PSU broke and took HDD with it

[callum92]

As the title says, the PSU broke in my brother's computer and it seems to have damaged the hard drive.
The drive doesn't spin up when connected to PC and I was wondering what are the chances it has just burnt something out on the drive's PCB?
If this is the case would it be possible to swap the PCB from a donor drive, I have an almost identical drive here as well?

Here are the specs for the drive:

Model: ST3300831AS
S/N: 3NF00XJS
P/N: 9Y7384-301
Firmware: 3.01
Data Code: 05243

[jermy]

was wondering what are the chances it has just burnt something out on the drive's PCB?

there is

See http://www.users.on.net/~fzabkar/HDD/TVS_diode_FAQ.html

If this is the case would it be possible to swap the PCB from a donor drive, I have an almost identical drive here as well?

what does "almost" mean

[callum92]

Hi jermy, thanks for the reply. Well they seem pretty similar to me but I am not sure what information is important.
The two drives have the same model, P/N and firmware. The serial numbers on the PCB are similar: 7522EJZZ and 7522LX7U.

These are pics of the broken drive:
PCB - http://imgur.com/gPC46z6
Front Label - http://imgur.com/COwVMkz

And these are pictues of the possible donor drive:
PCB - http://imgur.com/qlHSARh
Front Label - http://imgur.com/PzQRTMX

[callum92]

I read the article that you posted about TVS diode's, it looks promising. Can you determine where the TVS diode is on the drive from the pics I posted?

Last edited by callum92 on June 26th, 2015, 7:49, edited 1 time in total.

[jermy]

S/N doesn't matter, it's a match

[jermy]

I read the article that you posted about TVS diode's

i didn't posted it, fzabkar did
it's below the power connection

[callum92]

Ok then I read the article you linked to, all credits go to fzabkar

One of the two I have highlighted or both of them?
http://imgur.com/HUYXVoa

Do you think that if I removed the TVS diode there is a good chance the drive will work? Are you aware of this drive having any additional protection as mentioned in the article?

[callum92]

After looking closely at the markings on the two diode's, the one of the left in the picture is marked 'AE GP435', is this the right one?

[hddguy]

Just make straight swap of PCB - ROM on this series is not adaptive and its very possible you need no transfer of adaptives. Also, do not connect to same system that causes this issue initially. If you exchange PCB and it makes strange sounds or knocking consult a professional. If it appears fine but no identification it could be due to differences in ROM so you should swap it. Any issues or problems or if there is any uncertainty then you should consult a professional - at this point it should be a relatively simple fix for a decent pro.

[jermy]

One of the two I have highlighted or both of them?

yes

Do you think that if I removed the TVS diode there is a good chance the drive will work? Are you aware of this drive having any additional protection as mentioned in the article?

why all the trouble ? you have a match PCB, just do as hddguy suggested

Just make straight swap of PCB - ROM on this series is not adaptive and its very possible you need no transfer of adaptives. Also, do not connect to same system that causes this issue initially. If you exchange PCB and it makes strange sounds or knocking consult a professional. If it appears fine but no identification it could be due to differences in ROM so you should swap it. Any issues or problems or if there is any uncertainty then you should consult a professional - at this point it should be a relatively simple fix for a decent pro.

After looking closely at the markings on the two diode's, the one of the left in the picture is marked 'AE GP435', is this the right one?

this is the 5V the other one is the 12V

[Amarbir[CDR-Labs]]

After looking closely at the markings on the two diode's, the one of the left in the picture is marked 'AE GP435', is this the right one?

Well,
If PCB Model Marking Is Same ,Just Swap The ROM From Patient To Donor And You Are Good With The Same

[jermy]

Well,
If PCB Model Marking Is Same ,Just Swap The ROM From Patient To Donor And You Are Good With The Same

as hddguy said

ROM on this series is not adaptive and its very possible you need no transfer of adaptives.

this one most likely don't need a ROM swap

[callum92]

So I followed your advice and just swapped the PCB's and the drive is working again. Thanks to everyone who posted with suggestions, I really appreciate your help .
Now just out of curiosity, which component(s) do you think are likely to have failed on the broken PCB?
Im assuming it will be something inside the red circle:
http://imgur.com/Zn7a82j
According to fzabkar's article, it could be just either the 5v or 12v TVS diode, but the drive may have additional overvoltage protection like "a fuse, polyswitch, inductor or zero-ohm resistor". I am not even sure what polyswitch's or inductor's are.

[Jared]

Just take a multimeter, set it to check resistance and read both TVS diodes. If one reads 0 ohms or near that, just remove it and it may "fix" the PCB. Though I wouldn't trust it without the TVS as it'll be more susceptible to damage from power surges that might make it un-recoverable next time.

[callum92]

I measured the two TVS diodes and they both give a reading of around 70-80 ohms so I don't think they are the problem.
There are however another two diodes further away from the power connector that give a reading of 0 when measured, could this be the problem? I have highlighted them in the picture below:

Attachments

[fzabkar]

A shorted TVS diode would have shut down the PSU. As expected, your diodes were not shorted.

Those other two diodes are the flywheel diodes in the Vcore and Vio buck regulators. They should not measure 0 ohms. That is not to say they are faulty. Rather, it suggests that the ST MCU is damaged.

If you measure 0.0 ohms (not "OL") across the corresponding capacitors (see attachment), then the Vcore and Vio supplies are shorted. Better still, measure the voltage across the capacitors when the PCB is powered up. You can remove the PCB from the drive for these tests.

Attachments

Vcore_Vio.jpg (53.84 KiB) Viewed 15469 times

[callum92]

So it seems I made a mistake measuring the flywheel diodes, only the one on the 5v supply is bad giving a reading of 0.08ohms, the one on the 12v supply reads 15.8ohms.

The 4 capacitor's on the 12v supply have a resistance of around 12.7ohms and measure 4.7v when the PCB is powered up.

The 3 capacitor's on the 5v supply all have a resistance of 0.08ohms and measure 0.08v when the PCB is powered up, also the component to the left of the flywheel diode, marked 3R3 has a resistance of 0.05ohms. So it seems everything on the 5v supply fried when the PSU died, Is it likely that the damage is limited to these 5 components?

[fzabkar]

Those are not 5V and 12V flywheel diodes. That area of the circuit generates the Vcore (+1V - +2V ?) and Vio (+3.3V) voltages for the ST MCU chip plus the Samsung SDRAM and 8-pin serial flash memory at the bottom left corner. AFAICT, both circuits would be powered from the 5V supply.

I expect that the 4-capacitor circuit would be the Vcore while the 3-capacitor circuit would be the Vio. You could verify this by testing for continuity with pin #1 of the SDRAM or pin #8 of the flash memory. Note that one of these supplies is shorted to ground. The reading of 4.7V suggests that the associated 8-pin PWM controller IC (2957 ?) is faulty (possible shorted chopper), which in turn means that any component connected to that supply may have failed.