regardless of Sector Count register contents.  
X‘04’  
X‘05’  
X‘33’  
X‘42’  
X‘54’  
X‘55’  
X‘66’  
X‘77’  
X‘81’  
X‘82’  
X‘84’  
X‘85’  
X‘88’  
X‘89’  
X‘AA’  
X‘AB’  
X‘BB’  
No operation.  
Enable the advanced power management function.  
No operation.  
Enable Automatic Acoustic Management feature set  
No operation.  
Disables the read cache function.  
Disables the reverting to power-on default settings after software reset.  
No operation.  
No operation.  
Disables the write cache function.  
No operation.  
Disable the advanced power management function.  
No operation.  
No operation.  
Enables the read cache function.  
No operation.  
Specifies the transfer of 4-byte ECC for READ LONG and WRITE LONG  
commands.  
X‘C2’  
X‘CC’  
Disable Automatic Acoustic Management feature set  
Enables the reverting to power-on default settings after software reset.  
At power-on or after hardware reset, the default mode is the same as that is set with a value greater  
than X‘AA’ (except for write cache). If X‘66’ is specified, it allows the setting value greater than  
X‘AA’ which may have been modified to a new value since power-on, to remain the same even  
after software reset.  
5 - 36  
C141-E116-01EN  
At command issuance (I/O registers setting contents)  
1F7H(CM)  
1F6H(DH)  
1
1
1
0
1
1
1
1
DV  
xx  
×
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(FR)  
xx  
xx  
xx  
xx or transfer mode  
[See Table 5.6]  
At command completion (I/O registers contents to be read)  
1F7H(ST)  
Status information  
DV  
1F6H(DH)  
xx  
×
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(ER)  
xx  
xx  
xx  
xx  
Error information  
The host sets X'03' to the Features register. By issuing this command with setting a value to the  
Sector Count register, the transfer mode can be selected. Upper 5 bits of the Sector Count register  
defines the transfer type and lower 3 bits specifies the binary mode value.  
However, the IDD can operate with the PIO transfer mode 4 and multiword DMA transfer mode 2  
regardless of reception of the SET FEATURES command for transfer mode setting.  
The IDD supports following values in the Sector Count register value. If other value than below is  
specified, an ABORTED COMMAND error is posted.  
PIO default transfer mode  
00000 000 (X‘00’)  
PIO flow control transfer mode X  
00001 000 (X‘08’: Mode 0)  
00001 001 (X‘09’: Mode 1)  
00001 010 (X‘0A’: Mode 2)  
00001 011 (X‘0B’: Mode 3)  
00001 100 (X‘0C’: Mode 4)  
Multiword DMA transfer mode X  
Ultra DMA transfer mode X  
00100 000 (X‘20’: Mode 0)  
00100 001 (X‘21’: Mode 1)  
00100 010 (X‘22’: Mode 2)  
01000 000 (X‘40’: Mode 0)  
01000 001 (X‘41’: Mode 1)  
01000 010 (X‘42’: Mode 2)  
01000 011 (X‘43’: Mode 3)  
01000 100 (X‘44’: Mode 4)  
01000 101 (X‘45’: Mode 5)  
C141-E116-01EN  
5 - 37  
Subcommand code 42h allows the host to enable the Automatic Acoustic Management feature set.  
To enable the Automatic Acoustic Management feature set, the host writes the Sector Count  
register with the requested automatic acoustic management level and executes a SET FEATURES  
command with subcommand code 42h. The acoustic management level is selected on a scale from  
01h to FEh. Following table shows the acoustic management level values.  
Enabling or disabling of the Automatic Acoustic Management feature set, and the current  
automatic acoustic management level setting will be preserved by the device across all forms of  
reset, i.e., power on, hardware, and software resets.  
Automatic management levels  
Level  
Sector Count value  
FFh  
Reserved  
Maximum performance  
C0h - FEh  
80h - BFh  
01h - 7Fh  
00h  
Minimum acoustic emanation level  
Retired  
Vendor Specific (Maximum performance)  
Subcommand code C2h disables the Automatic Acoustic Management feature set. Devices that  
implement SET FEATURES subcommand 42h are not required to implement subcommand C2h.  
If device successfully completes execution of this subcommand, then the acoustic behavior of the  
device shall be vendor-specific, and the device return zeros in bits 0-7 of word 94 and bit 9 of  
word 86 of the IDENTIFY DEVICE data.  
(15)  
SET MULTIPLE MODE (X'C6')  
This command enables the device to perform the READ MULTIPLE and WRITE MULTIPLE  
commands. The block count (number of sectors in a block) for these commands are also specified  
by the SET MULTIPLE MODE command.  
The number of sectors per block is written into the Sector Count register. The IDD supports 2, 4,  
8 and 16 (sectors) as the block counts.  
Upon receipt of this command, the device sets the BSY bit of the Status register and checks the  
contents of the Sector Count register. If the contents of the Sector Count register is valid and is a  
supported block count, the value is stored for all subsequent READ MULTIPLE and WRITE  
MULTIPLE commands. Execution of these commands is then enabled. If the value of the Sector  
Count register is not a supported block count, an ABORTED COMMAND error is posted and the  
READ MULTIPLE and WRITE MULTIPLE commands are disabled.  
If the contents of the Sector Count register is 0 when the SET MULTIPLE MODE command is  
issued, the READ MULTIPLE and WRITE MULTIPLE commands are disabled.  
When the SET MULTIPLE MODE command operation is completed, the device clears the BSY  
bit and generates an interrupt.  
5 - 38  
C141-E116-01EN  
At command issuance (I/O registers setting contents)  
1F7H(CM)  
1F6H(DH)  
1
1
0
0
0
1
1
0
DV  
xx  
×
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(FR)  
xx  
xx  
xx  
Sector count/block  
xx  
At command completion (I/O registers contents to be read)  
1F7H(ST)  
Status information  
DV  
1F6H(DH)  
xx  
×
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(ER)  
xx  
xx  
xx  
Sector count/block  
Error information  
After power-on or after hardware reset, the READ MULTIPLE and WRITE MULTIPLE  
command operation are disabled as the default mode.  
Regarding software reset, the mode set prior to software reset is retained after software reset.  
The parameters for the multiple commands which are posted to the host system when the  
IDENTIFY DEVICE command is issued are listed below. See Subsection 5.3.2 for the  
IDENTIFY DEVICE command.  
Word 47 = 8010:  
Maximum number of sectors that can be transferred per interrupt by the  
READ MULTIPLE and WRITE MULTIPLE commands are 16 (fixed).  
Word 59 = 0000:  
= 01xx:  
The READ MULTIPLE and WRITE MULTIPLE commands are disabled.  
The READ MULTIPLE and WRITE MULTIPLE commands are enabled.  
"xx" indicates the current setting for number of sectors that can be  
transferred per interrupt by the READ MULTIPLE and WRITE  
MULTIPLE commands.  
e.g. 0110 = Block count of 16 has been set by the SET MULTIPLE MODE  
command.  
(16)  
EXECUTE DEVICE DIAGNOSTIC (X'90')  
This command performs an internal diagnostic test (self-diagnosis) of the device. This command  
usually sets the DRV bit of the Drive/Head register is to 0 (however, the DV bit is not checked). If  
two devices are present, both devices execute self-diagnosis.  
If device 1 is present:  
C141-E116-01EN  
5 - 39  
Both devices shall execute self-diagnosis.  
The device 0 waits for up to 5 seconds until device 1 asserts the PDIAG- signal.  
If the device 1 does not assert the PDIAG- signal but indicates an error, the device 0 shall  
append X‘80’ to its own diagnostic status.  
The device 0 clears the BSY bit of the Status register and generates an interrupt. (The device  
1 does not generate an interrupt.)  
A diagnostic status of the device 0 is read by the host system. When a diagnostic failure of the  
device 1 is detected, the host system can read a status of the device 1 by setting the DV bit  
(selecting the device 1).  
When device 1 is not present:  
The device 0 posts only the results of its own self-diagnosis.  
The device 0 clears the BSY bit of the Status register, and generates an interrupt.  
Table 5.7 lists the diagnostic code written in the Error register which is 8-bit code.  
If the device 1 fails the self-diagnosis, the device 0 "ORs" X‘80’ with its own status and sets that  
code to the Error register.  
Table 5.7 Diagnostic code  
Code  
Result of diagnostic  
No error detected.  
Data buffer compare error  
ROM sum check error  
Failure of device 1  
X‘01’  
X‘03’  
X‘05’  
X‘8x’  
At command issuance (I/O registers setting contents)  
1F7H(CM)  
1F6H(DH)  
1
0
0
1
0
0
0
0
DV  
xx  
×
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(FR)  
xx  
xx  
xx  
xx  
xx  
At command completion (I/O registers contents to be read)  
1F7H(ST)  
Status information  
00  
1F6H(DH)  
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(ER)  
00  
00  
01H  
01H  
Diagnostic code  
5 - 40  
C141-E116-01EN  
(17)  
FORMAT TRACK (X'50')  
Upon receipt of this command, the device sets the DRQ bit and waits the completion of 512-byte  
format parameter transfer from the host system. After completion of transfer, the device clears the  
DRQ bits, sets the BSY bit. However the device does not perform format operation, but the drive  
clears the BYS bit and generates an interrupt soon. When the command execution completes, the  
device clears the BSY bit and generates an interrupt.  
The drive supports this command for keep the compatibility with previous drive only.  
READ LONG (X'22' or X'23')  
(18)  
This command operates similarly to the READ SECTOR(S) command except that the device  
transfers the data in the requested sector and the ECC bytes to the host system. The ECC error  
correction is not performed for this command. This command is used for checking ECC function  
by combining with the WRITE LONG command.  
The READ LONG command supports only single sector operation.  
At command issuance (I/O registers setting contents)  
1F7H(CM)  
1F6H(DH)  
0
0
1
0
0
0
1
R
L
DV  
Head No. /LBA [MSB]  
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(FR)  
Cylinder No. [MSB] / LBA  
Cylinder No. [LSB] / LBA  
Sector No.  
/ LBA [LSB]  
Number of sectors to be transferred  
xx  
R = 0 or 1  
At command completion (I/O registers contents to be read)  
1F7H(ST) Status information  
DV Head No. /LBA [MSB]  
1F6H(DH)  
L
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(ER)  
Cylinder No. [MSB] / LBA  
Cylinder No. [LSB] / LBA  
Sector No.  
/ LBA [LSB]  
00 (*1)  
Error information  
*1 If the command is terminated due to an error, this register indicates 01.  
(19)  
WRITE LONG (X'32' or X'33')  
This command operates similarly to the READ SECTOR(S) command except that the device  
writes the data and the ECC bytes transferred from the host system to the disk medium. The device  
does not generate ECC bytes by itself. The WRITE LONG command supports only single sector  
operation.  
This command is operated under the following conditions:  
C141-E116-01EN  
5 - 41  
The command is issued in a sequence of the READ LONG or WRITE LONG (to the same  
address) command issuance. (WRITE LONG command can be continuously issued after the  
READ LONG command.)  
If above condition is not satisfied, the command operation is not guaranteed.  
At command issuance (I/O registers setting contents)  
1F7H(CM)  
1F6H(DH)  
0
0
1
1
0
0
1
R
L
DV  
Head No. /LBA [MSB]  
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(FR)  
Cylinder No. [MSB] / LBA  
Cylinder No. [LSB] / LBA  
Sector No.  
/ LBA [LSB]  
Number of sectors to be transferred  
xx  
R = 0 or 1  
At command completion (I/O registers contents to be read)  
1F7H(ST) Status information  
DV Head No. /LBA [MSB]  
1F6H(DH)  
L
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(ER)  
Cylinder No. [MSB] / LBA  
Cylinder No. [LSB] / LBA  
Sector No.  
/ LBA [LSB]  
00 (*1)  
Error information  
*1 If the command is terminated due to an error, this register indicates 01.  
(20)  
READ BUFFER (X'E4')  
The host system can read the current contents of the sector buffer of the device by issuing this  
command. Upon receipt of this command, the device sets the BSY bit of Status register and sets  
up the sector buffer for a read operation. Then the device sets the DRQ bit of Status register,  
clears the BSY bit, and generates an interrupt. After that, the host system can read up to 512 bytes  
of data from the buffer.  
At command issuance (I/O registers setting contents)  
1F7H(CM)  
1F6H(DH)  
1
1
1
0
0
1
0
0
DV  
xx  
×
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(FR)  
xx  
xx  
xx  
xx  
xx  
5 - 42  
C141-E116-01EN  
At command completion (I/O registers contents to be read)  
1F7H(ST)  
Status information  
DV  
1F6H(DH)  
xx  
×
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(ER)  
xx  
xx  
xx  
xx  
Error information  
(21)  
WRITE BUFFER (X'E8')  
The host system can overwrite the contents of the sector buffer of the device with a desired data  
pattern by issuing this command. Upon receipt of this command, the device sets the BSY bit of the  
Status register. Then the device sets the DRQ bit of Status register and clears the BSY bit when  
the device is ready to receive the data. After that, 512 bytes of data is transferred from the host  
and the device writes the data to the sector buffer, then generates an interrupt.  
At command issuance (I/O registers setting contents)  
1F7H(CM)  
1F6H(DH)  
1
1
1
0
1
0
0
0
DV  
xx  
×
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(FR)  
xx  
xx  
xx  
xx  
xx  
At command completion (I/O registers contents to be read)  
1F7H(ST)  
Status information  
DV  
1F6H(DH)  
xx  
×
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(ER)  
xx  
xx  
xx  
xx  
Error information  
C141-E116-01EN  
5 - 43  
(22)  
IDLE (X'97' or X'E3')  
Upon receipt of this command, the device sets the BSY bit of the Status register, and enters the  
idle mode. Then, the device clears the BSY bit, and generates an interrupt. The device generates an  
interrupt even if the device has not fully entered the idle mode. If the spindle of the device is  
already rotating, the spin-up sequence shall not be implemented.  
If the contents of the Sector Count register is other than 0, the automatic power-down function is  
enabled and the timer starts countdown immediately. When the timer reaches the specified time,  
the device enters the standby mode.  
If the contents of the Sector Count register is 0, the automatic power-down function is disabled.  
Enabling the automatic power-down function means that the device automatically enters the  
standby mode after a certain period of time. When the device enters the idle mode, the timer starts  
countdown. If any command is not issued while the timer is counting down, the device  
automatically enters the standby mode. If any command is issued while the timer is counting  
down, the timer is initialized and the command is executed. The timer restarts countdown after  
completion of the command execution.  
The period of timer count is set depending on the value of the Sector Count register as shown  
below.  
Sector Count register value  
[X'00']  
Point of timer  
Disable of timer  
0
1 to 240  
[X'01' to X'F0']  
(Value ×5) seconds  
(Value – 240) ×30 minutes  
21 minutes  
241 to 251 [X'F1' to X'FB']  
252  
253  
[X'FC']  
[X'FD']  
8 hours  
254 to 255 [X'FE' to X'FF']  
21 minutes 15 seconds  
At command issuance (I/O registers setting contents)  
1F7H(CM)  
X'97' or X'E3'  
DV  
1F6H(DH)  
xx  
×
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(FR)  
xx  
xx  
xx  
Period of timer  
xx  
5 - 44  
C141-E116-01EN  
At command completion (I/O registers contents to be read)  
1F7H(ST)  
Status information  
DV  
1F6H(DH)  
xx  
×
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(ER)  
xx  
xx  
xx  
xx  
Error information  
(23)  
IDLE IMMEDIATE (X'95' or X'E1')  
Upon receipt of this command, the device sets the BSY bit of the Status register, and enters the  
idle mode. Then, the device clears the BSY bit, and generates an interrupt. This command does not  
support the automatic power-down function.  
At command issuance (I/O registers setting contents)  
1F7H(CM)  
1F6H(DH)  
X'95' or X'E1'  
DV  
xx  
×
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(FR)  
xx  
xx  
xx  
xx  
xx  
At command completion (I/O registers contents to be read)  
1F7H(ST)  
Status information  
DV  
1F6H(DH)  
xx  
×
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(ER)  
xx  
xx  
xx  
xx  
Error information  
C141-E116-01EN  
5 - 45  
(24)  
STANDBY (X'96' or X'E2')  
Upon receipt of this command, the device sets the BSY bit of the Status register and enters the  
standby mode. The device then clears the BSY bit and generates an interrupt. The device  
generates an interrupt even if the device has not fully entered the standby mode. If the device has  
already spun down, the spin-down sequence is not implemented.  
If the contents of the Sector Count register is other than 0, the automatic power-down function is  
enabled and the timer starts countdown when the device returns to idle mode.  
When the timer value reaches 0 (passed a specified time), the device enters the standby mode.  
If the contents of the Sector Count register is 0, the automatic power-down function is disabled.  
Under the standby mode, the spindle motor is stopped. Thus, when the command involving a seek  
such as the READ SECTOR(s) command is received, the device processes the command after  
driving the spindle motor.  
At command issuance (I/O registers setting contents)  
1F7H(CM)  
1F6H(DH)  
X'96' or X'E2'  
DV  
xx  
×
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(FR)  
xx  
xx  
xx  
Period of timer  
xx  
At command completion (I/O registers contents to be read)  
1F7H(ST)  
Status information  
DV  
1F6H(DH)  
xx  
×
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(ER)  
xx  
xx  
xx  
xx  
Error information  
(25)  
STANDBY IMMEDIATE (X'94' or X'E0')  
Upon receipt of this command, the device sets the BSY bit of the Status register and enters the  
standby mode. The device then clears the BSY bit and generates an interrupt. This command does  
not support the automatic power-down sequence.  
5 - 46  
C141-E116-01EN  
At command issuance (I/O registers setting contents)  
1F7H(CM)  
X'94' or X'E0'  
DV  
1F6H(DH)  
xx  
×
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(FR)  
xx  
xx  
xx  
xx  
xx  
At command completion (I/O registers contents to be read)  
1F7H(ST)  
Status information  
DV  
1F6H(DH)  
xx  
×
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(ER)  
xx  
xx  
xx  
xx  
Error information  
(26)  
SLEEP (X'99' or X'E6')  
This command is the only way to make the device enter the sleep mode.  
Upon receipt of this command, the device sets the BSY bit of the Status register and enters the  
sleep mode. The device then clears the BSY bit and generates an interrupt. The device generates  
an interrupt even if the device has not fully entered the sleep mode.  
In the sleep mode, the spindle motor is stopped and the ATA interface section is inactive. All I/O  
register outputs are in high-impedance state.  
The only way to release the device from sleep mode is to execute a software or hardware reset.  
At command issuance (I/O registers setting contents)  
1F7H(CM)  
1F6H(DH)  
X'99' or X'E6'  
DV  
xx  
×
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(FR)  
xx  
xx  
xx  
xx  
xx  
C141-E116-01EN  
5 - 47  
At command completion (I/O registers contents to be read)  
1F7H(ST)  
Status information  
DV  
1F6H(DH)  
xx  
×
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(ER)  
xx  
xx  
xx  
xx  
Error information  
(27)  
CHECK POWER MODE (X'98' or X'E5')  
The host checks the power mode of the device with this command.  
The host system can confirm the power save mode of the device by analyzing the contents of the  
Sector Count and Sector Number registers.  
The device sets the BSY bit and sets the following register value. After that, the device clears the  
BSY bit and generates an interrupt.  
Power save mode  
Sector Count register  
X'00'  
Sector Number register  
N/A  
• During moving to standby mode  
• Standby mode  
• During returning from the standby mode  
• Idle mode  
X'FF'  
X'FF'  
X'00'  
X'FF'  
• Active mode  
5 - 48  
C141-E116-01EN  
At command issuance (I/O registers setting contents)  
1F7H(CM)  
X'98' or X'E5'  
DV  
1F6H(DH)  
xx  
×
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(FR)  
xx  
xx  
xx  
xx  
xx  
At command completion (I/O registers contents to be read)  
1F7H(ST)  
Status information  
DV  
1F6H(DH)  
xx  
×
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(ER)  
xx  
xx  
xx  
X'00' or X'FF'  
Error information  
(28)  
SMART (X'B0)  
This command performs operations for device failure predictions according to a subcommand  
specified in the FR register. If the value specified in the FR register is supported, the Aborted  
Command error is posted.  
It is necessary for the host to set the keys (CL = 4Fh and CH = C2h) in the CL and CH registers  
prior to issuing this command. If the keys are set incorrectly, the Aborted Command error is  
posted.  
When the failure prediction feature is disabled, the Aborted Command error is posted in response  
to subcommands other than SMART Enable Operations (FR register = D8h).  
When the failure prediction feature is enabled, the device collects or updates several items to  
forecast failures. In the following sections, note that the values of items collected or updated by  
the device to forecast failures are referred to as attribute values.  
C141-E116-01EN  
5 - 49  
Table 5.8 Features Register values (subcommands) and functions (1/2)  
Features Resister Function  
X’D0’  
X’D1’  
X’D2’  
SMART Read Attribute Values:  
A device that received this subcommand asserts the BSY bit and saves all the  
updated attribute values. The device then clears the BSY bit and transfers 512-byte  
attribute value information to the host.  
* For information about the format of the attribute value information, see Table 5.9.  
SMART Read Attribute Thresholds:  
This subcommand is used to transfer 512-byte insurance failure threshold value data  
to the host.  
* For information about the format of the insurance failure threshold value data, see  
Table 5.10.  
SMART Enable-Disable Attribute AutoSave:  
This subcommand is used to enable (SC register 00h) or disable (SC register =  
00h) the setting of the automatic saving feature for the device attribute data. The  
setting is maintained every time the device is turned off and then on. When the  
automatic saving feature is enabled, the attribute values are saved after 15 minutes  
passed since the previous saving of the attribute values. However, if the failure  
prediction feature is disabled, the attribute values are not automatically saved.  
When the device receives this subcommand, it asserts the BSY bit, enables or  
disables the automatic saving feature, then clears the BSY bit.  
SMART Save Attribute Values:  
X’D3’  
X’D4’  
When the device receives this subcommand, it asserts the BSY bit, saves device  
attribute value data, then clears the BSY bit.  
SMART Execute off-line Immediate/Execute Self Test:  
The device that received these subcommands shall execute off-line data collection or  
Self Test, or device shall abort current Self Test.  
The setting of SN register is described as following.  
Off-line data collection:  
Self Test functions:  
(SN register = 00h)  
- Quick Test – Off-line Mode  
(SN register = 01h)  
(SN register = 81h)  
(SN register = 02h)  
(SN register = 82h)  
(SN register = 7Fh)  
- Quick Test – Captive Mode  
- Comprehensive Test – Off-line Mode  
- Comprehensive Test – Captive Mode  
- Self Test Stop  
The device that received subcommand (SN register is described 00h, 01h or 02h)  
shall execute Off-line data collection or Self Test after asserts the BSY bit and clears  
it.  
The device that received subcommand (SN register is described 81h or 82h) shall  
execute Self Test after assert the BSY bit, then clears the BSY bit after completes  
these command process.  
The device that received subcommand (SN register is described 7Fh) shall assert the  
BSY bit. When the device is in process of performing Self Test or off-line data  
collection, it should abort the current Self Test or off-line data collection, then clears  
the BSY bit.  
5 - 50  
C141-E116-01EN  
Table 5.8 Features Register values (subcommands) and functions (2/2)  
X’D5’  
SMART Read Logging Data:  
This subcommand is used to transfer logging data that transfer length specified in SC  
register is to the host.  
The setting of SN register is described Log Sector Address below. When SN register  
is set 00h, 01h, or 06h, SC register should be set to 01h. When SN register is set  
80h-9Fh, SC register should be set 01h to maximum value that shows in the Log  
Directory.  
Log Sector Address  
01h (SC register 01h):  
06h (SC register 01h):  
80h-9Fh:  
SMART Error Log  
(See Table 5.12)  
SMART Self Test Log  
(See Table 5.13)  
Host vendor specific  
When the device receives these subcommands, it asserts the BSY bit and transfers  
logging data to the host, then clears the BSY bit.  
X’D6’  
SMART Write Logging Data:  
The device receives the logging data that transfer length specified in SC register from  
the host. SN register should be set the Log Sector Address below, and SC register  
should be set 01h to maximum value that shows in the Log Directory.  
Log Sector Address  
80h-9Fh:  
Host vendor specific  
When the device receives these subcommands, it asserts the BSY bit and receives  
logging data from the host, then clears the BSY bit.  
X’D8’  
X’D9’  
X’DA’  
SMART Enable Operations:  
This subcommand enables the failure prediction feature. The setting is maintained  
even when the device is turned off and then on.  
When the device receives this subcommand, it asserts the BSY bit, enables the failure  
prediction feature, then clears the BSY bit.  
SMART Disable Operations:  
This subcommand disables the failure prediction feature. The setting is maintained  
even when the device is turned off and then on.  
When the device receives this subcommand, it asserts the BSY bit, disables the  
failure prediction feature, then clears the BSY bit.  
SMART Return Status:  
When the device receives this subcommand, it asserts the BSY bit and saves the  
current device attribute values. Then the device compares the device attribute values  
with insurance failure threshold values. If there is an attribute value exceeding the  
threshold, F4h and 2Ch are loaded into the CL and CH registers. If there are no  
attribute values exceeding the thresholds, 4Fh and C2h are loaded into the CL and  
CH registers. After the settings for the CL and CH registers have been determined,  
the device clears the BSY bit.  
X’DB’  
SMART Enable/Disable Automatic Off-line:  
The device that receives this subcommand enables (SC register 00h) or disables  
(SC register = 00h) automatic off-line data collection. The condition is maintained  
even when the device is turned off and turned on.  
When the failure prediction feature and automatic off-line data collection are  
enabled, an off-line data collection shall be performed regardless of host issued  
command after more than 4 hours passed since power-on or the previous off-line data  
collection.  
When the device receives this subcommand, it asserts the BSY bit, enables or  
disables automatic off-line data collection, then clears the BSY bit.  
C141-E116-01EN  
5 - 51  
The host must regularly issue the SMART Read Attribute Values subcommand (FR register =  
D0h), SMART Save Attribute Values subcommand (FR register = D3h), or SMART Return Status  
subcommand (FR register = DAh) to save the device attribute value data on a medium.  
Alternative, the device must issue the SMART Enable-Disable Attribute AutoSave subcommand  
(FR register = D2h) to use a feature which regularly save the device attribute value data to a  
medium.  
The host can predict failures in the device by periodically issuing the SMART Return Status  
subcommand (FR register = DAh) to reference the CL and CH registers.  
If an attribute value is below the insurance failure threshold value, the device is about to fail or the  
device is nearing the end of it life . In this case, the host recommends that the user quickly backs  
up the data.  
At command issuance (I-O registers setting contents)  
1F7H(CM)  
1F6H(DH)  
1
0
1
1
0
0
0
0
DV  
xx  
×
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(FR)  
Key (C2h)  
Key (4Fh)  
xx  
xx  
Subcommand  
At command completion (I-O registers setting contents)  
1F7H(ST)  
Status information  
DV  
1F6H(DH)  
xx  
×
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(ER)  
Key-failure prediction status (C2h-2Ch)  
Key-failure prediction status (4Fh-F4h)  
xx  
xx  
Error information  
The attribute value information is 512-byte data; the format of this data is shown in Table 5.9. The  
host can access this data using the SMART Read Attribute Values subcommand (FR register =  
D0h). The insurance failure threshold value data is 512-byte data; the format of this data is shown  
in Table 5.10. The host can access this data using the SMART Read Attribute Thresholds  
subcommand (FR register = D1h).  
5 - 52  
C141-E116-01EN  
Table 5.9 Device attribute data structure  
Byte (Hex)  
Description  
00  
01  
Data structure revision number *1  
1st attribute  
Attribute ID number *2  
Status flag *3  
02  
03  
04  
05  
Normalized attribute values *4  
Worst ever normalized *5  
Raw attribute values *6  
06  
07  
to  
0C  
0D  
Reserved  
2nd to 30th  
attribute  
Reserved  
0E  
to  
169  
(Each attribute format is the same as 1st attribute.)  
Off-line data collection status *7  
Self Test execution status byte *8  
Off-line data collection executing times (sec)  
16A  
16B  
Off-line data  
collection  
16C  
16D  
16E  
Reserved  
16F  
Off-line data collection capability *9  
170  
171  
172  
SMART capability flag *10  
Drive error logging capability *11  
Self Test failure checkpoint  
173  
174  
175  
Quick Test completion time (min) *12  
Comprehensive Test completion time (min) *13  
Reserved  
176  
to  
181  
182  
to  
1FE  
1FF  
Vendor unique  
Check sum *14  
C141-E116-01EN  
5 - 53  
Table 5.10 Warranty failure threshold data structure  
Byte (Hex)  
Description  
Data structure revision number *1  
00  
01  
1st drive threshold Attribute ID number *2  
Attribute threshold *15  
02  
03  
04  
05  
06  
Reserved  
07  
to  
0C  
0D  
2nd to 30th drive  
threshold  
Reserved  
0E  
to  
169  
(Each threshold format is the same as 1st drive  
threshold.)  
16A  
to  
Reserved  
17B  
17C  
to  
Vendor unique  
Check sum  
1FE  
1FF  
*1 Data structure revision number  
It indicates the revision number of device attribute and warranty failure threshold. They will  
have the same “Data structure revision number”.  
5 - 54  
C141-E116-01EN  
*2 Attribute ID  
The attribute ID is defined as follows:  
Attribute ID (Dec)  
Description  
(Indicates unused attribute data)  
0
1
Read error rate  
2
Throughput performance  
Spin up time  
3
4
Number of times the spindle motor is activated  
Number of alternative sectors  
Read channel margin (Not supported)  
Seek error rate  
5
6
7
8
9
Seek time performance  
Power-on time  
10  
Number of retries made to activate the spindle motor  
Number of retries to calibration  
Number of turn on/off times  
Reserved  
11  
12  
13 to 193  
194  
HDA Temperature  
195 to 198  
199  
Reserved  
Ultra ATA CRC Error Rate  
Write error rate  
200  
201 to 255  
(Vendor unique)  
C141-E116-01EN  
5 - 55  
*3 Status flag  
Bit  
Description  
0
If this bit is set to 1, it indicates the attribute is guaranteed for  
normal operation when an attribute value exceeds the threshold.  
1
If this bit is set to 1 (0), it indicates the attribute is updated only  
by on-line test (off-line test).  
2
3
4
5
If this bit is set to 1, it indicates a performance attribute.  
If this bit is set to 1, it indicates an error-rate attribute.  
If this bit is set to 1, it indicates an event count attribute.  
If this bit is set to 1, it indicates the attribute shall be collected and  
saved even if the failure prediction feature is disabled.  
6 to 15  
Reserved bits  
*4 Normalized attribute value  
The current attribute value is the normalized raw attribute data. The value varies between 01h  
and 64h. The closer the value gets to 01h, the higher the possibility of a failure. The device  
compares the attribute values with thresholds. When the attribute values are larger than the  
thresholds, the device is operating normally.  
*5 Worst ever normalized  
This is the worst attribute value among the attribute values collected to date. This value  
indicates the state nearest to a failure so far.  
*6 Raw attribute value  
Raw attributes data is retained.  
5 - 56  
C141-E116-01EN  
*7 Off-line data collection status  
Values  
Description  
00h or 80h  
01h or 81h  
02h or 82h  
03h or 83h  
04h or 84h  
Off-line data collection is not started.  
Reserved  
Off-line data collection has been completed without error.  
Reserved  
Off-line data collection has been suspended by an interrupt  
command from the host.  
05h or 85h  
06h or 86h  
Off-line data collection has been aborted by an interrupt command  
from the host.  
Off-line data collection has been aborted with a fatal error. (Not  
used)  
40h to 7Fh  
C0h to FFh  
Vendor unique (Not used)  
07h to 3Fh  
87h to BFh  
Reserved  
If bit [7] is 1, it indicates that automatic off-line data collection function is enabled.  
*8 Self Test execution status byte [16Bh]  
Bit 0-3: Self Test remain time.  
The values in these bits indicate the remaining percentage (0% - 90%) of Self Test until  
completion by 0h-9h.  
Bit 4-7: Self Test execution status  
00h: Self Test has been completed normally. Otherwise Self Test has not  
performed.  
01h: Self Test has been interrupted by host.  
02h: Self Test has been interrupted by the soft/hard reset from the host.  
03h: Self Test has been aborted for a final error.  
04h: Self Test has been completed abnormally for an unknown meaning.  
05h: Self Test has been completed abnormally by Write/Read Test.  
06h: Self Test has been completed abnormally by servo analysis.  
07h: Self Test has been completed abnormally by Read Scan Test  
08h-0Eh:  
Reserved  
0Fh: Self Test is being performed.  
C141-E116-01EN  
5 - 57  
*9 Off-line data collection capability [16Fh]  
Bit  
Description  
0
If this bit is set to 1, it indicates SMART EXECUTE OFF-LINE  
IMMEDIATE subcommand is supported. (FR register = D4)  
1
2
Vendor unique  
If this bit is set to 1, it indicates off-line data collection being  
aborted when a new command is received.  
3
If this bit is set to 1, it indicates SMART off-line read scanning is  
supported.  
4
If this bit is set to 1, it indicates SMART Self Test is supported.  
Reserved bits  
5 to 7  
*10  
SMART capability flag [170-171h]  
Bit  
Description  
0
If this bit is set to 1, it indicates the attribute data is saved to the  
media before the drive enters power save mode.  
1
If this bit is set to 1, it indicates the device saves the attribute  
automatically according to fixed operation.  
2-15  
Reserved bits  
*11  
Drive error logging capability [172h]  
Bit  
Description  
0
If this bit is set to 1, it indicates the drive error logging is  
supported.  
1 to 15  
Reserved bits  
*12  
*13  
*14  
Quick Test completion time [minutes]  
This value indicates the processing time of the Quick Test (off-line mode).  
Comprehensive Test completion time [minutes]  
This value indicates the processing time of the Comprehensive Test (off-line mode).  
Check sum  
Twos complement of the lower byte, obtained by adding 511-byte data one by at a time from  
the beginning.  
5 - 58  
C141-E116-01EN  
*15  
Attribute threshold  
The limit of a varying attribute value. The host compares the attribute values with the  
thresholds to identify a failure.  
Table 5.11 Log directory data structure  
Address  
Description  
(Hex)  
00 to 01  
02  
03 to 0B  
0C  
0D to FF  
100  
SMART logging version  
Number of sectors for Log Sector Address 01h  
Reserved  
Number of sectors for Log Sector Address 06h  
Reserved  
Number of sectors  
Reserved  
Address 80h  
101  
Address 81h  
Data format is the same at Address 100  
to 101  
102 to 13F  
140 to 1FF  
to  
Address 9Fh  
Reserved  
C141-E116-01EN  
5 - 59  
Table 5.12 Error logging data structure  
Address  
(Hex)  
Description  
00  
01  
SMART error logging version 01h  
Index pointer of latest error data structure  
02 to 31  
32  
Reserved  
Command data  
structure  
Error log data  
stucture 1  
Device control register  
Features register  
33  
34  
35  
36  
37  
38  
39  
Sector count register  
Sector number register  
Cylinder low register  
Cylinder high register  
Drive/Head register  
Command register  
Elapsed time from power-on  
[ms]  
3A to 3D  
3E  
3F  
40  
41  
42  
43  
44  
45  
Reserved  
Error register  
Error data  
structure  
Sector count register  
Sector number register  
Cylinder low register  
Cylinder high register  
Drive/Head register  
Status register  
46 to 58  
59  
Vendor unique  
State  
5A  
5B  
Cumulative elapsed time [h]  
5C to 1C3  
Error log data  
structure 2-5  
Data structure same as error log data structure 1  
1C4  
1C5  
Total error count  
1C6 to 1FE Reserved  
1FF Check sum  
5 - 60  
C141-E116-01EN  
Table 5.13 Self Test log data structure  
Byte  
Description  
0 to 1  
Data Structure Revision Number = 0x0001  
1st Descriptor Entry  
2
3
Self Test Number  
Self Test Execution Status  
Life time power on hours  
Self Test Error No.  
4 to 5  
6
7 to 10  
11 to 25  
26 to 49  
482 to 505  
506 to 507  
508  
Error LBA  
Vender Specific bytes  
2nd Descriptor Entry  
21st Descriptor Entry  
Vender Specific  
Self Test Index  
Reserved  
509 to 510  
511  
Checksum  
Self Test Number  
This byte shows executed Self Test Number.  
Self Test Execution Status  
Refer to explanation of Attribute Byte 16Bh.  
Self Test Index  
When it is “00h”, it means that Self Test has not been performed.  
Check Sum  
Twos complement of the lower byte, obtained by adding 511-byte data one by at a  
time from the beginning.  
*16  
Command data structure  
It indicates that the device has received structure of the command when the drive occurs the  
error.  
*17  
*18  
*19  
Error data structure  
It indicates that structure of the ATA taskfile register which the drive occurs the error.  
Total error count  
It indicates that the total count of the error registered in the error log.  
Check sum  
Twos complement of the lower byte, obtained by adding 511-byte data one by at a time from  
the beginning.  
C141-E116-01EN  
5 - 61  
(29)  
FLUSH CACHE (X ‘E7’)  
This command is use by the host to request the device to flush the write cache. If the write cache  
is to be flushed, all data cached shall be written to the media. The BSY bit shall remain set to one  
until all data has been successfully written or an error occurs. The device should use all error  
recovery methods available to ensure the data is written successfully. The flushing of write cache  
may take several seconds to complete depending upon the amount of data to be flushed and the  
success of the operation.  
NOTE: This command may take longer than 30 s to complete.  
If the command is not supported, the device shall set the ABRT bit to one. An unrecoverable error  
encountered during execution of writing data results in the termination of the command and the  
Command Block registers contain the sector address of the sector where the first unrecoverable  
error occurred. The sector is removed from the cache. Subsequent FLUSH CACHE commands  
continue the process of flushing the cache.  
At command issuance (I/O registers setting contents)  
1F7H(CM)  
1F6H(DH)  
X'E7'  
DV  
xx  
×
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(FR)  
xx  
xx  
xx  
xx  
xx  
At command completion (I/O registers contents to be read)  
1F7H(ST)  
Status information  
DV  
1F6H(DH)  
xx  
×
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(ER)  
xx  
xx  
xx  
xx  
Error information  
5 - 62  
C141-E116-01EN  
(30)  
SECURITY DISABLE PASSWORD (F6h)  
This command invalidates the user password already set and releases the lock function.  
The host transfers the 512-byte data shown in Table5.14 to the device. The device compares the  
user password or master password in the transferred data with the user password or master  
password already set, and releases the lock function if the passwords are the same.  
Although this command invalidates the user password, the master password is retained. To  
recover the master password, issue the SECURITY SET PASSWORD command and reset the user  
password.  
If the user password or master password transferred from the host does not match, the Aborted  
Command error is returned.  
Issuing this command while in LOCKED MODE or FROZEN MODE returns the Aborted  
Command error.  
(The section about the SECURITY FREEZE LOCK command describes LOCKED MODE and  
FROZEN MODE.)  
Table 5.14 Contents of security password  
Word  
0
Contents  
Control word  
Bit 0: Identifier  
0 = Compares the user passwords.  
1 = Compares the master passwords.  
Bits 1 to 15: Reserved  
Password (32 bytes)  
1 to 16  
17 to 255  
Reserved  
C141-E116-01EN  
5 - 63  
At command issuance (I-O registers setting contents)  
1F7H(CM)  
1F6H(DH)  
1
1
1
1
0
1
1
0
DV  
xx  
×
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(FR)  
xx  
xx  
xx  
xx  
xx  
At command completion (I-O registers contents to be read)  
1F7H(ST)  
Status information  
DV  
1F6H(DH)  
xx  
×
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(ER)  
xx  
xx  
xx  
xx  
Error information  
(31)  
SECURITY ERASE PREPARE (F3h)  
The SECURITY ERASE UNIT command feature is enabled by issuing the SECURITY ERASE  
PREPARE command and then the SECURITY ERASE UNIT command. The SECURITY  
ERASE PREPARE command prevents data from being erased unnecessarily by the SECURITY  
ERASE UNIT command.  
Issuing this command during FROZEN MODE returns the Aborted Command error.  
5 - 64  
C141-E116-01EN  
At command issuance (I-O registers setting contents)  
1F7H(CM)  
1F6H(DH)  
1
1
1
1
0
0
1
1
DV  
xx  
×
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(FR)  
xx  
xx  
xx  
xx  
xx  
At command completion (I-O registers contents to be read)  
1F7H(ST)  
Status information  
DV  
1F6H(DH)  
xx  
×
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(ER)  
xx  
xx  
xx  
xx  
Error information  
(32)  
SECURITY ERASE UNIT (F4h)  
This command erases all user data. This command also invalidates the user password and releases  
the lock function.  
The host transfers the 512-byte data shown in Table 5.14 to the device. The device compares the  
user password or master password in the transferred data with the user password or master  
password already set. The device erases user data, invalidates the user password, and releases the  
lock function if the passwords are the same.  
Although this command invalidates the user password, the master password is retained. To  
recover the master password, issue the SECURITY SET PASSWORD command and reset the user  
password.  
If the SECURITY ERASE PREPARE command is not issued immediately before this command is  
issued, the Aborted Command error is returned.  
Issuing this command while in FROZEN MODE returns the Aborted Command error.  
C141-E116-01EN  
5 - 65  
At command issuance (I-O registers setting contents)  
1F7H(CM)  
1F6H(DH)  
1
1
1
1
0
1
0
0
DV  
xx  
×
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(FR)  
xx  
xx  
xx  
xx  
xx  
At command completion (I-O registers contents to be read)  
1F7H(ST)  
Status information  
DV  
1F6H(DH)  
xx  
×
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(ER)  
xx  
xx  
xx  
xx  
Error information  
(33)  
SECURITY FREEZE LOCK (F5h)  
This command puts the device into FROZEN MODE. The following commands used to change  
the lock function return the Aborted Command error if the device is in FROZEN MODE.  
SECURITY SET PASSWORD  
SECURITY UNLOCK  
SECURITY DISABLE PASSWORD  
SECURITY ERASE PREPARE  
SECURITY ERASE UNIT  
FROZEN MODE is canceled when the power is turned off. If this command is reissued in  
FROZEN MODE, the command is completed and FROZEN MODE remains unchanged.  
Issuing this command during LOCKED MODE returns the Aborted Command error.  
The following medium access commands return the Aborted Command error when the device is in  
LOCKED MODE:  
5 - 66  
C141-E116-01EN  
READ DMA  
WRITE LONG  
SECURITY DISABLE PASSWORD  
SECURITY FREEZE LOCK  
SECURITY SET PASSWORD  
FORMAT TRACK  
READ LONG  
WRITE MULTIPLE  
WRITE SECTORS  
WRITE VERIFY  
SET MAX ADDRESS  
READ MULTIPLE  
READ SECTORS  
READ VERIFY  
WRITE DMA  
FLASH CACHE  
At command issuance (I-O registers setting contents)  
1F7H(CM)  
1F6H(DH)  
1
1
1
1
0
1
0
1
DV  
xx  
×
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(FR)  
xx  
xx  
xx  
xx  
xx  
At command completion (I-O registers contents to be read)  
1F7H(ST)  
Status information  
DV  
1F6H(DH)  
xx  
×
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(ER)  
xx  
xx  
xx  
xx  
Error information  
(34)  
SECURITY SET PASSWORD (F1h)  
This command enables a user password or master password to be set.  
The host transfers the 512-byte data shown in Table 5.15 to the device. The device determines the  
operation of the lock function according to the specifications of the Identifier bit and Security level  
bit in the transferred data. (Table 5.16)  
Issuing this command in LOCKED MODE or FROZEN MODE returns the Aborted Command  
error.  
C141-E116-01EN  
5 - 67  
Table 5.15 Contents of SECURITY SET PASSWORD data  
Word  
0
Contents  
Control word  
Bit 0 Identifier  
0 = Sets a user password.  
1 = Sets a master password.  
Bits 1 to 7 Reserved  
Bit 8 Security level  
0 = High  
1 = Maximum  
Bits 9 to 15 Reserved  
Password (32 bytes)  
Reserved  
1 to 16  
17 to 255  
Table 5.16 Relationship between combination of Identifier and Security level, and  
operation of the lock function  
Indentifier  
User  
Level  
High  
Description  
The specified password is saved as a new user password. The lock  
function is enabled after the device is turned off and then on.  
LOCKED MODE can be canceled using the user password or the  
master password already set.  
Master  
User  
High  
The specified password is saved as a new master password. The  
lock function is not enabled.  
Maximum  
The specified password is saved as a new user password. The lock  
function is enabled after the device is turned off and then on.  
LOCKED MODE can be canceled using the user password only.  
The master password already set cannot cancel LOCKED MODE.  
Master  
Maximum  
The specified password is saved as a new master password. The  
lock function is not enabled.  
5 - 68  
C141-E116-01EN  
At command issuance (I-O registers setting contents)  
1F7H(CM)  
1F6H(DH)  
1
1
1
1
0
0
0
1
DV  
xx  
×
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(FR)  
xx  
xx  
xx  
xx  
xx  
At command completion (I-O registers contents to be read)  
1F7H(ST)  
Status information  
DV  
1F6H(DH)  
xx  
×
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(ER)  
xx  
xx  
xx  
xx  
Error information  
(35)  
SECURITY UNLOCK (F2h)  
This command cancels LOCKED MODE.  
The host transfers the 512-byte data shown in Table 5.14 to the device. Operation of the device  
varies as follows depending on whether the host specifies the master password or user password.  
When the master password is selected  
When the security level in LOCKED MODE is high, the password is compared with the master  
password already set. If the passwords are the same, LOCKED MODE is canceled.  
Otherwise, the Aborted Command error is returned. If the security level in LOCKED MODE  
is set to the highest level, the Aborted Command error is always returned.  
When the user password is selected  
The password is compared with the user password already set. If the passwords are the same,  
LOCKED MODE is canceled. Otherwise, the Aborted Command error is returned.  
If the password comparison fails, the device decrements the UNLOCK counter. The UNLOCK  
counter initially has a value of five. When the value of the UNLOCK counter reaches zero, this  
command or the SECURITY ERASE UNIT command causes the Aborted Command error until  
the device is turned off and then on, or until a hardware reset is executed. Issuing this command  
with LOCKED MODE canceled (in UNLOCK MODE) has no affect on the UNLOCK counter.  
Issuing this command in FROZEN MODE returns the Aborted Command error.  
C141-E116-01EN  
5 - 69  
At command issuance (I-O registers setting contents)  
1F7H(CM)  
1F6H(DH)  
1
1
1
1
0
0
1
0
DV  
xx  
×
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(FR)  
xx  
xx  
xx  
xx  
xx  
At command completion (I-O registers contents to be read)  
1F7H(ST)  
Status information  
DV  
1F6H(DH)  
xx  
×
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(ER)  
xx  
xx  
xx  
xx  
Error information  
(36-1) SET MAX ADDRESS (F9)  
This command allows the maximum address accessible by the user to be set in LBA or CHS mode.  
Upon receipt of the command, the device sets the BSY bit and saves the maximum address  
specified in the DH, CH, CL and SN registers. Then, it clears BSY and generates an interrupt.  
The new address information set by this command is reflected in Words 1, 54, 57, 58, 60 and 61 of  
IDENTIFY DEVICE information. If an attempt is made to perform a read or write operation for an  
address beyond the new address space, an ID Not Found error will result.  
When SC register bit 0, VV (Value Volatile), is 1, the value set by this command is held even after  
power on and the occurrence of a hard reset. When the VV bit is 0, the value set by this command  
becomes invalid when the power is turned on or a hard reset occurs, and the maximum address  
returns to the value (default value if not set) most lately set when VV bit = 1.  
After power on and the occurrence of a hard reset, the host can issue this command only once  
when VV bit = 1. If this command with VV bit = 1 is issued twice or more, any command  
following the first time will result in an Aborted Command error.  
5 - 70  
C141-E116-01EN  
At command issuance (I/O registers setting contents)  
1F7H(CM)  
1F6H(DH)  
1
1
1
1
1
0
0
1
L
DV  
Max head/LBA [MSB]  
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
Max. cylinder [MSB]/Max. LBA  
Max. cylinder [LSB]/Max. LBA  
Max. sector/Max. LBA [LSB]  
1F2H(SC)  
1F1H(FR)  
xx  
xx  
VV  
At command completion (I/O registers contents to be read)  
1F7H(ST) Status information  
DV Max head/LBA [MSB]  
1F6H(DH)  
×
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(ER)  
Max. cylinder [MSB]/Max. LBA  
Max. cylinder [LSB]/Max. LBA  
Max. sector/Max. LBA [LSB]  
xx  
Error information  
(36-2) SET MAX SET PASSWORD (F9)  
This command requests a transfer of single sector of data from the host, and defines the content of  
this sector of information. The password is retained by the device until the next power cycle.  
At command issuance (I/O registers setting contents)  
1F7H(CM)  
1F6H(DH)  
1
1
1
1
1
1
1
0
0
0
1
1
L
DV  
xx  
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
xx  
xx  
xx  
xx  
1F1H(FR)  
0
0
0
0
0
0
C141-E116-01EN  
5 - 71  
At command completion (I/O registers contents to be read)  
1F7H(ST)  
Status information  
DV  
1F6H(DH)  
1
L
1
xx  
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(ER)  
xx  
xx  
xx  
xx  
Error information  
SET MAX SET PASSWORD data content  
Word Content  
0
Reserved  
1 - 16  
17 - 255  
Password (32 bytes)  
Reserved  
(36-3) SET MAX LOCK (F9)  
After this command is completed any other Set Max commands except SET MAX UNLOCK and  
SET MAX FREEZE LOCK are rejected. The device remains in this state until a power cycle or  
the acceptance of A SET MAX UNLOCK or SET MAX FREEZE LOCK command.  
At command issuance (I/O registers setting contents)  
1F7H(CM)  
1F6H(DH)  
1
1
1
1
1
1
1
0
0
1
0
L
DV  
xx  
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
xx  
xx  
xx  
xx  
1F1H(FR)  
0
0
0
0
0
0
1
At command completion (I/O registers contents to be read)  
1F7H(ST)  
Status information  
DV  
1F6H(DH)  
1
L
1
xx  
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(ER)  
xx  
xx  
xx  
xx  
Error information  
5 - 72  
C141-E116-01EN  
(36-4) SET MAX UNLOCK (F9)  
This command requests a transfer of a single sector of data from the host.  
The password supplied in the sector of data transferred shall be compared with the stored SET  
MAX password.  
If the password compare fails, then the device returns command aborted and decrements the  
unlock counter. On the acceptance of the SET MAX LOCK command, this counter is set to a  
value of five and shall be decremented for each password mismatch when SET MAX UNLOCK is  
issued and the device is locked.  
When this counter reaches zero, then the SET MAX UNLOCK command shall return command  
aborted until a power cycle.  
If the password compare matches, then the device shall make  
Set_Max_Unlocked state and all SET MAX commands will be accepted.  
a
transition to the  
At command issuance (I/O registers setting contents)  
1F7H(CM)  
1F6H(DH)  
1
1
1
1
1
1
1
0
0
1
1
L
DV  
xx  
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
xx  
xx  
xx  
xx  
1F1H(FR)  
0
0
0
0
0
0
1
At command completion (I/O registers contents to be read)  
1F7H(ST)  
Status information  
DV  
1F6H(DH)  
1
L
1
xx  
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(ER)  
xx  
xx  
xx  
xx  
Error information  
(36-5) SET MAX FREEZE LOCK (F9)  
This SET MAX FREEZE LOCK command sets the device to Set_Max_Frozen state. After  
command completion any subsequent SET MAX commands are rejected. Commands disabled by  
SET MAX FREEZE LOCK are:  
C141-E116-01EN  
5 - 73  
SET MAX ADDRESS  
SET MAX SET PASSWORD  
SET MAX LOCK  
SET MAX UNLOCK  
At command issuance (I/O registers setting contents)  
1F7H(CM)  
1F6H(DH)  
1
1
1
1
1
1
1
0
0
0
1
0
L
DV  
xx  
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
xx  
xx  
xx  
xx  
1F1H(FR)  
0
0
0
0
0
1
At command completion (I/O registers contents to be read)  
1F7H(ST)  
Status information  
DV  
1F6H(DH)  
1
L
1
xx  
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(ER)  
xx  
xx  
xx  
xx  
Error information  
(37)  
READ NATIVE MAX ADDRESS (F8)  
This command posts the maximum address intrinsic to the device, which can be set by the SET  
MAX ADDRESS command. Upon receipt of this command, the device sets the BSY bit and  
indicates the maximum address in the DH, CH, CL and SN registers. Then, it clears BSY and  
generates an interrupt.  
At command issuance (I/O registers setting contents)  
1F7H(CM)  
1F6H(DH)  
1
1
1
1
1
0
0
0
L
DV  
xx  
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(FR)  
xx  
xx  
xx  
xx  
xx  
5 - 74  
C141-E116-01EN  
At command completion (I/O registers contents to be read)  
1F7H(ST) Status information  
DV Max head/LBA [MSB]  
1F6H(DH)  
×
×
×
1F5H(CH)  
1F4H(CL)  
1F3H(SN)  
1F2H(SC)  
1F1H(ER)  
Max. cylinder [MSB]/Max. LBA  
Max. cylinder [LSB]/Max. LBA  
Max. sector/Max. LBA [LSB]  
xx  
Error information  
C141-E116-01EN  
5 - 75  
5.3.3  
Error posting  
Table 5.17 lists the defined errors that are valid for each command.  
Table 5.17 Command code and parameters  
Command name  
Error register (X'1F1')  
Status register (X'1F7')  
ICRC UNC  
V
INDF  
V
ABRT  
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
*
TR0NF  
DRDY  
V
DWF  
V
ERR  
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
READ SECTOR(S)  
WRITE SECTOR(S)  
READ MULTIPLE  
WRITE MULTIPLE  
READ DMA  
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
WRITE DMA  
V
V
V
WRITE VERIFY  
V
V
V
V
V
READ VERIFY SECTOR(S)  
RECALIBRATE  
V
V
V
V
V
V
SEEK  
V
V
V
INITIALIZE DEVICE PARAMETERS  
IDENTIFY DEVICE  
IDENTIFY DEVICE DMA  
SET FEATURES  
V
V
V
V
V
V
V
V
SET MULTIPLE MODE  
EXECUTE DEVICE DIAGNOSTIC  
FORMAT TRACK  
V
V
*
*
*
*
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
V
READ LONG  
WRITE LONG  
READ BUFFER  
WRITE BUFFER  
IDLE  
IDLE IMMEDIATE  
STANDBY  
STANDBY IMMEDIATE  
SLEEP  
CHECK POWER MODE  
SMART  
V
V
FLUSH CACHE  
SECURITY DISABLE PASSWORD  
SECURITY ERASE PREPARE  
SECURITY ERASE UNIT  
SECURITY FREEZE LOCK  
SECURITY SET PASSWORD  
SECURITY UNLOCK  
SET MAX ADDRESS  
READ NATIVE MAX ADDRESS  
V
Invalid command  
V: Valid on this command  
*: See the command descriptions.  
5 - 76  
C141-E116-01EN  
5.4  
Command Protocol  
The host should confirm that the BSY bit of the Status register of the device is 0 prior to issue a  
command. If BSY bit is 1, the host should wait for issuing a command until BSY bit is cleared to  
0.  
Commands can be executed only when the DRDY bit of the Status register is 1. However, the  
following commands can be executed even if DRDY bit is 0.  
EXECUTE DEVICE DIAGNOSTIC  
INITIALIZE DEVICE PARAMETERS  
5.4.1  
Data transferring commands from device to host  
The execution of the following commands involves data transfer from the device to the host.  
IDENTIFY DEVICE  
IDENTIFY DEVICE DMA  
READ SECTOR(S)  
READ LONG  
READ BUFFER  
SMART: SMART Read Attribute Values, SMART Read Attribute Thresholds  
The execution of these commands includes the transfer one or more sectors of data from the device  
to the host. In the READ LONG command, 516 bytes are transferred. Following shows the  
protocol outline.  
a) The host writes any required parameters to the Features, Sector Count, Sector Number,  
Cylinder, and Device/Head registers.  
b) The host writes a command code to the Command register.  
c) The device sets the BSY bit of the Status register and prepares for data transfer.  
d) When one sector (or block) of data is available for transfer to the host, the device sets DRQ bit  
and clears BSY bit. The drive then asserts INTRQ signal.  
e) After detecting the INTRQ signal assertion, the host reads the Status register. The host reads  
one sector of data via the Data register. In response to the Status register being read, the  
device negates the INTRQ signal.  
f) The drive clears DRQ bit to 0. If transfer of another sector is requested, the device sets the  
BSY bit and steps d) and after are repeated.  
Even if an error is encountered, the device prepares for data transfer by setting the DRQ bit.  
Whether or not to transfer the data is determined for each host. In other words, the host should  
receive the relevant sector of data (512 bytes of uninsured dummy data) or release the DRQ status  
by resetting.  
Figure 5.2 shows an example of READ SECTOR(S) command protocol, and Figure 5.3 shows an  
example protocol for command abort.  
C141-E116-01EN  
5 - 77  
Command  
b c  
Parameter write  
Status read  
e
Status read  
~
• • • •  
• • • •  
a
e
BSY  
d
f
d
DRDY  
DRQ  
INTRQ  
Data transfer  
Expanded  
Command  
Min. 30 µs (*1)  
• • •  
DRQ  
INTRQ  
Data Reg.  
Selection  
• • • •  
• • • •  
Data  
IOR-  
• • • •  
Word  
0
1
2
255  
IOCS16-  
*1 When the IDD receives a command that hits the cache data during read-ahead, and  
transfers data from the buffer without reading from the disk medium.  
Figure 5.2 Read Sector(s) command protocol  
Even if the error status exists, the drive makes a preparation (setting the DRQ bit) of data transfer.  
It is up to the host whether data is transferred. In other words, the host should receive the data of  
the sector (512 bytes of uninsured dummy data) or release the DRQ by resetting.  
5 - 78  
C141-E116-01EN  
Note:  
For transfer of a sector of data, the host needs to read Status register (X'1F7') in order to clear  
INTRQ (interrupt) signal. The Status register should be read within a period from the DRQ  
setting by the device to 5 µs after the completion of the sector data transfer. Note that the host  
does not need to read the Status register for the reading of a single sector or the last sector in  
multiple-sector reading. If the timing to read the Status register does not meet above  
condition, normal data transfer operation is not guaranteed.  
When the host new command even if the device requests the data transfer (setting in DRQ bit),  
the correct device operation is not guaranteed.  
Command  
Status read  
Parameter write  
~
BSY  
DRDY  
DRQ  
INTRQ  
Data transfer  
* Transfers dummy data  
* The host should receive 512-byte dummy data or release the DRQ set state by resetting.  
Figure 5.3 Protocol for command abort  
5.4.2  
Data transferring commands from host to device  
The execution of the following commands involves Data transfer from the host to the drive.  
FORMAT TRACK  
WRITE SECTOR(S)  
WRITE LONG  
WRITE BUFFER  
WRITE VERIFY  
SECURITY DISABLE PASSWORD  
SECURITY ERASE UNIT  
SECURITY SET PASSWORD  
SECURITY UNLOCK  
The execution of these commands includes the transfer one or more sectors of data from the host to  
the device. In the WRITE LONG command, 516 bytes are transferred. Following shows the  
protocol outline.  
C141-E116-01EN  
5 - 79  
a) The host writes any required parameters to the Features, Sector Count, Sector Number,  
Cylinder, and Device/Head registers.  
b) The host writes a command code in the Command register. The drive sets the BSY bit of the  
Status register.  
c) When the device is ready to receive the data of the first sector, the device sets DRQ bit and  
clears BSY bit.  
d) The host writes one sector of data through the Data register.  
e) The device clears the DRQ bit and sets the BSY bit.  
f) When the drive completes transferring the data of the sector, the device clears BSY bit and  
asserts INTRQ signal. If transfer of another sector is requested, the drive sets the DRQ bit.  
g) After detecting the INTRQ signal assertion, the host reads the Status register.  
h) The device resets INTRQ (the interrupt signal).  
i) If transfer of another sector is requested, steps d) and after are repeated.  
Figure 5.4 shows an example of WRITE SECTOR(S) command protocol, and Figure 5.3 shows an  
example protocol for command abort.  
Command  
Parameter write  
Status read  
g
Status read  
g
~
b
• • • •  
• • • •  
a
f
BSY  
DRDY  
DRQ  
c
e
h
INTRQ  
Data transfer  
d
d
Expanded  
Command  
DRQ  
Max. 1  
s
µ
• • •  
Data Reg. Selection  
• • • •  
• • • •  
Data  
IOR-  
• • • •  
Word  
0
1
2
255  
IOCS16  
Figure 5.4 WRITE SECTOR(S) command protocol  
5 - 80  
C141-E116-01EN  
Note:  
For transfer of a sector of data, the host needs to read Status register (X'1F7') in order to clear  
INTRQ (interrupt) signal. The Status register should be read within a period from the DRQ  
setting by the device to 5 µs after the completion of the sector data transfer. Note that the host  
does not need to read the Status register for the first and the last sector to be transferred. If the  
timing to read the Status register does not meet above condition, normal data transfer operation  
is not assured guaranteed.  
When the host issues the command even if the drive requests the data transfer (DRQ bit is set),  
or when the host executes resetting, the device correct operation is not guaranteed.  
5.4.3  
Commands without data transfer  
Execution of the following commands does not involve data transfer between the host and the  
device.  
RECALIBRATE  
SEEK  
READY VERIFY SECTOR(S)  
EXECUTE DEVICE DIAGNOSTIC  
INITIALIZE DEVICE PARAMETERS  
SET FEATURES  
SET MULTIPLE MODE  
IDLE  
IDLE IMMEDIATE  
STANDBY  
STANDBY IMMEDIATE  
CHECK POWER MODE  
FLUSH CACHE  
SECURITY ERASE PREPARE  
SECURITY FREEZE LOCK  
SMART: except for SMART Read Attribute values and SMART Read Attribute Thresholds  
SET MAX ADDRESS  
READ NATIVE MAX ADDRESS  
Figure 5.5 shows the protocol for the command execution without data transfer.  
Command  
Parameter write  
Status read  
~
BSY  
DRDY  
INTRQ  
Figure 5.5 Protocol for the command execution without data transfer  
C141-E116-01EN  
5 - 81  
5.4.4  
Other commands  
READ MULTIPLE  
SLEEP  
WRITE MULTIPLE  
See the description of each command.  
5.4.5  
DMA data transfer commands  
READ DMA  
WRITE DMA  
Starting the DMA transfer command is the same as the READ SECTOR(S) or WRITE  
SECTOR(S) command except the point that the host initializes the DMA channel preceding the  
command issuance.  
The interrupt processing for the DMA transfer differs the following point.  
The interrupt processing for the DMA transfer differs the following point.  
a) The host writes any parameters to the Features, Sector Count, Sector Number, Cylinder,  
and Device/Head register.  
b) The host initializes the DMA channel  
c) The host writes a command code in the Command register.  
d) The device sets the BSY bit of the Status register.  
e) The device asserts the DMARQ signal after completing the preparation of data transfer.  
The device asserts either the BSY bit during DMA data transfer.  
f) When the command execution is completed, the device clears both BSY and DRQ bits  
and asserts the INTRQ signal.  
g) The host reads the Status register.  
h) The host resets the DMA channel.  
5 - 82  
C141-E116-01EN  
Command  
c, d  
Status read  
Parameter write  
~
a
BSY  
• •  
DRDY  
INTRQ  
f
g
e
• •  
DRQ  
• •  
Data transfer  
Expanded  
[Multiword DMA transfer]  
DRQ  
• • • •  
• • • •  
DMARQ  
DMACK-  
• • • •  
• • • •  
IOR- or  
IOW-  
Word  
0
1
n-1  
n
Figure 5.6 Normal DMA data transfer  
C141-E116-01EN  
5 - 83  
5.5  
Ultra DMA feature set  
Overview  
5.5.1  
Ultra DMA is a data transfer protocol used with the READ DMA and WRITE DMA commands.  
When this protocol is enabled it shall be used instead of the Multiword DMA protocol when these  
commands are issued by the host. This protocol applies to the Ultra DMA data burst only. When  
this protocol is used there are no changes to other elements of the ATA protocol (e.g.: Command  
Block Register access).  
Several signal lines are redefined to provide new functions during an Ultra DMA burst. These  
lines assume these definitions when 1) an Ultra DMA Mode is selected, and 2) a host issues a  
READ DMA or a WRITE DMA, command requiring data transfer, and 3) the host asserts  
DMACK-. These signal lines revert back to the definitions used for non-Ultra DMA transfers  
upon the negation of DMACK- by the host at the termination of an Ultra DMA burst. All of the  
control signals are unidirectional. DMARQ and DMACK- retain their standard definitions.  
With the Ultra DMA protocol, the control signal (STROBE) that latches data from DD (15:0) is  
generated by the same agent (either host or device) that drives the data onto the bus. Ownership of  
DD (15:0) and this data strobe signal are given either to the device during an Ultra DMA data in  
burst or to the host for an Ultra DMA data out burst.  
During an Ultra DMA burst a sender shall always drive data onto the bus, and after a sufficient  
time to allow for propagation delay, cable settling, and setup time, the sender shall generate a  
STROBE edge to latch the data. Both edges of STROBE are used for data transfers so that the  
frequency of STROBE is limited to the same frequency as the data. The highest fundamental  
frequency on the cable shall be 16.67 million transitions per second or 8.33 MHz (the same as the  
maximum frequency for PIO Mode 4 and DMA Mode 2).  
Words in the IDENTIFY DEVICE data indicate support of the Ultra DMA feature and the Ultra  
DMA Modes the device is capable of supporting. The Set transfer mode subcommand in the SET  
FEATURES command shall be used by a host to select the Ultra DMA Mode at which the system  
operates. The Ultra DMA Mode selected by a host shall be less than or equal to the fastest mode  
of which the device is capable. Only the Ultra DMA Mode shall be selected at any given time. All  
timing requirements for a selected Ultra DMA Mode shall be satisfied. Devices supporting Ultra  
DMA Mode 2 shall also support Ultra DMA Modes 0 and 1. Devices supporting Ultra DMA  
Mode 1 shall also support Ultra DMA Mode 0.  
An Ultra DMA capable device shall retain its previously selected Ultra DMA Mode after  
executing a Software reset sequence. An Ultra DMA capable device shall clear any previously  
selected Ultra DMA Mode and revert to its default non-Ultra DMA Modes after executing a Power  
on or hardware reset.  
Both the host and device perform a CRC function during an Ultra DMA burst. At the end of an  
Ultra DMA burst the host sends the its CRC data to the device. The device compares its CRC data  
to the data sent from the host. If the two values do not match the device reports an error in the  
error register at the end of the command. If an error occurs during one or more Ultra DMA bursts  
for any one command, at the end of the command, the device shall report the first error that  
occurred.  
5 - 84  
C141-E116-01EN  
5.5.2  
Phases of operation  
An Ultra DMA data transfer is accomplished through a series of Ultra DMA data in or data out  
bursts. Each Ultra DMA burst has three mandatory phases of operation: the initiation phase, the  
data transfer phase, and the Ultra DMA burst termination phase. In addition, an Ultra DMA burst  
may be paused during the data transfer phase (see 5.5.3 and 5.5.4 for the detailed protocol  
descriptions for each of these phases, 5.6 defines the specific timing requirements). In the  
following rules DMARDY- is used in cases that could apply to either DDMARDY- or  
HDMARDY-, and STROBE is used in cases that could apply to either DSTROBE or HSTROBE.  
The following are general Ultra DMA rules.  
a) An Ultra DMA burst is defined as the period from an assertion of DMACK- by the host to the  
subsequent negation of DMACK-.  
b) A recipient shall be prepared to receive at least two data words whenever it enters or resumes  
an Ultra DMA burst.  
5.5.3  
Ultra DMA data in commands  
5.5.3.1 Initiating an Ultra DMA data in burst  
The following steps shall occur in the order they are listed unless otherwise specifically allowed  
(see 5.6.3.1 and 5.6.3.2 for specific timing requirements):  
1) The host shall keep DMACK- in the negated state before an Ultra DMA burst is initiated.  
2) The device shall assert DMARQ to initiate an Ultra DMA burst. After assertion of DMARQ  
the device shall not negate DMARQ until after the first negation of DSTROBE.  
3) Steps (3), (4) and (5) may occur in any order or at the same time. The host shall assert STOP.  
4) The host shall negate HDMARDY-.  
5) The host shall negate CS0-, CS1-, DA2, DA1, and DA0. The host shall keep CS0-, CS1-,  
DA2, DA1, and DA0 negated until after negating DMACK- at the end of the burst.  
6) Steps (3), (4) and (5) shall have occurred at least tACK before the host asserts DMACK-. The  
host shall keep DMACK- asserted until the end of an Ultra DMA burst.  
7) The host shall release DD (15:0) within tAZ after asserting DMACK-.  
8) The device may assert DSTROBE tZIORDY after the host has asserted DMACK-. Once the  
device has driven DSTROBE the device shall not release DSTROBE until after the host has  
negated DMACK- at the end of an Ultra DMA burst.  
9) The host shall negate STOP and assert HDMARDY- within tENV after asserting DMACK-.  
After negating STOP and asserting HDMARDY-, the host shall not change the state of either  
signal until after receiving the first transition of DSTROBE from the device (i.e., after the first  
data word has been received).  
10) The device shall drive DD (15:0) no sooner than tZAD after the host has asserted DMACK-,  
negated STOP, and asserted HDMARDY-.  
C141-E116-01EN  
5 - 85  
11) The device shall drive the first word of the data transfer onto DD (15:0). This step may occur  
when the device first drives DD (15:0) in step (10).  
12) To transfer the first word of data the device shall negate DSTROBE within tFS after the host  
has negated STOP and asserted HDMARDY-. The device shall negate DSTROBE no sooner  
than tDVS after driving the first word of data onto DD (15:0).  
5.5.3.2 The data in transfer  
The following steps shall occur in the order they are listed unless otherwise specifically allowed  
(see 5.6.3.3 and 5.6.3.2 for specific timing requirements):  
1) The device shall drive a data word onto DD (15:0).  
2) The device shall generate a DSTROBE edge to latch the new word no sooner than tDVS after  
changing the state of DD (15:0). The device shall generate a DSTROBE edge no more  
frequently than tCYC for the selected Ultra DMA Mode. The device shall not generate two  
rising or two falling DSTROBE edges more frequently than 2tCYC for the selected Ultra DMA  
mode.  
3) The device shall not change the state of DD (15:0) until at least tDVH after generating a  
DSTROBE edge to latch the data.  
4) The device shall repeat steps (1), (2) and (3) until the data transfer is complete or an Ultra  
DMA burst is paused, whichever occurs first.  
5.5.3.3 Pausing an Ultra DMA data in burst  
The following steps shall occur in the order they are listed unless otherwise specifically allowed  
(see 5.6.3.4 and 5.6.3.2 for specific timing requirements).  
a) Device pausing an Ultra DMA data in burst  
1) The device shall not pause an Ultra DMA burst until at least one data word of an Ultra  
DMA burst has been transferred.  
2) The device shall pause an Ultra DMA burst by not generating DSTROBE edges.  
NOTE - The host shall not immediately assert STOP to initiate Ultra DMA burst  
termination when the device stops generating STROBE edges. If the device does not  
negate DMARQ, in order to initiate ULTRA DMA burst termination, the host shall  
negate HDMARDY- and wait tRP before asserting STOP.  
3) The device shall resume an Ultra DMA burst by generating a DSTROBE edge.  
b) Host pausing an Ultra DMA data in burst  
1) The host shall not pause an Ultra DMA burst until at least one data word of an Ultra  
DMA burst has been transferred.  
2) The host shall pause an Ultra DMA burst by negating HDMARDY-.  
5 - 86  
C141-E116-01EN  
3) The device shall stop generating DSTROBE edges within tRFS of the host negating  
HDMARDY-.  
4) If the host negates HDMARDY- within tSR after the device has generated a DSTROBE  
edge, then the host shall be prepared to receive zero or one additional data words. If the  
host negates HDMARDY- greater than tSR after the device has generated a DSTROBE  
edge, then the host shall be prepared to receive zero, one or two additional data words.  
The additional data words are a result of cable round trip delay and tRFS timing for the  
device.  
5) The host shall resume an Ultra DMA burst by asserting HDMARDY-.  
5.5.3.4 Terminating an Ultra DMA data in burst  
a) Device terminating an Ultra DMA data in burst  
The following steps shall occur in the order they are listed unless otherwise specifically  
allowed (see 5.6.3.5 and 5.6.3.2 for specific timing requirements):  
1) The device shall initiate termination of an Ultra DMA burst by not generating DSTROBE  
edges.  
2) The device shall negate DMARQ no sooner than tSS after generating the last DSTROBE  
edge. The device shall not assert DMARQ again until after the Ultra DMA burst is  
terminated.  
3) The device shall release DD (15:0) no later than tAZ after negating DMARQ.  
4) The host shall assert STOP within tLI after the device has negated DMARQ. The host  
shall not negate STOP again until after the Ultra DMA burst is terminated.  
5) The host shall negate HDMARDY- within tLI after the device has negated DMARQ. The  
host shall continue to negate HDMARDY- until the Ultra DMA burst is terminated.  
Steps (4) and (5) may occur at the same time.  
6) The host shall drive DD (15:0) no sooner than tZAH after the device has negated DMARQ.  
For this step, the host may first drive DD (15:0) with the result of its CRC calculation (see  
5.5.5):  
7) If DSTROBE is negated, the device shall assert DSTROBE within tLI after the host has  
asserted STOP. No data shall be transferred during this assertion. The host shall ignore  
this transition on DSTROBE. DSTROBE shall remain asserted until the Ultra DMA  
burst is terminated.  
8) If the host has not placed the result of its CRC calculation on DD (15:0) since first driving  
DD (15:0) during (6), the host shall place the result of its CRC calculation on DD (15:0)  
(see 5.5.5).  
9) The host shall negate DMACK- no sooner than tMLI after the device has asserted  
DSTROBE and negated DMARQ and the host has asserted STOP and negated  
HDMARDY-, and no sooner than tDVS after the host places the result of its CRC  
calculation on DD (15:0).  
C141-E116-01EN  
5 - 87  
10) The device shall latch the host's CRC data from DD (15:0) on the negating edge of  
DMACK-.  
11) The device shall compare the CRC data received from the host with the results of its own  
CRC calculation. If a miscompare error occurs during one or more Ultra DMA bursts for  
any one command, at the end of the command the device shall report the first error that  
occurred (see 5.5.5).  
12) The device shall release DSTROBE within tIORDYZ after the host negates DMACK-.  
13) The host shall not negate STOP no assert HDMARDY- until at least tACK after negating  
DMACK-.  
14) The host shall not assert DIOR-, CS0-, CS1-, DA2, DA1, or DA0 until at least tACK after  
negating DMACK.  
b) Host terminating an Ultra DMA data in burst  
The following steps shall occur in the order they are listed unless otherwise specifically  
allowed (see 5.6.3.6 and 5.6.3.2 for specific timing requirements):  
1) The host shall not initiate Ultra DMA burst termination until at least one data word of an  
Ultra DMA burst has been transferred.  
2) The host shall initiate Ultra DMA burst termination by negating HDMARDY-. The host  
shall continue to negate HDMARDY- until the Ultra DMA burst is terminated.  
3) The device shall stop generating DSTROBE edges within tRFS of the host negating  
HDMARDY-.  
4) If the host negates HDMARDY- within tSR after the device has generated a DSTROBE  
edge, then the host shall be prepared to receive zero or one additional data words. If the  
host negates HDMARDY- greater than tSR after the device has generated a DSTROBE  
edge, then the host shall be prepared to receive zero, one or two additional data words.  
The additional data words are a result of cable round trip delay and tRFS timing for the  
device.  
5) The host shall assert STOP no sooner than tRP after negating HDMARDY-. The host  
shall not negate STOP again until after the Ultra DMA burst is terminated.  
6) The device shall negate DMARQ within tLI after the host has asserted STOP. The device  
shall not assert DMARQ again until after the Ultra DMA burst is terminated.  
7) If DSTROBE is negated, the device shall assert DSTROBE within tLI after the host has  
asserted STOP. No data shall be transferred during this assertion. The host shall ignore  
this transition on DSTROBE. DSTROBE shall remain asserted until the Ultra DMA  
burst is terminated.  
8) The device shall release DD (15:0) no later than tAZ after negating DMARQ.  
9) The host shall drive DD (15:0) no sooner than tZAH after the device has negated DMARQ.  
For this step, the host may first drive DD (15:0) with the result of its CRC calculation (see  
5.5.5).  
5 - 88  
C141-E116-01EN  
10) If the host has not placed the result of its CRC calculation on DD (15:0) since first driving  
DD (15:0) during (9), the host shall place the result of its CRC calculation on DD (15:0)  
(see 5.5.5).  
11) The host shall negate DMACK- no sooner than tMLI after the device has asserted  
DSTROBE and negated DMARQ and the host has asserted STOP and negated  
HDMARDY-, and no sooner than tDVS after the host places the result of its CRC  
calculation on DD (15:0).  
12) The device shall latch the host's CRC data from DD (15:0) on the negating edge of  
DMACK-.  
13) The device shall compare the CRC data received from the host with the results of its own  
CRC calculation. If a miscompare error occurs during one or more Ultra DMA burst for  
any one command, at the end of the command, the device shall report the first error that  
occurred (see 5.5.5).  
14) The device shall release DSTROBE within tIORDYZ after the host negates DMACK-.  
15) The host shall neither negate STOP nor assert HDMARDY- until at least tACK after the  
host has negated DMACK-.  
16) The host shall not assert DIOR-, CS0-, CS1-, DA2, DA1, or DA0 until at least tACK after  
negating DMACK.  
5.5.4  
Ultra DMA data out commands  
5.5.4.1 Initiating an Ultra DMA data out burst  
The following steps shall occur in the order they are listed unless otherwise specifically allowed  
(see 5.6.3.7 and 5.6.3.2 for specific timing requirements):  
1) The host shall keep DMACK- in the negated state before an Ultra DMA burst is initiated.  
2) The device shall assert DMARQ to initiate an Ultra DMA burst.  
3) Steps (3), (4), and (5) may occur in any order or at the same time. The host shall assert STOP.  
4) The host shall assert HSTROBE.  
5) The host shall negate CS0-, CS1-, DA2, DA1, and DA0. The host shall keep CS0-, CS1-,  
DA2, DA1, and DA0 negated until after negating DMACK- at the end of the burst.  
6) Steps (3), (4), and (5) shall have occurred at least tACK before the host asserts DMACK-. The  
host shall keep DMACK- asserted until the end of an Ultra DMA burst.  
7) The device may negate DDMARDY- tZIORDY after the host has asserted DMACK-. Once the  
device has negated DDMARDY-, the device shall not release DDMARDY- until after the host  
has negated DMACK- at the end of an Ultra DMA burst.  
8) The host shall negate STOP within tENV after asserting DMACK-. The host shall not assert  
STOP until after the first negation of HSTROBE.  
C141-E116-01EN  
5 - 89  
9) The device shall assert DDMARDY- within tLI after the host has negated STOP. After  
asserting DMARQ and DDMARDY- the device shall not negate either signal until after the  
first negation of HSTROBE by the host.  
10) The host shall drive the first word of the data transfer onto DD (15:0). This step may occur  
any time during Ultra DMA burst initiation.  
11) To transfer the first word of data: the host shall negate HSTROBE no sooner than tLI after the  
device has asserted DDMARDY-. The host shall negate HSTROBE no sooner than tDVS after  
the driving the first word of data onto DD (15:0).  
5.5.4.2 The data out transfer  
The following steps shall occur in the order they are listed unless otherwise specifically allowed  
(see 5.6.3.8 and 5.6.3.2 for specific timing requirements):  
1) The host shall drive a data word onto DD (15:0).  
2) The host shall generate an HSTROBE edge to latch the new word no sooner than tDVS after  
changing the state of DD (15:0). The host shall generate an HSTROBE edge no more  
frequently than tCYC for the selected Ultra DMA Mode. The host shall not generate two rising  
or falling HSTROBE edges more frequently than 2 tCYC for the selected Ultra DMA mode.  
3) The host shall not change the state of DD (15:0) until at least tDVH after generating an  
HSTROBE edge to latch the data.  
4) The host shall repeat steps (1), (2) and (3) until the data transfer is complete or an Ultra DMA  
burst is paused, whichever occurs first.  
5.5.4.3 Pausing an Ultra DMA data out burst  
The following steps shall occur in the order they are listed unless otherwise specifically allowed  
(see 5.6.3.9 and 5.6.3.2 for specific timing requirements).  
a) Host pausing an Ultra DMA data out burst  
1) The host shall not pause an Ultra DMA burst until at least one data word of an Ultra  
DMA burst has been transferred.  
2) The host shall pause an Ultra DMA burst by not generating an HSTROBE edge.  
Note: The device shall not immediately negate DMARQ to initiate Ultra DMA burst  
termination when the host stops generating HSTROBE edges. If the host does not assert  
STOP, in order to initiate Ultra DMA burst termination, the device shall negate  
DDMARDY- and wait tRP before negating DMARQ.  
3) The host shall resume an Ultra DMA burst by generating an HSTROBE edge.  
5 - 90  
C141-E116-01EN  
b) Device pausing an Ultra DMA data out burst  
1) The device shall not pause an Ultra DMA burst until at least one data word of an Ultra  
DMA burst has been transferred.  
2) The device shall pause an Ultra DMA burst by negating DDMARDY-.  
3) The host shall stop generating HSTROBE edges within tRFS of the device negating  
DDMARDY-.  
4) If the device negates DDMARDY- within tSR after the host has generated an HSTROBE  
edge, then the device shall be prepared to receive zero or one additional data words. If  
the device negates DDMARDY- greater than tSR after the host has generated an  
HSTROBE edge, then the device shall be prepared to receive zero, one or two additional  
data words. The additional data words are a result of cable round trip delay and tRFS  
timing for the host.  
5) The device shall resume an Ultra DMA burst by asserting DDMARDY-.  
5.5.4.4 Terminating an Ultra DMA data out burst  
a) Host terminating an Ultra DMA data out burst  
The following stops shall occur in the order they are listed unless otherwise specifically  
allowed (see 5.6.3.10 and 5.6.3.2 for specific timing requirements):  
1) The host shall initiate termination of an Ultra DMA burst by not generating HSTROBE  
edges.  
2) The host shall assert STOP no sooner than tSS after it last generated an HSTROBE edge.  
The host shall not negate STOP again until after the Ultra DMA burst is terminated.  
3) The device shall negate DMARQ within tLI after the host asserts STOP. The device shall  
not assert DMARQ again until after the Ultra DMA burst is terminated.  
4) The device shall negate DDMARDY- with tLI after the host has negated STOP. The  
device shall not assert DDMARDY- again until after the Ultra DMA burst termination is  
complete.  
5) If HSTROBE is negated, the host shall assert HSTROBE with tLI after the device has  
negated DMARQ. No data shall be transferred during this assertion. The device shall  
ignore this transition on HSTROBE. HSTROBE shall remain asserted until the Ultra  
DMA burst is terminated.  
6) The host shall place the result of its CRC calculation on DD (15:0) (see 5.5.5)  
7) The host shall negate DMACK- no sooner than tMLI after the host has asserted HSTROBE  
and STOP and the device has negated DMARQ and DDMARDY-, and no sooner than  
tDVS after placing the result of its CRC calculation on DD (15:0).  
8) The device shall latch the host's CRC data from DD (15:0) on the negating edge of  
DMACK-.  
C141-E116-01EN  
5 - 91  
9) The device shall compare the CRC data received from the host with the results of its own  
CRC calculation. If a miscompare error occurs during one or more Ultra DMA bursts for  
any one command, at the end of the command, the device shall report the first error that  
occurred (see 5.5.5).  
10) The device shall release DDMARDY- within tIORDYZ after the host has negated DMACK-.  
11) The host shall neither negate STOP nor negate HSTROBE until at least tACK after  
negating DMACK-.  
12) The host shall not assert DIOW-, CS0-, CS1-, DA2, DA1, or DA0 until at least tACK after  
negating DMACK.  
b) Device terminating an Ultra DMA data out burst  
The following steps shall occur in the order they are listed unless otherwise specifically  
allowed (see 5.6.3.11 and 5.6.3.2 for specific timing requirements):  
1) The device shall not initiate Ultra DMA burst termination until at least one data word of  
an Ultra DMA burst has been transferred.  
2) The device shall initiate Ultra DMA burst termination by negating DDMARDY-.  
3) The host shall stop generating an HSTROBE edges within tRFS of the device negating  
DDMARDY-.  
4) If the device negates DDMARDY- within tSR after the host has generated an HSTROBE  
edge, then the device shall be prepared to receive zero or one additional data words. If  
the device negates DDMARDY- greater than tSR after the host has generated an  
HSTROBE edge, then the device shall be prepared to receive zero, one or two additional  
data words. The additional data words are a result of cable round trip delay and tRFS  
timing for the host.  
5) The device shall negate DMARQ no sooner than tRP after negating DDMARDY-. The  
device shall not assert DMARQ again until after the Ultra DMA burst is terminated.  
6) The host shall assert STOP with tLI after the device has negated DMARQ. The host shall  
not negate STOP again until after the Ultra DMA burst is terminated.  
7) If HSTROBE is negated, the host shall assert HSTROBE with tLI after the device has  
negated DMARQ. No data shall be transferred during this assertion. The device shall  
ignore this transition of HSTROBE. HSTROBE shall remain asserted until the Ultra  
DMA burst is terminated.  
8) The host shall place the result of its CRC calculation on DD (15:0) (see 5.5.5).  
9) The host shall negate DMACK- no sooner than tMLI after the host has asserted HSTROBE  
and STOP and the device has negated DMARQ and DDMARDY-, and no sooner than  
tDVS after placing the result of its CRC calculation on DD (15:0).  
10) The device shall latch the host's CRC data from DD (15:0) on the negating edge of  
DMACK-.  
5 - 92  
C141-E116-01EN  
11) The device shall compare the CRC data received from the host with the results of its own  
CRC calculation. If a miscompare error occurs during one or more Ultra DMA bursts for  
any one command, at the end of the command, the device shall report the first error that  
occurred (see 5.5.5).  
12) The device shall release DDMARDY- within tIORDYZ after the host has negated DMACK-.  
13) The host shall neither negate STOP nor HSTROBE until at least tACK after negating DMACK-.  
14) The host shall not assert DIOW-, CS0-, CS1-, DA2, DA1, or DA0 until at least tACK after  
negating DMACK.  
5.5.5  
Ultra DMA CRC rules  
The following is a list of rules for calculating CRC, determining if a CRC error has occurred  
during an Ultra DMA burst, and reporting any error that occurs at the end of a command.  
a) Both the host and the device shall have a 16-bit CRC calculation function.  
b) Both the host and the device shall calculate a CRC value for each Ultra DMA burst.  
c) The CRC function in the host and the device shall be initialized with a seed of 4ABAh at the  
beginning of an Ultra DMA burst before any data is transferred.  
d) For each STROBE transition used for data transfer, both the host and the device shall calculate  
a new CRC value by applying the CRC polynomial to the current value of their individual  
CRC functions and the word being transferred. CRC is not calculated for the return of  
STROBE to the asserted state after the Ultra DMA burst termination request has been  
acknowledged.  
e) At the end of any Ultra DMA burst the host shall send the results of its CRC calculation  
function to the device on DD (15:0) with the negation of DMACK-.  
f) The device shall then compare the CRC data from the host with the calculated value in its own  
CRC calculation function. If the two values do not match, the device shall save the error and  
report it at the end of the command. A subsequent Ultra DMA burst for the same command  
that does not have a CRC error shall not clear an error saved from a previous Ultra DMa burst  
in the same command. If a miscompare error occurs during one or more Ultra DMA bursts for  
any one command, at the end of the command, the device shall report the first error that  
occurred.  
g) For READ DMA or WRITE DMA commands: When a CRC error is detected, it shall be  
reported by setting both ICRC and ABRT (bit 7 and bit 2 in the Error register) to one. ICRC is  
defined as the "Interface CRC Error" bit. The host shall respond to this error by re-issuing the  
command.  
h) A host may send extra data words on the last Ultra DMA burst of a data out command. If a  
device determines that all data has been transferred for a command, the device shall terminate  
the burst. A device may have already received more data words than were required for the  
command. These extra words are used by both the host and the device to calculate the CRC,  
but, on an Ultra DMA data out burst, the extra words shall be discarded by the device.  
C141-E116-01EN  
5 - 93  
i) The CRC generator polynomial is : G (X) = X16 + X12 + X5 + 1.  
Note: Since no bit clock is available, the recommended approach for calculating CRC is to use  
a word clock derived from the bus strobe. The combinational logic shall then be equivalent to  
shifting sixteen bits serially through the generator polynomial where DD0 is shifted in first and  
DD15 is shifted in last.  
5.5.6  
Series termination required for Ultra DMA  
Series termination resistors are required at both the host and the device for operation in any of the  
Ultra DMA Modes. The following table describes recommended values for series termination at  
the host and the device.  
Table 5.18 Recommended series termination for Ultra DMA  
Signal  
DIOR-:HDMARDY-:HSTROBE  
DIOW-:STOP  
Host Termination  
22 ohm  
Device Termination  
82 ohm  
22 ohm  
82 ohm  
CS0-, CS1-  
33 ohm  
82 ohm  
DA0, DA1, DA2  
DMACK-  
33 ohm  
82 ohm  
22 ohm  
82 ohm  
DD15 through DD0  
DMARQ  
33 ohm  
33 ohm  
82 ohm  
22 ohm  
INTRQ  
82 ohm  
22 ohm  
IORDY:DDMARDY-:DSTROBE  
RESET-  
82 ohm  
22 ohm  
33 ohm  
82 ohm  
Note: Only those signals requiring termination are listed in this table. If a signal is  
not listed, series termination is not required for operation in an Ultra DMA Mode.  
For signals also requiring a pull-up or pull-down resistor at the host see Figure 5.7.  
Vcc  
Figure 5.7 Ultra DMA termination with pull-up or pull-down  
5 - 94  
C141-E116-01EN  
5.6  
Timing  
5.6.1  
PIO data transfer  
Figure 5.8 shows of the data transfer timing between the device and the host system.  
t0  
Addresses  
t1  
t9  
t2  
DIOR-/DIOW-  
t2i  
Write data  
DD0-DD15  
t3  
t4  
Read data  
DD0-DD15  
t5  
t6  
t10  
t11  
IORDY  
Symbol  
t12  
Timing parameter  
Min.  
120  
25  
70  
25  
20  
10  
5
Max.  
Unit  
ns  
ns  
ns  
ns  
ns  
ns  
ns  
ns  
ns  
ns  
ns  
ns  
t0  
t1  
Cycle time  
Data register selection setup time for DIOR-/DIOW-  
Pulse width of DIOR-/DIOW-  
t2  
t2i  
t3  
Recovery time of DIOR-/DIOW-  
Data setup time for DIOW-  
t4  
Data hold time for DIOW-  
t5  
Time from DIOR- assertion to read data available  
Data hold time for DIOR-  
50  
t6  
t9  
Data register selection hold time for DIOR-/DIOW-  
Time from DIOR-/DIOW- assertion to IORDY "low" level  
Time from validity of read data to IORDY "high" level  
Pulse width of IORDY  
10  
0
t10  
t11  
t12  
35  
1,250  
Figure 5.8 PIO data transfer timing  
C141-E116-01EN  
5 - 95  
5.6.2  
Multiword data transfer  
Figure 5.9 shows the multiword DMA data transfer timing between the device and the host system.  
t0  
DMARQ  
DMACK-  
tJ  
tC  
tI  
tK  
tD  
DIOR-/DIOW-  
Write data  
DD0-DD15  
tG  
tH  
Read data  
DD0-DD15  
tE  
tF  
Symbol  
Timing parameter  
Min.  
120  
70  
5
Max.  
35  
30  
Unit  
ns  
ns  
ns  
ns  
ns  
ns  
ns  
ns  
ns  
ns  
t0  
tC  
tD  
tE  
tF  
tG  
tH  
tI  
Cycle time  
Delay time from DMACK assertion to DMARQ negation  
Pulse width of DIOR-/DIOW-  
Data setup time for DIOR-  
Data hold time for DIOR-  
Data setup time for DIOW-  
20  
10  
0
Data hold time for DIOW-  
DMACK setup time for DIOR-/DIOW-  
DMACK hold time for DIOR-/DIOW-  
Continuous time of high level for DIOR-/DIOW-  
tJ  
5
tK  
25  
Figure 5.9 Multiword DMA data transfer timing (mode 2)  
5 - 96  
C141-E116-01EN  
5.6.3  
Ultra DMA data transfer  
Figures 5.10 through 5.19 define the timings associated with all phases of Ultra DMA bursts.  
Table 5.19 contains the values for the timings for each of the Ultra DMA Modes.  
5.6.3.1 Initiating an Ultra DMA data in burst  
5.6.3.2 contains the values for the timings for each of the Ultra DMA Modes.  
DMARQ  
(device)  
tUI  
DMACK-  
(host)  
tFS  
tACK  
tENV  
tZAD  
STOP  
(host)  
tACK  
tFS  
tENV  
HDMARDY-  
(host)  
tZAD  
tZFS  
tZIORDY  
DSTROBE  
(device)  
tDZFS  
tVDS  
tAZ  
tDVH  
DD (15:0)  
tACK  
DA0,DA1,DA2,  
CS0-,CS1-  
Note:  
The definitions for the STOP, HDMARDY-and DSTROBE signal lines are not in effect until  
DMARQ and DMACK- are asserted.  
Figure 5.10 Initiating an Ultra DMA data in burst  
C141-E116-01EN  
5 - 97  
5.6.3.2 Ultra DMA data burst timing requirements  
Table 5.19 Ultra DMA data burst timing requirements (1 of 2)  
MODE 0  
(in ns)  
MODE 1  
(in ns)  
MODE 2  
(in ns)  
MODE 3  
(in ns)  
MODE 4  
(in ns)  
MODE 5  
(in ns)  
NAME  
COMMENT  
MIN MAX MIN MAX MIN MAX MIN MAX MIN MAX MIN MAX  
t2CYCTYP 240  
160  
120  
90  
60  
40  
Typical sustained average two  
cycle time  
tCYC  
112  
230  
73  
54  
39  
25  
16.8  
Cycle time allowing for  
asymmetry and clock variations  
(from STROBE edge to STROBE  
edge)  
t2CYC  
153  
115  
86  
57  
38  
Two cycle time allowing for clock  
variations (from rising edge to  
next rising edge or from falling  
edge to next falling edge of  
STROBE)  
tDS  
15  
5
10  
5
7
5
7
5
5
5
4
Data setup time at recipient (from  
data valid until STROBE edge)  
(*2), (*5)  
tDH  
4.6  
Data hold time at recipient (from  
STROBE edge until data may  
become invalid) (*2), (*5)  
tDVS  
tDVH  
70  
48  
31  
20  
6.7  
6.2  
4.8  
4.8  
Data valid setup time at sender (from  
data valid until STROBE edge) (*3)  
6.2  
6.2  
6.2  
6.2  
Data valid hold time at sender  
(from STROBE edge until data  
may become invalid) (*3)  
tCS  
15  
10  
7
7
5
5
CRC word setup time at device  
(*2)  
tCH  
5
5
5
5
5
5
CRC word hold time device (*2)  
tCVS  
70  
48  
31  
20  
6.7  
10  
CRC word valid setup time at host  
(from CRC valid until DMACK-  
negation) (*3)  
tCVH  
6.2  
6.2  
6.2  
6.2  
6.2  
10  
CRC word valid hold time at  
sender (from DMACK-negation  
until CRC may become invalid)  
(*3)  
tZFS  
tDZFS  
tFS  
0
0
0
0
0
35  
25  
Time from STROBE output  
released-to-driving until the first  
transition of critical timing  
70  
48  
31  
20  
6.7  
Time from data output released-  
to-driving until the first transition  
of critical timing  
230  
150  
200  
150  
170  
150  
130  
100  
120  
100  
90 First STROBE time (for device to  
first negate DSTROBE from  
STOP during a data in burst)  
tLI  
0
20  
0
0
20  
0
0
20  
0
0
20  
0
0
20  
0
0
20  
0
75 Limited interlock time (*1)  
Interlock time with minimum (*1)  
Unlimited interlock time (*1)  
tMLI  
TUI  
5 - 98  
C141-E116-01EN  
Table 5.19 Ultra DMA data burst timing requirements (2 of 2)  
MODE 0  
(in ns)  
MODE 1  
(in ns)  
MODE 2  
(in ns)  
MODE 3  
(in ns)  
MODE 4  
(in ns)  
MODE 5  
(in ns)  
NAME  
tAZ  
COMMENT  
MIN MAX MIN MAX MIN MAX MIN MAX MIN MAX MIN MAX  
10  
10  
10  
10  
10  
10 Maximum time allowed for output  
drivers to release (from asserted or  
negated)  
tZAH  
tZAD  
tENV  
20  
0
20  
0
20  
0
20  
0
20  
0
20  
0
Minimum delay time required for  
output  
Drivers to assert or negate (from  
released)  
20  
70  
75  
20  
70  
70  
20  
70  
60  
20  
55  
60  
20  
55  
60  
20  
50 Envelope time (from DMACK- to  
STOP and HDMARDY- during  
data in burst initiation and from  
DMACK to STOP during data out  
burst initiation)  
tRFS  
50 Ready-to-final-STROBE time (no  
STROBE edges shall be sent this  
long after negation of DMARDY-)  
tRP  
160  
125  
100  
100  
100  
85  
Ready-to-pause time (that  
recipient shall wait to pause after  
negating DMARDY-)  
tIORDYZ  
tZIORDY  
tACK  
20  
20  
20  
20  
20  
20 Maximum time before releasing  
IORDY  
0
0
0
0
0
0
Minimum time before driving  
IORDY (*4)  
20  
20  
20  
20  
20  
20  
Setup and hold times for  
DMACK- (before assertion or  
negation)  
tSS  
50  
50  
50  
50  
50  
50  
Time from STROBE edge to  
negation of DMARQ or assertion  
of STOP (when sender terminates  
a burst)  
*1: Except for some instances of tMLI that apply to host signals only, the parameters tUI, tMLI and tLI indicate sender-to-recipient or  
recipient-to-sender interlocks, i.e., one agent (either sender or recipient) is waiting for the other agent to respond with a signal  
before proceeding. tUI is an unlimited interlock that has no maximum time value. tMLI is a limited time-out that has a defined  
minimum. tLI is a limited time-out that has a defined maximum.  
*2: 80-conductor cabling shall be required in order to meet setup (tDS, tCS) and hold (tDH, tCH) times in modes greater than 2.  
*3: Timing for tDVS, tDVH, tCVS and tCVH shall be met for lumped capacitive loads of 15 and 40 pf at the connector where all signals  
(Data and STROBE) have the same capacitive load value. Due to reflections on the cable, the measurement of these timings is not  
valid in a normally functioning system.  
*4: For all modes the parameter tZIORDY may be greater than tENV due to the fact that the host has a pull up on IORDY- giving it a  
known state when not actively driven.  
*5: The parameters tDS, and tDH for mode 5 is defined for a recipient at the end of the cable only in a configuration with one device at  
the end of the cable.  
Note:  
All timing measurement switching points (low to high and high to low) shall be taken at 1.5V.  
C141-E116-01EN  
5 - 99  
Table 5.20 Ultra DMA sender and recipient timing requirements  
MODE 0  
(in ns)  
MODE 1  
(in ns)  
MODE 2  
(in ns)  
MODE 3  
(in ns)  
MODE 4  
(in ns)  
MODE 5  
(in ns)  
NAME  
tDSIC  
COMMENT  
MIN MAX MIN MAX MIN MAX MIN MAX MIN MAX MIN MAX  
14.7  
4.8  
72.9  
9
9.7  
4.8  
50.9  
9
6.8  
4.8  
33.9  
9
6.8  
4.8  
22.6  
9
4.8  
4.8  
9.5  
9
2.3  
2.8  
6
Recipient IC data setup time (from  
data valid until STROBE edge)  
(*1)  
tDHIC  
Recipient IC data hold time (from  
STROBE edge until data may  
become invalid) (*1)  
tDVSIC  
Sender IC data valid setup time  
(from data valid until STROBE  
edge) (*2)  
tDVHIC  
6
Sender IC data valid hold time  
(from STROBE edge until data  
may become invalid) (*2)  
*1: The correct data value shall be captured by the recipient given input data with a slew rate of 0.4 V/ns rising and falling and the  
input STROBE with a slew rate of 0.4 V/ns rising and falling at tDSIC and tDHIC timing (as measured through 1.5V).  
*2: The parameters tDVSIC and tDVHIC shall be met for lumped capacitive loads of 15 and 40 pf at the IC where all signals have the same  
capacitive load value. Noise that may couple onto the output signals from external sources in a normally functioning system has not  
been included in these values.  
Note:  
All timing measurement switching points (low to high and high to low) shall be taken at 1.5V.  
5 - 100  
C141-E116-01EN  
5.6.3.3 Sustained Ultra DMA data in burst  
5.6.3.2 contains the values for the timings for each of the Ultra DMA Modes.  
t2CYC  
tCYC  
tCYC  
t2CYC  
DSTROBE  
at device  
tDVH  
tDVHIC  
tDVS  
tDVSIC  
tDVH  
tDVHIC  
tDVS  
tDVSIC  
tDVH  
tDVHIC  
DD(15:0)  
at device  
DSTROBE  
at host  
tDH  
tDHIC  
tDH  
tDHIC  
tDH  
tDHIC  
tDS  
tDSIC  
tDS  
tDSIC  
DD(15:0)  
at host  
Note:  
DD (15:0) and DSTROBE signals are shown at both the host and the device to emphasize that  
cable setting time as well as cable propagation delay shall not allow the data signals to be  
considered stable at the host until some time after they are driven by the device.  
Figure 5.11 Sustained Ultra DMA data in burst  
C141-E116-01EN  
5 - 101  
5.6.3.4 Host pausing an Ultra DMA data in burst  
5.6.3.2 contains the values for the timings for each of the Ultra DMA Modes.  
DMARQ  
(device)  
DMACK-  
(host)  
tRP  
STOP  
(host)  
HDMARDY-  
(host)  
tRFS  
DSTROBE  
(device)  
DD(15:0)  
(device)  
Notes:  
1) The host may assert STOP to request termination of the Ultra DMA burst no sooner than  
tRP after HDMARDY- is negated.  
2) After negating HDMARDY-, the host may receive zero, one, two or three more data  
words from the device.  
Figure 5.12 Host pausing an Ultra DMA data in burst  
5 - 102  
C141-E116-01EN  
5.6.3.5 Device terminating an Ultra DMA data in burst  
5.6.3.2 contains the values for the timings for each of the Ultra DMA Modes.  
DMARQ  
(device)  
tMLI  
DMACK-  
(host)  
tACK  
tLI  
tLI  
STOP  
(host)  
tACK  
tLI  
HDMARDY-  
(host)  
tSS  
tIORDYZ  
DSTROBE  
(device)  
tZAH  
tAZ  
tCVS  
tCVH  
DD(15:0)  
CRC  
tACK  
DA0, DA1, DA2,  
CS0-, CS1-  
Note:  
The definitions for the STOP, HDMARDY- and DSTROBE signal lines are no longer in effect  
after DMARQ and DMACK- are negated.  
Figure 5.13 Device terminating an Ultra DMA data in burst  
C141-E116-01EN  
5 - 103  
5.6.3.6 Host terminating an Ultra DMA data in burst  
5.6.3.2 contains the values for the timings for each of the Ultra DMA Modes.  
DMARQ  
(device)  
tLI  
tMLI  
DMACK-  
(host)  
tZAH  
tACK  
tAZ  
tRP  
STOP  
(host)  
tACK  
HDMARDY-  
(host)  
tMLI  
tLI  
tRFS  
tIORDYZ  
DSTROBE  
(device)  
tCVS  
tCVH  
DD(15:0)  
CRC  
tACK  
DA0, DA1, DA2,  
CS0, CS1  
Note:  
The definitions for the STOP, HDMARDY- and DSTROBE signal lines are no longer in effect  
after DMARQ and DMACK- are negated.  
Figure 5.14 Host terminating an Ultra DMA data in burst  
5 - 104  
C141-E116-01EN  
5.6.3.7 Initiating an Ultra DMA data out burst  
5.6.3.2 contains the values for the timings for each of the Ultra DMA Modes.  
DMARQ  
(device)  
tUI  
DMACK-  
(host)  
tENV  
tACK  
STOP  
(host)  
tLI  
tUI  
tZIORDY  
DDMARDY-  
(device)  
tACK  
HSTROBE  
(host)  
tDZFS  
tDVS  
tDVH  
DD(15:0)  
(host)  
tACK  
DA0, DA1, DA2  
CS0-, CS1-  
Note:  
The definitions for the STOP, DDMARDY- and HSTROBE signal lines are not in effect until  
DMARQ and DMACK- are asserted.  
Figure 5.15 Initiating an Ultra DMA data out burst  
C141-E116-01EN  
5 - 105  
5.6.3.8 Sustained Ultra DMA data out burst  
5.6.3.2 contains the values for the timings for each of the Ultra DMA Modes.  
t2CYC  
tCYC  
tCYC  
t2CYC  
HSTROBE  
at host  
tDVH  
tDVHIC  
tDVH  
tDVHIC  
tDVH  
tDVHIC  
tDVS  
tDVSIC  
tDVS  
tDVSIC  
DD(15:0)  
at host  
HSTROBE  
at device  
tDH  
tDHIC  
tDS  
tDSIC  
tDH  
tDHIC  
tDS  
tDSIC  
tDH  
tDHIC  
DD(15:0)  
at device  
Note:  
DD (15:0) and HSTROBE signals are shown at both the device and the host to emphasize that  
cable setting time as well as cable propagation delay shall not allow the data signals to be  
considered stable at the device until some time after they are driven by the host.  
Figure 5.16 Sustained Ultra DMA data out burst  
5 - 106  
C141-E116-01EN  
5.6.3.9 Device pausing an Ultra DMA data out burst  
5.6.3.2 contains the values for the timings for each of the Ultra DMA Modes.  
tRP  
DMARQ  
(device)  
DMACK-  
(host)  
STOP  
(host)  
DDMARDY-  
(device)  
tRFS  
HSTROBE  
(host)  
DD(15:0)  
(host)  
Notes:  
1) The device may negate DMARQ to request termination of the Ultra DMA burst no sooner  
than tRP after DDMARDY- is negated.  
2) After negating DDMARDY-, the device may receive zero, one two or three more data  
words from the host.  
Figure 5.17 Device pausing an Ultra DMA data out burst  
C141-E116-01EN  
5 - 107  
5.6.3.10 Host terminating an Ultra DMA data out burst  
5.6.3.2 contains the values for the timings for each of the Ultra DMA Modes.  
tLI  
DMARQ  
(device)  
tMLI  
DMACK-  
(host)  
tLI  
tACK  
tSS  
STOP  
(host)  
tLI  
tIORDYZ  
DDMARDY-  
(device)  
tACK  
HSTROBE  
(host)  
tCVS  
tCVH  
DD(15:0)  
(host)  
CRC  
tACK  
DA0, DA1, DA2  
CS0-, CS1-  
Note:  
The definitions for the STOP, DDMARDY- and HSTROBE signal lines are no longer in effect  
after DMARQ and DMACK- are negated.  
Figure 5.18 Host terminating an Ultra DMA data out burst  
5 - 108  
C141-E116-01EN  
5.6.3.11 Device terminating an Ultra DMA data in burst  
5.6.3.2 contains the values for the timings for each of the Ultra DMA Modes.  
DMARQ  
(device)  
DMACK-  
(host)  
tACK  
tLI  
tMLI  
STOP  
(host)  
tRP  
tIORDYZ  
DDMARDY-  
(device)  
tRFS  
tLI  
tMLI  
tACK  
HSTROBE  
(host)  
tCVS  
tCVH  
DD(15:0)  
(host)  
CRC  
tACK  
DA0, DA1, DA2,  
CS0-, CS1-  
Note:  
The definitions for the STOP, DDMARDY- and HSTROBE signal lines are no longer in effect  
after DMARQ and DMACK- are negated.  
Figure 5.19 Device terminating an Ultra DMA data out burst  
C141-E116-01EN  
5 - 109  
5.6.4  
Power-on and reset  
Figure 5.20 shows power-on and reset (hardware and software reset) timing.  
(1)  
Only master device is present  
Clear Reset *1  
Power-on  
tM  
RESET-  
Software reset  
tN  
BSY  
DASP-  
tP  
*1: Reset means including Power-on-Reset, Hardware Reset (RESET-), and Software Reset.  
Master and slave devices are present (2-drives configuration)  
(2)  
Clear Reset  
[Master device]  
tN  
BSY  
DASP-  
[Slave device]  
BSY  
tQ  
tP  
PDIAG-  
DASP-  
tS  
tR  
Timing parameter  
Symbol  
tM  
tN  
Min.  
25  
Max.  
Unit  
µs  
ns  
ms  
s
Pulse width of RESET-  
Time from RESET- negation to BSY set  
400  
1
tP  
Time from RESET- negation to DASP- or DIAG- negation  
Self-diagnostics execution time  
tQ  
30  
tR  
Time from RESET- negation to DASP- assertion (slave device)  
Duration of DASP- assertion  
400  
31  
ms  
s
tS  
Figure 5.20 Power-on Reset Timing  
5 - 110  
C141-E116-01EN  
CHAPTER 6  
OPERATIONS  
6.1  
6.2  
6.3  
6.4  
6.5  
6.6  
Device Response to the Reset  
Address Translation  
Power Save  
Defect Management  
Read-Ahead Cache  
Write Cache  
6.1  
Device Response to the Reset  
This section describes how the PDIAG- and DASP- signals responds when the power of the IDD is  
turned on or the IDD receives a reset or diagnostic command.  
C141-E116-01EN  
6 - 1  
6.1.1  
Response to power-on  
After the master device (device 0) releases its own power-on reset state, the master device shall  
check a DASP- signal for up to 450 ms to confirm presence of a slave device (device 1). The  
master device recognizes presence of the slave device when it confirms assertion of the DASP-  
signal. Then, the master device checks a PDIAG- signal to see if the slave device has successfully  
completed the power-on diagnostics.  
If the master device cannot confirm assertion of the DASP- signal within 450 ms, the master  
device recognizes that no slave device is connected.  
After the slave device (device 1) releases its own power-on reset state, the slave device shall report  
its presence and the result of power-on diagnostics to the master device as described below:  
DASP- signal: Asserted within 400 ms, and negated after the first command is received from  
the host or within 31 seconds or after executing software reset, which ever  
comes first.  
PDIAG- signal: Negated within 1 ms and asserted within 30 seconds, then negated within 31  
seconds.  
Power on  
Master device  
Power On Reset-  
Status Reg.  
BSY bit  
Max. 31 sec.  
Checks DASP- for up to  
450 ms.  
If presence of a slave device is  
confirmed, PDIAG- is checked for  
up to 31 seconds.  
Slave device  
Power On Reset-  
BSY bit  
Max. 1 ms.  
PDIAG-  
DASP-  
Max. 30 sec.  
Max. 400 ms.  
Max. 31 sec.  
Figure 6.1 Response to power-on  
6 - 2  
C141-E116-01EN  
6.1.2  
Response to hardware reset  
Response to RESET- (hardware reset through the interface) is similar to the power-on reset.  
Upon receipt of hardware reset, the master device checks a DASP- signal for up to 450 ms to  
confirm presence of a slave device. The master device recognizes the presence of the slave device  
when it confirms assertion of the DASP- signal. Then the master device checks a PDIAG- signal  
to see if the slave device has successfully completed the self-diagnostics.  
If the master device cannot confirm assertion of the DASP- signal within 450 ms, the master  
device recognizes that no slave device is connected.  
After the slave device receives the hardware reset, the slave device shall report its presence and the  
result of the self-diagnostics to the master device as described below:  
DASP- signal: Asserted within 400 ms, and negated after the first command is received from  
the host or within 31 seconds or after executing software reset, which ever  
comes first.  
PDIAG- signal: Negated within 1 ms and asserted within 30 seconds, then negated within 31  
seconds  
Reset-  
Master device  
Status Reg.  
BSY bit  
Max. 31 sec.  
If presence of a slave device is  
Checks DASP- for up to  
450 ms.  
confirmed, PDIAG- is checked for  
up to 31 seconds.  
Slave device  
BSY bit  
Max. 1 ms.  
PDIAG-  
DASP-  
Max. 30 sec.  
Max. 400 ms.  
Max. 31 sec.  
.
Figure 6.2 Response to hardware reset  
C141-E116-01EN  
6 - 3  
6.1.3  
Response to software reset  
The master device does not check the DASP- signal for a software reset. If a slave device is  
present, the master device checks the PDIAG- signal for up to 31 seconds to see if the slave device  
has completed the self-diagnosis successfully.  
After the slave device receives the software reset, the slave device shall report its presence and the  
result of the self-diagnostics to the master device as described below:  
PDIAG- signal: negated within 1 ms and asserted within 30 seconds then negated within 31  
seconds.  
When the IDD is set to a slave device, the IDD asserts the DASP- signal when negating the  
PDIAG- signal, and negates the DASP- signal when asserting the PDIAG- signal.  
X'3F6' Reg.  
Master device  
X"0C"  
or X"04"  
X"00"  
Status Reg.  
BSY bit  
Max. 31 sec.  
If the slave device is preset, DASP- is checked for up to  
31 seconds.  
Slave device  
BSY bit  
Max. 1 ms.  
PDIAG-  
DASP-  
Max. 30 sec.  
Figure 6.3 Response to software reset  
6 - 4  
C141-E116-01EN  
6.1.4  
Response to diagnostic command  
When the master device receives an EXECUTE DEVICE DIAGNOSTIC command and the slave  
device is present, the master device checks the PDIAG- signal for up to 6 seconds to see if the  
slave device has completed the self-diagnosis successfully.  
The master device does not check the DASP- signal.  
After the slave device receives the EXECUTE DEVICE DIAGNOSTIC command, it shall report  
the result of the self-diagnostics to the master device as described below:  
PDIAG- signal: negated within 1 ms and asserted within 5 seconds then negated within 6  
seconds.  
When the IDD is set to a slave device, the IDD asserts the DASP- signal when negating the  
PDIAG- signal, and negates the DASP- signal when asserting the PDIAG- signal.  
X'1F7' Reg.  
Write  
Master device  
Status Reg.  
BSY bit  
Max. 6 sec.  
If the slave device is preset, DASP- signal is checked for up to  
6 seconds.  
Slave device  
BSY bit  
Max. 1 ms.  
PDIAG-  
DASP-  
Max. 5 sec.  
Figure 6.4 Response to diagnostic command  
C141-E116-01EN  
6 - 5  
6.2  
Address Translation  
When the IDD receives any command which involves access to the disk medium, the IDD always  
implements the address translation from the logical address (a host-specified address) to the  
physical address (logical to physical address translation).  
Following subsections explains the CHS translation mode.  
6.2.1  
Default parameters  
In the logical to physical address translation, the logical cylinder, head, and sector addresses are  
translated to the physical cylinder, head, and sector addresses based on the number of heads and  
the number of sectors per track which are specified with an INITIALIZE DEVICE  
PARAMETERS command. This is called as the current translation mode.  
If the number of heads and the number of sectors are not specified with an INITIALIZE DEVICE  
PARAMETERS command, the default values listed in Table 6.1 are used. This is called as the  
default translation mode. The parameters in Table 6.1 are called BIOS specification.  
Table 6.1 Default parameters  
MPG3204AH-E  
MPG3307AH-E  
MPG3409AH-E  
Number of cylinders  
Number of head  
16,383  
16  
Parameters  
(logical)  
Number of sectors/track  
63  
Formatted capacity (GB)  
20.49  
30.74  
40.99  
As long as the formatted capacity of the IDD does not exceed the value shown on Table 6.1, the  
host can freely specify the number of cylinders, heads, and sectors per track.  
Generally, the device recognizes the number of heads and sectors per track with the INITIALIZE  
DEVICE PARAMETER command. However, it cannot recognize the number of cylinders. In  
other words, there is no way for the device to recognize a host access area on logical cylinders.  
Thus the host should manage cylinder access to the device.  
The host can specify a logical address freely within an area where an address can be specified  
(within the specified number of cylinders, heads, and sectors per track) in the current translation  
mode.  
The host can read an addressable parameter information from the device by the IDENTIFY  
DEVICE command (Words 54 to 56).  
6 - 6  
C141-E116-01EN  
6.2.2  
Logical address  
(1)  
CHS mode  
Logical address assignment starts from physical cylinder (PC) 0, physical head (PH) 0, and  
physical sector (PS) 1 and is assigned by calculating the number of sectors per track which is  
specified by the INITIALIZE DEVICE PARAMETERS command. The head address is advanced  
at the subsequent sector from the last sector of the current physical head address. The first physical  
sector of the subsequent physical sector is the consecutive logical sector from the last sector of the  
current physical sector.  
Figure 6.5 shows an example (assuming there is no track skew).  
Physical sector  
1
2
3
62 63 64  
LS LS  
..  
..  
.. 126 127 ..189 190.. 734 735 736  
Physical cylinder 0  
LS LS  
1
LS LS  
63  
LS LS  
63  
LS LS LS  
41 42 43  
2
63  
1
1
1
Physical head 0  
..  
..  
..  
LH0  
LH1  
LH2  
LH11  
Physical sector  
1
LS  
44  
20 21  
LS LS  
63  
83 84  
LS LS  
735 736  
..146  
..............  
..  
..  
Physical cylinder 0  
Physical head 1  
LS LS  
63  
1
63  
1
1
..  
..  
..  
..............  
....  
LH11  
LH12  
LH13  
LH14  
ex: Zone 0  
Physical parameter  
- Physical sector: 1 to 736  
Specification of INITIALIZE DEVICE PARAMETERS command  
- Logical head: LH=0 to 15  
- Logical sector: LS=1 to 63  
Figure 6.5 Address translation (example in CHS mode)  
C141-E116-01EN  
6 - 7  
(2)  
LBA mode  
Logical address assignment in the LBA mode starts from physical cylinder 0, physical head 0, and  
physical sector 1. The logical address is advanced at the subsequent sector from the last sector of  
the current track. The first physical sector of the subsequent physical track is the consecutive  
logical sector from the last sector of the current physical track.  
Figure 6.6 shows an example of (assuming there is no track skew).  
Physical sector  
1
2
3
733 734 735 736  
......................  
......................  
Physical cylinder 0  
Physical head 0  
LBA LBA LBA  
LBA LBA LBA LBA  
732 733 734 735  
0
1
2
733 734 735 736  
1
2
3
......................  
......................  
Physical cylinder 0  
Physical head 1  
LBA LBA LBA LBA  
1464 1465 1466 1467  
LBA LBA LBA  
736 737 738  
ex: Zone 0  
Physical parameter  
- Physical sector: 1 to 736  
Figure 6.6 Address translation (example in LBA mode)  
6.3  
Power Save  
The host can change the power consumption state of the device by issuing a power command to the  
device.  
6.3.1  
Power save mode  
There are four types of power consumption state of the device including active mode where all  
circuits are active.  
In the power save mode, power supplying to the part of the circuit is turned off. There are three  
types of power save modes:  
Idle mode  
Standby mode  
Sleep mode  
6 - 8  
C141-E116-01EN  
(1)  
Active mode  
In this mode, all the electric circuit in the device are active or the device is under seek, read or  
write operation.  
A device enters the active mode under the following conditions:  
A command with Seek or Write or Read is issued.  
(2)  
Idle mode  
In this mode, circuits on the device is set to power save mode.  
The device enters the Idle mode under the following conditions:  
A IDLE or IDLE IMMEDIATE command is issued in the active or standby mode.  
When one of the following command is issued, the command is executed normally and the  
device is still stayed in the idle mode.  
- Reset (hardware or software)  
- IDLE command  
- IDLE IMMEDIATE command  
- A command without Seek or Write or Read is issued.  
(3)  
Standby mode  
In this mode, the VCM circuit is turned off and the spindle motor is stopped.  
The device can receive commands through the interface. However if a command with disk access  
is issued, response time to the command under the standby mode takes longer than the active or  
Idle mode because the access to the disk medium cannot be made immediately.  
The drive enters the standby mode under the following conditions:  
A STANDBY or STANDBY IMMEDIATE command is issued in the active or idle mode.  
When automatic power down sequence is enabled, the timer has elapsed.  
A reset is issued in the sleep mode.  
When one of following commands is issued, the command is executed normally and the device is  
still stayed in the standby mode.  
Reset (hardware or software)  
STANDBY command  
STANDBY IMMEDIATE command  
INITIALIZE DEVICE PARAMETERS command  
C141-E116-01EN  
6 - 9  
CHECK POWER MODE command  
(4)  
Sleep mode  
The power consumption of the drive is minimal in this mode. The drive enters only the standby  
mode from the sleep mode. The only method to return from the standby mode is to execute a  
software or hardware reset.  
The drive enters the sleep mode under the following condition:  
A SLEEP command is issued.  
Issued commands are invalid (ignored) in this mode.  
6.3.2  
Power commands  
The following commands are available as power commands.  
IDLE  
IDLE IMMEDIATE  
STANDBY  
STANDBY IMMEDIATE  
SLEEP  
CHECK POWER MODE  
6.4  
Defect Management  
Defective sectors of which the medium defect location is registered in the system space are  
replaced with spare sectors in the formatting at the factory shipment.  
All the user space area are formatted at shipment from the factory based on the default parameters  
listed in Table 6.1.  
6 - 10  
C141-E116-01EN  
6.4.1  
Spare area  
Following two types of spare area are provided in the user space.  
1) Spare sector for sector slip:  
used for alternating defective sectors at formatting in shipment (128 sectors/32 cylinders)  
2) Spare cylinder for alternative assignment:  
used by automatic alternative assignment. (4 cylinders/drive)  
6.4.2  
Alternating defective sectors  
The two alternating methods described below are available:  
(1)  
Sector slip processing  
A defective sector is not used and is skipped and a logical sector address is assigned to the  
subsequent normal sector (physically adjacent sector to the defective sector).  
When defective sector is present, the sector slip processing is performed in the formatting.  
Figure 6.7 shows an example where (physical) sector 5 is defective on head 0 in cylinder 0.  
Index  
Sector (physical)  
1
1
2
2
3
4
5
6
5
7
8
7
734  
733  
735  
734  
736  
735  
Cylinder 0  
Head 0  
Defective  
sector  
3
4
6
unused  
If an access request to sector 5 is specified, the device accesses physical sector 6 instead of sector 5.  
Figure 6.7 Sector slip processing  
C141-E116-01EN  
6 - 11  
(2)  
Alternate cylinder assignment  
A defective sector is assigned to the spare sector in the alternate cylinder.  
This processing is performed when a physical track contains three or more defective sectors, and  
when the automatic alternate processing is performed.  
Figure 6.8 shows an example where (physical) sector 5 is detective on head 0 in cylinder 0.  
Index  
Sector (physical)  
1
2
3
4
4
5
6
6
7
7
735  
735  
736  
736  
Cylinder 0  
Defective  
sector  
Head 0  
1
2
3
(unused)  
Alternate  
cylinder  
Already  
assigned  
Head 0  
Defective sector is assigned to unassigned sector.  
1 alternate cylinder is provided in outer side.  
When an access request to sector 5 is specified, the device accesses the alternated sector in the  
alternate cylinder instead of sector 5. When an access request to sectors next to sector 5 is  
specified, the device seeks to cylinder 0, head 0, and continues the processing.  
Figure 6.8 Alternate cylinder assignment  
(3)  
Automatic alternate assignment  
The device performs the automatic assignment at following case.  
1) When ECC correction performance is increased during read error retry, a read error is  
recovered.  
Before automatic alternate assignment, the device performs rewriting the corrected data to the  
erred sector and rereading. If no error occurs at rereading, the automatic alternate assignment  
is not performed.  
2) When a write error occurs and the error does not recovered.  
6 - 12  
C141-E116-01EN  
6.5  
Read-Ahead Cache  
After a read command which reads the data from the disk medium is completed, the read-ahead  
cache function reads the subsequent data blocks automatically and stores the data in the data  
buffer.  
When the next command requests to read the read-ahead data, the data can be transferred from the  
data buffer without accessing the disk medium. The host can access the data at higher speed.  
6.5.1  
Data buffer configuration  
The device has a 2MB data buffer. The buffer is used by divided into two and other commands  
parts; for MPU work, for read cache of read commands and other commands (see Figure 6.9).  
2,048 KB (2,097,152 bytes)  
For R/W command  
for MPU work  
1,968 KB  
80 KB  
(2,015,232 bytes)  
(81,920 bytes)  
Figure 6.9 Data buffer configuration  
The read-ahead operation is performed at execution of the READ SECTOR(S), READ  
MULTIPLE, or READ DMA command, and read-ahead data are stored in the buffer for read  
cache.  
C141-E116-01EN  
6 - 13  
6.5.2  
Caching operation  
The caching operation is performed only at receipt of the following commands. The device transfers  
data from the data buffer to the host system if the following data exist in the data buffer.  
All sector data to be processed by the command  
A part of data including the starting sector to be processed by the command  
When a part of data to be processed exist in the data buffer, the remaining data are read from the  
disk medium and are transferred to the host system.  
(1)  
Commands that are object of caching operation  
The following commands are object of caching operation.  
READ SECTOR (S)  
READ MULTIPLE  
READ DMA  
When the caching operation is disabled by the SET FEATURES command, no caching operation  
is performed.  
(2)  
Data that are object of caching operation  
The following data are object of caching operation.  
1) Read-ahead data read from the disk medium in the data buffer after completion of the  
command that are object of caching operation.  
2) Data transferred to the host system once by requesting with the command that are object of  
caching operation. When the sector data requested by the host does not finish storing in the  
buffer for read cache, it is not object of caching operation. And also, when the sequential hit  
occurs continuously, the caching data required by the host becomes invalid.  
(3)  
Invalidating caching data  
Caching data in the data buffer is invalidated in the following case.  
1) Commands other than the following commands are issued (all caching data are invalidated)  
WRITE SECTOR(S)  
WRITE DMA  
WRITE MULTIPLE  
CHECK POWER MODE  
2) Caching operation is disabled by the SET FEATURES command.  
3) Command issued by the host is terminated with an error.  
4) Soft reset or hard reset is executed, or power is turned off.  
6 - 14  
C141-E116-01EN  
6.5.3  
Usage of read segment  
This subsection explains the usage of the read segment buffer at following cases.  
(1)  
Miss-hit (no hit)  
A lead block of the read-requested data is not stored in the data buffer. The requested data is read  
from the disk media.  
1) Sets the host address pointer (HAP) and the disk address pointer (DAP) to the sequential  
address to the last read segment.  
HAP  
Segment only for read  
DAP  
2) Transfers the requested data that already read to the host system with reading the requested  
data from the disk media.  
Stores the read-requested  
data upto this point  
HAP  
Empty area  
Read-requested data  
DAP  
3) After reading the requested data and transferring the requested data to the host system had  
been completed, the disk drive continues to read till a certain amount of data is stored.  
HAP  
(stopped)  
Read Ahead Data  
Read-requested data  
(stopped)  
DAP  
4) Following shows the cache enabled data for next read command.  
Cache enabled data  
Start LBA  
Last LBA  
6 - 15  
C141-E116-01EN  
(3)  
Sequential read  
When the disk drive receives the read command that targets the sequential address to the previous  
read command, the disk drive tries to fill the buffer space with the read ahead data.  
a. Sequential command just after non-sequential command  
1) At receiving the sequential read command, the disk drive sets the DAP and HAP to the  
sequential address of the last read command and reads the requested data.  
HAP  
Mis-hit data  
Empty data  
DAP  
2) The disk drive transfers the requested data that is already read to the host system with  
reading the requested data.  
HAP  
Mis-hit data  
Requested data  
Empty data  
DAP  
3) After completion of the reading and transferring the requested data to the host system, the  
disk drive performs the read-ahead operation continuously till a certain amount of data is  
stored.  
HAP  
Read-  
ahead  
data  
Empty  
data  
Mis-hit data  
Requested data  
DAP  
6 - 16  
C141-E116-01EN  
b. Sequential hit  
When the last sector address of the previous read command is sequential to the lead sector  
address of the received read command, the disk drive transfers the hit data in the buffer to the  
host system.  
The disk drive performs the read-ahead operation of the new continuous data to the empty area  
that becomes vacant by data transfer at the same time as the disk drive starts transferring data  
to the host system.  
1) In the case that the contents of buffer is as follows at receiving a read command;  
HAP (Completion of transferring requested data)  
Read-ahead data  
Hit data  
DAP  
Last LBA Start LBA  
2) The disk drive starts the read-ahead operation to the empty area that becomes vacant by  
data transfer at the same time as the disk drive starts transferring hit data.  
HAP  
Read-ahead data  
New read-ahead data  
Hit data  
DAP  
3) After completion of data transfer of hit data, the disk drive performs the read-ahead  
operation for the data area of which the disk drive transferred hit data.  
HAP  
DAP  
Read-ahead data  
C141-E116-01EN  
6 - 17  
(3)  
Full hit (hit all)  
All requested data are stored in the data buffer. The disk drive starts transferring the requested  
data from the address of which the requested data is stored. After completion of command, a  
previously existed cache data before the full hit reading are still kept in the buffer, and the disk  
drive does not perform the read-ahead operation. If the disk drive receives a full hit command  
while performing the read-ahead operation, the disk drive starts transferring the requested data  
without stopping the read-ahead operation.  
1) In the case that the contents of the data buffer is as follows for example and the previous  
command is a sequential read command, the disk drive sets the HAP to the address of which  
the hit data is stored.  
Last position at previous read command  
HAP  
HAP (set to hit position for data transfer)  
Cache data  
Full hit data  
Cache data  
DAP  
Last position at previous read command  
2) The disk drive transfers the requested data but does not perform the read-ahead operation.  
HAP  
(stopped)  
Cache data  
Full hit data  
Cache data  
(4)  
Partially hit  
A part of requested data including a lead sector are stored in the data buffer. The disk drive starts  
the data transfer from the address of the hit data corresponding to the lead sector of the requested  
data, and reads remaining requested data from the disk media directly.  
Following is an example of partially hit to the cache data.  
Cache data  
Last LBA  
Start LBA  
6 - 18  
C141-E116-01EN  
1) The disk drive sets the HAP to the address where the partially hit data is stored, and sets the  
DAP to the address just after the partially hit data.  
HAP  
Partially hit data  
Lack data  
DAP  
2) The disk drive starts transferring partially hit data and reads lack data from the disk media at  
the same time.  
HAP  
(stopped)  
Requested data to be transferred  
Partially hit data  
Lack data  
DAP  
C141-E116-01EN  
6 - 19  
6.6  
Write Cache  
The write cache function of the drive makes a high speed processing in the case that data to be  
written by a write command is logically sequent the data of previous command and random write  
operation is performed.  
When the drive receives a write command, the drive starts transferring data of sectors requested by  
the host system and writing on the disk medium. After transferring data of sectors requested by the  
host system, the drive generates the interrupt of command complete. Also, the drive sets the  
normal end status in the Status register. The drive continues writing data on the disk medium.  
When all data requested by the host are written on the disk medium, actual write operation is  
completed.  
The drive receives the next command continuously. If the received command is a "sequential  
write" (data to be written by a command is logically sequent to data of previous command), the  
drive starts data transfer and receives data of sectors requested by the host system. At this time, if  
the write operation of the previous command is still been executed, the drive continuously executes  
the write operation of the next command from the sector next to the last sector of the previous  
write operation. Thus, the latency time for detecting a target sector of the next command is  
eliminated. This shortens the access time. The drive generates an interrupt of command complete  
after completion of data transfer requested by the host system as same as at previous command.  
When the write operation of the previous command had been completed, the latency time occurs to  
search the target sector.  
If the received command is not a "sequential write", the drive receives data of sectors requested by  
the host system as same as "sequential write". The drive generates the interrupt of command  
complete after completion of data transfer requested by the host system. Received data is  
processed after completion of the write operation to the disk medium of the previous command.  
Even if a hard reset or soft reset is received or the write cache function is disabled by the SET  
FEATURES command during unwritten data is kept, the instruction is not executed until  
remaining unwritten data is written onto the disk medium.  
The drive uses a write data as a read cache data. When a read command is issued to the same  
address after the write command, the read operation to the disk medium is not performed.  
When an error occurs during the write operation, the drive makes retry as much as possible. If the  
error cannot be recovered by retry, the drive stops the write operation to the erred sector, and  
continues the write operation from the next sector if the write data is remained. (If the drive stacks  
a write command, for that the drive posts the command completion, next to the command that write  
operation is stopped by error occurrence.) After an error occurs at above write operation, the drive  
posts the error status to the host system at next command. (The drive does not execute this  
command, sets the error status that occurred at the write operation, and generates the interrupt for  
abnormal end. However, when the drive receives a write command after the completion of error  
processing, the drive posts the error after writing the write data of the write command.)  
6 - 20  
C141-E116-01EN  
At the time that the drive has stopped the command execution after the error recovery has failed,  
the write cache function is disabled automatically. The releasing the disable state can be done by  
the SET FEATURES command. When the power of the drive is turned on after the power is  
turned off once, the status of the write cache function returns to the default state. The default state  
is “write cache enable”, and can be disable by the SET FEATURES command.  
The write cache function is operated with the following command.  
WRITE SECTOR(S)  
WRITE MULTIPLE  
WRITE DMA  
IMPORTANT  
When the write cache function is enabled, the transferred data from the  
host by the WRITE SECTOR(S) is not completely written on the disk  
medium at the time that the interrupt of command complete is  
generated. When the unrecoverable error occurs during the write  
operation, the command execution is stopped. Then, when the drive  
receives the next command, it generates an interrupt of abnormal end.  
However an interrupt of abnormal end is not generated when a write  
automatic assignment succeeds. However, since the host may issue  
several write commands before the drive generates an interrupt of  
abnormal end, the host cannot recognize that the occurred error is for  
which command generally. Therefore, it is very hard to retry the  
unrecoverable write error for the host in the write cache operation  
generally. So, take care to use the write cache function.  
C141-E116-01EN  
6 - 21  
Comments concerning this manual can be directed to one of the following addresses:  
FUJITSU LIMITED  
Business Planning  
FUJITSU ICL ESPAÑA S.A.  
Almagro 40, 28010 Madrid, SPAIN  
Solid Square East Tower  
580 Horikawa-cho,Saiwai-ku, Kawasaki,  
210-0913, Japan  
TEL:  
FAX:  
34-1-581-8100  
34-1-581-8125  
TEL:  
FAX:  
81-44-540-4056  
81-44-540-4123  
FUJITSU AUSTRALIA LIMITED  
2 Julius Avenue (Cnr Delhi Road) North Ryde N.S.W. 2113,  
AUSTRALIA  
FUJITSU COMPUTER PRODUCTS OF AMERICA, INC.  
2904 Orchard Parkway, San Jose,  
TEL:  
FAX:  
61-2-9776-4555  
61-2-9776-4556  
California 95134-2009, U.S.A.  
TEL:  
FAX:  
1-408-432-6333  
1-408-432-3908  
FUJITSU HONG KONG LTD.  
10/F., Lincoln House, 979 King’s Road, Taikoo Place, Island East,  
HONG KONG  
FUJITSU CANADA INC.  
2800 Matheson Blvd. East, Mississauga, Toronto,  
Ontario L4W 4X5, CANADA  
TEL:  
FAX:  
852-2827-5780  
852-2827-4724  
TEL:  
FAX:  
1-905-602-5454  
1-905-602-5457  
FUJITSU KOREA LTD.  
Coryo Finance Center Bldg, 23-6, YoulDo-Dong,  
Young DungPo-Gu, Seoul, Republic of KOREA  
FUJITSU EUROPE LIMITED  
2, Longwalk Road, Stockley Park, Uxbridge,  
Middlesex UB11 1AB, ENGLAND  
TEL:  
FAX:  
82-2-3787-6000  
82-2-3787-6070  
TEL:  
FAX:  
44-81-573-4444  
44-81-573-2643  
FUJITSU COMPUTERS (SINGAPORE) PTE. LTD  
20 Science Park Road #03-01,  
TELETECH PARK SINGAPORE SCIENCE PARK II,  
Singapore 117674  
FUJITSU DEUTSCHLAND GmbH  
Frankfurter Ring 211, 80807 München, GERMANY  
TEL:  
FAX:  
65-777-6577  
65-771-5499  
TEL:  
FAX:  
49-89-323780  
49-89-32378100  
FUJITSU TAIWAN LTD.  
FUJITSU NORDIC AB  
Kung Hans Väg 12, S-192 68 Sollentura, SWEDEN  
8F, Hun Tai Center, 168-170, Tun Hwa North Road,  
1st Sec., Taipei, TAIWAN  
TEL:  
FAX:  
46-8-626-4500  
46-8-626-4588  
TEL:  
FAX:  
886-2-545-7700  
886-2-717-4644  
FUJITSU ITALIA S.p.A.  
Via Nazario Sauro, 38 20099 Sesto S. Giovanni (MI), ITALY  
FUJITSU SYSTEMS BUSINESS (MALAYSIA) SDN. BHD.  
Fujitsu Plaza, 1A, Jalan Tandang 204, P.O. Box 636 Pejabat Pos  
Jalan Sultan  
TEL:  
FAX:  
39-2-26294-1  
39-2-26294-201  
46770 Petaling Jaya, Selangor Darul Ehsan, Malaysia  
TEL:  
FAX:  
60-3-793-3888  
60-3-793-0888  
FUJITSU FRANCE S.A.  
I, Place des Etas-Unis, SILIC 310,  
94588 Rungis Cedex, FRANCE  
FUJITSU SYSTEMS BUSINESS (THAILAND) LTD.  
12th F1., Olympia Thai Tower, 444 Rachadapisek Road,  
Samsennok, Huay Kwang, Bangkok 10320, Thailand  
TEL:  
FAX:  
33-1-41-80-38-80  
33-1-41-80-38-66  
TEL:  
FAX:  
66-2-512-6066  
66-2-512-6068  
11  
FUJITSU LIMITED  
Reader Comment Form  
We would appreciate your comments and suggestions for improving this publication.  
Publication No.  
Rev. Letter Title  
Current Date  
How did you use this publication?  
Learning  
Is the material presented effectively?  
Sales  
Installing  
Fully  
Well  
Illustrated  
Well  
Organized  
Clean  
covered  
Reference  
Operating  
Maintaining  
What is your overall rating of this publication?  
What is your occupation?  
Very Good  
Good  
Fair  
Very Poor  
Poor  
Your other comments may be entered here. Please be specific and give page,  
paragraph and line number references where applicable.  
Your Name & Return Address  
Thank you for your interest. Please send this sheet to one of the addresses in a left page.  

Mr Coffee CUBIKA User Manual
Motorola V60p User Manual
Microsoft Sql Server 2008 R2 Small Business Edition C9C00500 User Manual
LG Electronics MMBB000000010 User Manual
Insignia X7510 User Manual
HTC FUZE User Manual
HP Hewlett Packard Hewlett Packard Computer Accessories 11865A User Manual
GPX Clock Radio C224B User Manual
Echo Bear Cat CH6614H 72620 User Manual
Carrier WEATHERMASTER Z8075 105 User Manual