ST5000DM003 Startup Fail

[stef12345]

Hi,

my Segate ST5000DM003 suddenly failed - it doesnt seem to go fully through the Init Sequence.

Symptoms:

- Spins up, Heads move, but acting somehow "abnormally"
- no Connection via TTL, Logic Analyzer on TX shows no activity
- SATA sends requests, but no valid reply at all, its stuck in Identifying the drive
- so of course smartctl would be senseless

After power up, i hear the head assembly moving quite a good bit, but it doesnt do any longer the "harsh" "zig-zag" movement before being ready, as it was before.
Listening to the HDD tells me that the head assembly moves a bit at this moment, but does not fulfill the harsh movement as mentioned before.

For some reasons while trying to identify the problem, the 12V 0 Ohm Resistors and the TVS Diode "failed", causing a short Circuit , so i had to bridge the 0 Ohm Resistors and remove the Diode. Spins up again.
but still the same problem.

No visible damage otherwise. Tried to "reflow" everything a bit with the heatgun, but no success. Of course i cleaned the contacts aswell.

I suppose its a firmware Fail for any reason, since the heads are being initially moved, but at the moment i am clueless whether the firmware is integrated into the MCU or is provided externally.
Or could it be likely that the Platter calibration data is stored externally? Is the 8 Pin Part near the MCU a EEPROM?

Any chance to rewrite the firmware (i have RS232-TTL, an STM32 MCU programmer, the possibility to run some Code on a STM32/8051/AVR/Pi Pico).
Any Pinouts available for the Main MCU?

Attached is a picture form another Site which shows the Main PCB, altrough it provides a larger capacity, it seems to be identical.

Any ideas are very welcome - thank you!

Best regards
Stef

Attachments

[stef12345]

In addition - keeping it connected and wait until the SATA Controller gives up, then performing smartctl outputs the correct Drive Data (Model, Size, LBA etc), but in Depth Information cannot be gathered

Code:

=== START OF INFORMATION SECTION ===
Device Model:     ST5000DM003
Serial Number:    XXXXXXX
LU WWN Device Id: 5 000c50 0aa7247a2
Firmware Version: 0001
User Capacity:    5.000.981.078.016 bytes [5,00 TB]
Sector Size:      512 bytes logical/physical
Rotation Rate:    7200 rpm
Device is:        Not in smartctl database 7.3/5528
ATA Version is:   ATA8-ACS T13/1699-D revision 4
SATA Version is:  SATA 3.0, 6.0 Gb/s
Local Time is:    Mon Apr 28 04:28:03 2025 CEST
SMART support is: Available - device has SMART capability.
SMART support is: Enabled
AAM feature is:   Unavailable
APM feature is:   Unavailable
Rd look-ahead is: Enabled
Write cache is:   Enabled
DSN feature is:   Unavailable
ATA Security is:  Disabled, NOT FROZEN [SEC1]
Read SCT Status failed: scsi error device will be ready soon
Wt Cache Reorder: Unknown (SCT Feature Control command failed)

Read SMART Data failed: scsi error device will be ready soon

=== START OF READ SMART DATA SECTION ===
SMART Status command failed: scsi error device will be ready soon
SMART overall-health self-assessment test result: UNKNOWN!
SMART Status, Attributes and Thresholds cannot be read.

Read SMART Log Directory failed: scsi error device will be ready soon

ATA_READ_LOG_EXT (addr=0x00:0x00, page=0, n=1) failed: scsi error device will be ready soon
Read GP Log Directory failed

SMART Extended Comprehensive Error Log (GP Log 0x03) not supported

Read SMART Error Log failed: scsi error device will be ready soon

SMART Extended Self-test Log (GP Log 0x07) not supported

Read SMART Self-test Log failed: scsi error device will be ready soon

Selective Self-tests/Logging not supported

Read SCT Status failed: scsi error device will be ready soon

Read SCT Status failed: scsi error device will be ready soon
SCT (Get) Error Recovery Control command failed

Device Statistics (GP/SMART Log 0x04) not supported

Pending Defects log (GP Log 0x0c) not supported

ATA_READ_LOG_EXT (addr=0x11:0x00, page=0, n=1) failed: scsi error device will be ready soon
Read SATA Phy Event Counters failed


No Success with Seatools either. Any Request to perform Drive diagnostics does immediate fail.

That is all i got. Yes, meanwhile i found out that the Firmware resides in the 16 MBit Chip, but asking myself if the drive Calibration Data resides in there, too. I guess that i have to take the Chip from the PCB and read it, as usually, out.

[LeviFlix]

Hello, I have the same problem you are having with another model ST3000VN007. In my case, the HD only recognizes 3GB, but the HD is 3TB. I am also suspicious of the HD BIOS.

[SWM]

The flash chip of the HDD contains unique data for each disk. If this chip or its contents are damaged, this HDD is a brick.
If the HDD ROM is damaged, the disk usually does not spin up.
Seagates are diagnosed via the terminal. In your cases, the level is 1.8V(USB-TTL 1.8V).
And not everyone can even do this...

https://forum.hddguru.com/viewtopic.php?f=1&t=45269

And yes, they have SWD ARM. But only the manufacturer has the documentation...

[stef12345]

I used 3.3V TTL (fed my MAX3232 with 3.3 Volts); althrough i had this drive (got it from my nephew, already being used for quite some time) back then connected via Serial at this level, because i was wondering about its quite extreme SMR behaviour (never had one before and thought it was a fault, i was even not aware that this drive was working with SMR) and guessed i had do so some Preformatting in order to resolve this.

In the current faulty state, i got the drive, the same day before posting this issue, running once again by reconnecting power several times and waiting for the "magic click" to happen, after that the HDD just worked fine and of course i pulled then as much as i could from it. The first sign that the drive starts to fail was that it couldn't be unmounted by the LVM any longer. I am not quite sure if this is at the end a platter or head related issue, or that it wanted to perform some CMR-SMR Actions - and cannot continue for whatever reason, or the coil amplifiers gave up. But that doesnt explain the Serial behaviour.

Maybe i have some luck, go down to 1.8 Volts by using Transistors or whatever as a level shifter and get access again. Maybe even my logic analyzer couldn´t see this level. But as said, i had back then already access to this drive at 3.3 Volts.

I'm not mad if i couldn't use the drive any longer, because of its SMR nature, but 5 TB is still a quite good number.

[SWM]

Well, if you copied the data - You can do whatever you want with it. Maybe there is 3.3V or you have a good 3232.
Repairing the SMR, and just disks over 1TB - is not cost-effective in terms of time.
You need to look at the answer in the terminal when power is supplied to the HDD. With the SATA disconnected.
Sometimes you need to wait a few minutes for the reaction to the Ctrl-Z.

[stef12345]

Unfortunaly it doesn't react to CTRL-Z at all, no matter on which point....i see in the Terminal output a greyish Block Char, for maybe each 2-3 Minutes and sometimes it echoes the character back that i enter, but not always. If the MCU is fed with 1.8 Volts VCC, i am asking myself if enough voltage is present at all, eg if the Voltage converter 5V=>1.8V is failing at a certain point.

I have a look. For me its not any longer the need to get access again, but i'm curious about the fault.

[SWM]

Look at the oscilloscope on the Tx HDD pin when the HDD power is turned on. GPIO USART do not have external additional protection against overvoltage.

[stef12345]

Well, i've read on Reddit about a bug regarding the ST1000/3000DM Series, which eventually turn the Drives to be Stuck in BSY state, caused by a Media Cache error. My drive is from the F3 Series, altrough from the Kahuna Family and not the Grenada Family, where the ST1000DM belongs to, but maybe the bug made its way to Kahuna aswell. There is technically no way to fix this issue, even not by rewriting the Firmware - as long as the bug isn´t be sorted out.

One thing is for sure - never again Seagate; i had my best experiences with WD. They may be relatively slow here and there, but my drives from this MF were reliable.

[stef12345]

I saw indeed some High/Low transitions on TX, but it is barely repeatable. My suspicion is that there is a external Pullup on TX which failed, if its not the GPIO.

[SWM]

I have seen external pullup on Rx more often. GPIO Tx is configured with PushPull most likely.

[stef12345]

Anyway, my last message needs to be approved, but is far as i can tell there are similar issues with ST1000/3000-Drives related to the Media Cache, eventually leading to a endless BSY state, its a FW bug. Since i dont trust the given electronics any longer, i put it aside, maybe i get my hands sometime onto a replacement PCB. There is not much to choose from.