ST2000DM001 damaged electronics

[Jano952]

Hello guys,
Story: I have accidentally shorted all molex pins together, computer turned off. I turned it back on and went on toilet. When i came back i saw that disk is not detected, so i restarted PC, HDD beeped once and then bad smell came out. I quickly turned PC off, checked connectors, and turned it back on. Disk did not spin up and was not detect in BIOS. So i disassembled PCB and saw that one IC is probably burned up. There are two similar ICs, one is marked 7157B (probably LX7157B) and the other is half burned so I can see only 71 on it, but i expect they are same. I have checked motor coils resistance, they have around 3.3 Ohm.
I had lots of data there (and retard had old/no backups). What do you suggest ? In this topic http://forum.hddguru.com/viewtopic.php?t=27449&p=187906 fzabkar suggested using high current LDO regulator instead of 7157B switching converter. How much will probably cost replace of PCB ? What should i measure ?

Thanks

Attachments

[michael chiklis]

If you really need to get back your data i suggest you to send it to a pro, you might do things worse (example kill rom chip, would mean data lost permanently).
PCB issue should cost about 350 EUR to fix it, unless there are further problems.
This is not a DIY case if you don't know about electronics, soldering and multimeter testing.

[data-medics]

I'd forget about trying to repair the PCB. You see one burnt component, but I'm sure that there are more.

If you go to any data recovery place that has PC-3000 they can most likely read the ROM code and program a new board for you. I don't know what prices are like in your area, but here in the USA I charge about $500 for the whole process of replacing the PCB and imaging the data onto a new drive.

If you are intent on doing it yourself, you'll need to transfer the ROM chip circled in Blue on the picture to a compatible board (match the number printed in the green part on the opposite side - something like 100687658 REV B)

Hope that helps.

[data-medics]

Forgot the picture before.

Attachments

[michael chiklis]

data-medics explained everything to solve the problem by yourself, but keep in mind that DIY its risky if you don't know about electronics (you also need tools to deal with rom).

[data-medics]

Best way to unsolder is to use a hot air rework station, and use as little heat as possible to melt the solder. Make sure to add flux to the contacts before heating, and keep the airflow low so as not to blow other components off the board. I'd practice by removing the one from the donor PCB first before heating the original (if you cook that chip, your data is really gone).

Also make sure to pay attention to the orientation of the chip.

[data-medics]

data-medics explained everything to solve the problem by yourself, but keep in mind that DIY its risky if you don't know about electronics (you also need tools to deal with rom).

I agree, but I realize that the local options for pro recovery in Slovakia may be limited, and DIY might be the only real option.

[michael chiklis]

I can do this job for him if he sends the drive to Italy

[data-medics]

That's if he feels comfortable mailing it. 90% of my business is local because people don't want to risk mailing out.

[Jano952]

thank for replies,
soldering is not problem, I actually know a bit about electronics.

I did some research and measurements, and I will probably take it to some data recovery center ( found one in Slovakia, actually good prices), because that damage seems to be quite worse and there might be that SMOOTH IC damaged, so using simple external regulator would not be possible (and I would need to swap whole PCB and ROM chip). ROM chip is powered through regulator that seems to be okay, so it should be okay too.

I was also thinking about programing some Atmega microprocessor to save ROM chip data into computer before experimenting (its just classic SPI Flash, easy to read), but I am just too scared i will f**k it up completely .

well, sending such fragile item to italy is quite risky in my opinion, but thanks for help

Anyway, here is picture with everything i measured

BTW serial number is : 100664987 REV A

Attachments

[fzabkar]

@Jano952, ISTM that there should not be a low resistance path between the R47 and 1R0 inductors. Furthermore, I would have expected that the 1R0 inductor would have been the Vio supply for the SPI flash rather than the R47 coil.

I'm thinking that the following arrangement would make more sense:

1R0 coil + diode + SMOOTH -> Vio buck converter (+1.8V for SDRAM and SPI flash)
R47 coil + PWM controller #1 -> Vcore_1 for dual-core MCU
R47 coil + PWM controller #2 -> Vcore_2 for dual-core MCU
1R2 coil + 4184P FETKY (?) + SMOOTH -> -5V inverting buck regulator for preamp
coil + 6855A (near jumpers) -> some other DC-DC converter

If I'm right, then I believe that the MCU may be dead. Moreover, if there really is a short between the Vio and Vcore coils, then this would suggest that the short circuit exists within the MCU.

[Jano952]

ah, you re right, i was too quick with measurements and i did mistakes

Between 5V and 0 V is 25 Ohm, BOTH ways, I do not know what causes it (~shorted by dead PWM controller ?)
1R2 + 4184P seems to be fine, trasistor is OK, diode too, preamp might be okay (if 5V over voltage did not kill him),
between 1R0 and R47 from damaged IC is 60 Ohm one way, 50 Ohm other way (yep i did mistake here last time),
between R47 from undamaged IC and ROM there is really 0 Ohm. It probably powers both MCU and ROM.

how can i check preamp ? just check if it is not shorted with multimeter between 0 and 5 V, 0 and -5V ?

Well, anyway, I think I give up for now, I ll take it to recovery center and ask for price, if it cost too much then I ll do some experiments at home

[fzabkar]

If the problem is only a PCB fault, then a replacement PCB plus ROM transfer should cost no more than US$50.

http://www.onepcbsolution.com/100664987.html
http://www.hdd-parts.com/13022202.html

As for checking the preamp, I would do as you suggest.

The 25 ohm reading could possibly be due to the 5V TVS diode, but I suspect that your diagnosis is probably correct. In any case you should probably remove the diode to confirm whether it is responsible. Also, if the diode does turn out to be leaky, then this would prove that the drive was damaged by an external overvoltage.

MY Vio and Vcore guesses appear to be incorrect, but ISTM that the MCU may still be damaged. One last check would be to measure the resistances between ground and each of the coils. If the ROM has a low resistance, then there is a risk that it, too, may have been damaged.

[Jano952]

Well, i measured Flash chip, on 2000 Ohm range it has 330 Ohm and on 20k range it has 1,4k Ohm. I really dont understand why
preamp has 400 Ohm and 500 Ohm resistance on 2000 range and more than 2k on both supplies on 20k range.
actually i have no idea if it is good or not, I have never measured resistance on such a complex low voltage ICs

Problem is in coil R47 behind destroyed PWM reg. I has 0 Ohm to ground.. It might be shorted by PWM reg, or MCU might be dead, Coil resistance is too small to measure from which side of coil is that short (through MCU or PWM reg)

Coil 1R0 has 50 Ohm to Ground and I think it is quite low.

Without all Datasheets and schemes this is just lottery and repair is nearly impossible.
I ll just find similar PCB and change ROM, thanks for help .

[fzabkar]

The resistance readings for the flash and preamp appear to be OK. The reason for the strange readings is that you are measuring the devices by injecting a low current from a low voltage source within the meter. The meter's currents/voltages are insufficient large to turn on the PN junctions within semiconductor devices, so the measured resistances are fairly high. When the devices are operating, those same PN junctions are turned on and the current draw is relatively large. For example, an MCU core might draw 0.5A from 1V, say. That corresponds to a dynamic resistance of 2 ohms.

In short, I'd say that one of the MCU's cores is shorted to ground. I suspect that the PWM controller failed with a shorted MOSFET chopper. This would then have resulted in the full supply voltage (5V) being applied to the MCU, and this in turn would have resulted in a shorted core.

Good luck with the PCB swap. BTW, you might like to measure the resistances and voltages on the donor PCB for comparison purposes.

[Jano952]

Thanks, just one last question.

Mine PCB has Winbond W25Q40BW flash IC. I has more and slightly different commands (status read commands, supports Quad SPI), but basic commands (Read, Fast Read) are same as SST25WF040 flash. Can I clone that Winbond chip to 25WF flash with Atmega MCU and use 25WF flash (dont have any special tools, just MCUs with SPI) ? Original ROM is too valuable and I dont want to accidentally destroy it while soldering ( who knows what might happen )

[fzabkar]

I can't answer your question, but there was a recent thread where "michael chiklis" appears to have found that a Seagate PCB would not "boot" with a "compatible", fully functional, Atmel SPI flash whereas the original flash chip was OK. If his results are reproducible, then this would suggest that the PCB firmware may reject a non-native flash chip based on its ID.

viewtopic.php?f=1&t=29113

[Jano952]

Oh no, these Winbond flash chips are hard to find (only few suppliers on aliexpress or ebay), and they also have feature called "Unique ID Number" according to datasheet. I hope it uses only Manufacturer and Device ID and not these Unique IDs. Never mind, I ll copy it and while I ll be waiting for donor board I ll try to repair PWM regulator and see if it blows up

Thank you very much

[Jano952]

Hello, It s me again,
I have removed ROM Flash from PCB and saved its content with ATmega328, but I am not sure if everything is saved right ( I programmed in assembler, MCU is running at 1.8V, transfered data through UART to PC, thats just too many error factors )

Please, could you help me check it ? Here is link to the file i saved from flash https://www.dropbox.com/s/j37byzsgm9e01 ... 0DM001.bin

[Jano952]

Okay tripple post, sorry

I have been thinking and...I decided that data is worth max 30-50€ so I have just replaced LX7157B with LM338 to supply 1.7-1.8V. Magically Disk spins up, and also sends Terminal data

If voltage is 1.7V termina shows errors in log like this

Code:

Boot 0x40M

Spin Up
TCC-001F[0x000065B4][0x00006A20][0x00006E8C]

Trans.


Rst 0x40M

MC Internal LPC Process

Spin Up
TCC-001F
SIM Error 1000 LBA 0000000000038C07 FD FCFFF1FF
RW Error C44400CD
SIM Error 1002 LBA 0000000000038D33 FD FCFFF3FF
RW Error C44400BD
SIM Error 1002 LBA 0000000000038EF3 FD FCFF3DF
RW Er2or C44400CD
SIM Error 1002 LBA 0000000000038C07 FD FCFFF3FF
RW Error C44400CD

(P) SATA Reset


SIM Error 1002 LBA 0000000000034C5C FD FC37830A
RW Error C44400CD PARS
SIM ErroR 1002 LBA 0000000000039BF8 FD FCFFF3FF
RW Error C44400CD

When i push voltage up to 1,8V is starts clicking or writes this

Code:

TCC-001F[0x000065B4][0x00006A20][0x00006E8C]

Trans.


Rst 0x40M

MC Internal LPC Process

Spin Up
TCC-001F
(DOS) Table Read FAIL

DOS: MAGIC CONFIG NUMBER MISMATCH
(DOS) Requesting File Save

(DOS) Table Read FAIL
SIM Error 1002 LBA 0000000000039C16 FD FC37C137
RW Error C44400CD

ECMainPOR: EXCEPTION: SIM
SIM Error 1002 LBA 0000000000034C5C FD FC37830A
RW Error C44400CD
SIM Error 100" LBA 0000000000034C5C FD FC37830A
RW Error C44400CDWriteDCOInfo FAILED during Initialization

Starting LBA of RW Request=0000000000087B70  Length=0000180

ProcessRWError -Read-   at LBA 0000000000087B70  Sense Code=43110081

Starting LBA of RW Request=0000000000087B74  Length=0000017C

ProcessRWError -Read-   at LBA 0000000000087B74  Sense Code=43110081

Starting LBA ob RW Reqtest=000000000087B78  Length=00000178

ProcessRWError -Read-   at LBA 0000000000087B78  Sense Code=43110081

Starting LBA of RW Request=0000000000087B7C  Length=00000174

ProcessRWError -Read-   at LBA 0000000000087B7C  Sense Code=43110081

Starting LBA of RW Rdquest=0000000000087B80  Lafgth=00000170

ProcessRWError -Read-   at LBA 0000000000087B80  Sense CoDe=43110081

Starting LBA of RW Request=0000000000087B84  Length=0000016C

Some characters seems to be damaged, (voltage is not stable ?) I think there is nothing to do with this HDD, MCU is probably damaged, or I have damaged ROM, never mind, If anyone has good idea I ll be happy to hear it

It is intresting that it Writes everytime something differend and do different things ...