[SOLVED] FOUR of my 8TB Western Digital HDs just had their PCBs fry!

[help_me_please23]

I am sick to my stomach over this. I was just building a new Windows 10 PC with a ton of hard drives for the sole reason of not just a gaming machine, but one where I would store all of my data and I was going to begin a backup process tomorrow. This is the worst timing/luck ever!

Videos of my two 5 year old daughters and my two three year old sons are gone, as well as some VHS tapes I ripped that had the only remaining videos of my sister who passed away not long ago (I was literally planning on backing these all up TOMORROW). I have photos backed up to Amazon's cloud but they don't accept videos so those are all gone.

Here is what happened: I had a desktop gaming PC that had 8 hard drives in it. The CPU (Intel 9700k) processor I installed last year died, so I replaced it yesterday. I also bought a new PSU because I had 3 motherboards die (plus the CPU) in the span of a year, so I figured my PSU was somehow frying my hardware. So I installed the new CPU and PSU yesterday and my PC finally booted up for the first time in a while.

However, I noticed only 4 of my hard drives showed up in Windows. I went into my BIOS, and had the same thing. Not recognized.

I took the 4 problematic drives out of my case and tried to power them up in a friend's PC and had the same issue. Zero power. I also tried an external USB hard drive caddy (one that uses wall power of course since these are all 3.5" drives) and they simply aren't receiving power.

I believe somehow my old power supply fried them just as my CPU went down. Either that or the APC UPS battery backup device I have is somehow frying the insides of my computer. I have already ordered a new APC UPS battery backup device just because of how paranoid I am as to what could have caused this. It's definitely not the new PSU as I have tested the voltages on each SATA line and they are all within spec.

Here are the drives that died:

Drive 1 front:

https://imgur.com/aiChtaR

View: https://i.imgur.com/aiChtaR.jpg

Drive 1 back:

https://imgur.com/B4xVcE4

View: https://i.imgur.com/B4xVcE4.jpg


Drive 2 front:

https://imgur.com/fLHZHaB

View: https://i.imgur.com/fLHZHaB.jpg

Drive 2 back:

https://imgur.com/zAMH6ss

View: https://i.imgur.com/zAMH6ss.jpg


Drive 3 front:

https://imgur.com/vEOIdp7

View: https://i.imgur.com/vEOIdp7.jpg

Drive 3 back:

https://imgur.com/GnD6c8j

View: https://i.imgur.com/GnD6c8j.jpg


Drive 4 front:

https://imgur.com/1SlUbpe

View: https://i.imgur.com/1SlUbpe.jpg

Drive 4 back:

https://imgur.com/HS5O8KI

View: https://i.imgur.com/HS5O8KI.jpg


These are all 8TB Western Digital drives that came from Western Digital Elements/Essentials/MyBook USB enclosures that I shucked and took the drives out of, meaning they have no warranty. But it's not the $120 I spent on each drive that I care about. It's getting the data off of them. Also please note this is not the "3.3V pin issue" with the drives not receiving power; I am aware of that and this is unrelated. They were definitely fried by my old crappy PSU.

Now normally if I had dropped the drives or they were giving me the click of death I would just break down and cry for a few days because I don't have the thousands of dollars it would cost to "revive" these drives. But it's the fact that they simply aren't receiving power that gives me hope.

I took the PCB from one of the 8TB drives that is still good and swapped it onto one of the drives that do not receive power. To my delight, the drive began to spin! No clicks of death or anything. However, Windows 10 could not recognize the drive. I am guessing this is because the PCB wasn't a 100% exact match (it was pretty darn close from what I could see in regards to the little numbers on the back).

Is there some way for me to get identical PCB replacements for these 4 drives? If so, does it require a BIOS swap like I know Seagate drives require? I am pretty handy with a soldering iron but that's about it. I am not sure I would be able to desolder any of the extremely tiny chips I saw on the PCB when I tried the swap.

Any advice would be extremely appreciated. I haven't eaten in over a day because of how sick to my stomach I am over this and am praying these WD PCBs aren't "locked down" to a drive or something and if they are, there is a reasonable way to get them powered up again.

Thank you so much in advance.

 

[Negative Circumference]

Your data is very likely safe, so do not fret, the issue comes down to the platters now, those are hard to disrupt without physical tampering. You can buy identical drives, and swap the boards. There are also companies that will recover the data on those platters for a price. As long as the drives are identical, correct me if I'm wrong, but you can buy an identical drive and resolder the pcb connections to a new board You might need firmware and such, so it won't be an easy task.

I'm really just posting to say to keep your head up, if groups can recover data from drives lost in a fire, it should not be too difficult.

This site has some good documentation.

 

[help_me_please23]

Your data is very likely safe, so do not fret, the issue comes down to the platters now, those are hard to disrupt without physical tampering. You can buy identical drives, and swap the boards. There are also companies that will recover the data on those platters for a price. As long as the drives are identical, correct me if I'm wrong, but you can buy an identical drive and resolder the pcb connections to a new board You might need firmware and such, so it won't be an easy task.

I'm really just posting to say to keep your head up, if groups can recover data from drives lost in a fire, it should not be too difficult.

This site has some good documentation.

Thanks so much for the reply.

The PCB is very easy to swap. Just some torx screws and it's done.

I think unfortunately Western Digital drives do in fact have a BIOS that is tied to the PCB. So even if I somehow find the exact same replacement PCB, the BIOS will not let the drive be readable.

Supposedly this chip is very difficult to swap as you can't just get a desoldering iron and do it that way. I am also reading something about reprogramming a donor board but I am not familiar with any of that, nor whether or not it even works with Western Digital drives 🙁

 

[bryanl]

The best case failure would be limited to ruined TVS diodes and the accompanying blown fuses, and replacing both components should restore operation. However a different power supply should be used since it may have put out excessive voltage. Do not do as some people and merely replace the fuses without also replacing the TVS diodes since that can allow much more damage. Any electronics technician can perform the repair with common soldering equipment. http://www.users.on.net/~fzabkar/HDD/TVS_diode_FAQ.html

Circuit board replacement requires transplanting the BIOS chip to the new board., and a company named One PCB Solution offers that service along with board purchase. The soldering of that chip is not difficult if its pins are visible but otherwise requires special soldering equipment.

 

[fzabkar]

This person has posted the same tale of woe to at least two other data recovery forums. I am still waiting for photos of the component sides of each PCB.

There are several photos in this [long] thread:
http://www.hddoracle.com/viewtopic.php?f=21&t=2549

FYI, these drives are rebadged HGST helium models.

 

[help_me_please23]

The best case failure would be limited to ruined TVS diodes and the accompanying blown fuses, and replacing both components should restore operation. However a different power supply should be used since it may have put out excessive voltage. Do not do as some people and merely replace the fuses without also replacing the TVS diodes since that can allow much more damage. Any electronics technician can perform the repair with common soldering equipment. http://www.users.on.net/~fzabkar/HDD/TVS_diode_FAQ.html

Circuit board replacement requires transplanting the BIOS chip to the new board., and a company named One PCB Solution offers that service along with board purchase. The soldering of that chip is not difficult if its pins are visible but otherwise requires special soldering equipment.

Thank you!

I would be willing to try to solder in some diodes and/or fuses but are there photos on exactly how it would be done on these specific hard drives? That guide is 100% text which makes me a little leary.

This person has posted the same tale of woe to at least two other data recovery forums. I am still waiting for photos of the component sides of each PCB.

There are several photos in this [long] thread:
http://www.hddoracle.com/viewtopic.php?f=21&t=2549

FYI, these drives are rebadged HGST helium models.

I posted my reply there yesterday and the moderators there still haven't approved my replies. So I came here.

Here are photos of the other side:

https://imgur.com/a/eIfQ4nL

View: https://imgur.com/a/eIfQ4nL


Let me know if you need me to take better pics.

Thank you

 

[mdd1963]

When you swapped PSUs, did you reuse old modular cables from old PSU inserted into the new PSU to power the new PSU?

If so..that's what happened.... 🙁

 

[fzabkar]

These are the TVS diodes and fuses:

Test them with a multimeter, as explained in my TVS Diode FAQ.

 

[help_me_please23]

When you swapped PSUs, did you reuse old modular cables from old PSU inserted into the new PSU to power the new PSU?

If so..that's what happened.... 🙁

Nope, neither of my PSUs are modular. It's just that my old 2009 Antec Earthwatts PSU was apparently frying every damn thing in my desktop.

These are the TVS diodes and fuses:

Test them with a multimeter, as explained in my TVS Diode FAQ.

I apologize if this is a stupid question but do I need to have power going to the SATA port of the PCB while I test them?

 

[fzabkar]

Test them in the absence of power. It should be explained in my FAQ.

 

[help_me_please23]

These are the TVS diodes and fuses:

Test them with a multimeter, as explained in my TVS Diode FAQ.

OK thanks so much!

I got my multimeter and turned the dial to 200 in the "ohm" section.

For the 12V and 5V TVS diodes, the results were the same. If I have the black probe on the left side of the diode and the red probe on the right, I get a number anywhere from 120-190 but then it quickly goes down to 1 and stays there. If I put the black probe on the right side of the diode and the red probe on the left, I just get 1 no matter what.

For the fuses using the same multimeter setting as I used for the diodes I get the following: fuse on the left; 0.03. For the fuse on the right, nothing (1).

Please let me know if this helps, or if I did something wrong on the tests.

Thank you again so much.

 

[fzabkar]

The TVS diodes appear to be OK. However, the 5V fuse (on the right) is open circuit.

http://www.hddoracle.com/download/file.php?id=8606&mode=view

This suggests that there was an overvoltage on the 5V supply. I don't know why the TVS diode survived -- perhaps the fuse is a very fast acting type, or the diode is more robust than others, or the overvoltage was not excessive.

You could try to purchase a 2A fast acting smt fuse, or you could wire a 2A picofuse in place of the smt fuse. If the diode were shorted, I would have risked flowing a blob of solder over the fuse, but in the present circumstances I can't say if anything else was damaged.

http://www.farnell.com/datasheets/48294.pdf

If you need to replace the PCB, you will need to transfer the 8-pin serial flash memory IC (to the right of the SATA data connector). However, you must be careful not to damage this IC as it contains unique info. Hdd-parts.com includes a free firmware transfer service with each PCB.

 

[help_me_please23]

The TVS diodes appear to be OK. However, the 5V fuse (on the right) is open circuit.

http://www.hddoracle.com/download/file.php?id=8606&mode=view

This suggests that there was an overvoltage on the 5V supply. I don't know why the TVS diode survived -- perhaps the fuse is a very fast acting type, or the diode is more robust than others, or the overvoltage was not excessive.

You could try to purchase a 2A fast acting smt fuse, or you could wire a 2A picofuse in place of the smt fuse. If the diode were shorted, I would have risked flowing a blob of solder over the fuse, but in the present circumstances I can't say if anything else was damaged.

http://www.farnell.com/datasheets/48294.pdf

If you need to replace the PCB, you will need to transfer the 8-pin serial flash memory IC (to the right of the SATA data connector). However, you must be careful not to damage this IC as it contains unique info. Hdd-parts.com includes a free firmware transfer service with each PCB.

Thank you so much for all your help from the bottom of my heart.

Is this OK to buy for this? https://www.ebay.com/itm/5pcs-2A-Littelfuse-Pico-Fuse-2-Amp-125V-Littlefuse-5x-Fast-US-Shipping/254179618210?epid=1504535694

I do now notice I have a 5th hard drive that suffered from this bad PSU. The difference with this one is that it does in fact receive power. However, as soon as it is connected to a computer (Windows or Linux, it doesn't matter), it crashes the computer (blue screen in Windows with the error "SYSTEM_THREAD_EXCEPTION_NOT_HANDLED". )

I checked the two fuses on that one and both are reading 0.03.

Is there anything else I can try with this 5th hard drive? Any readings I can give you, or tips to get it to not crash the computer when it's plugged in? It has to be related to the bad power supply as it happened the same time as the other four drives.

 

[fzabkar]

The picofuse should be OK, although you will need to be careful how you mount it and insulate its leads. An smt fuse would be better, but more difficult to handle.

The risk with an overvoltage on the 5V supply is that the preamp on the headstack inside the drive may have been damaged. My concern is that the preamps in drives 1 to 4 may be shorted, in which case that would explain the open fuses. If that is the case, then game over. :-(

Does the 5th HDD identify itself correctly to BIOS? That is, does it report its correct model number and capacity? Does it make any clicking sounds? Does it spin down after clicking?

After replacing the fuse, I would advise you to test each board off the drive before installing it. In the meantime, if you can upload a good photo of the PCB from the 5th drive, I could help you to identify its voltage test points. These will serve as a reference for PCBs 1 to 4.

 

[help_me_please23]

The picofuse should be OK, although you will need to be careful how you mount it and insulate its leads. An smt fuse would be better, but more difficult to handle.

The risk with an overvoltage on the 5V supply is that the preamp on the headstack inside the drive may have been damaged. My concern is that the preamps in drives 1 to 4 may be shorted, in which case that would explain the open fuses. If that is the case, then game over. :-(

Does the 5th HDD identify itself correctly to BIOS? That is, does it report its correct model number and capacity? Does it make any clicking sounds? Does it spin down after clicking?

After replacing the fuse, I would advise you to test each board off the drive before installing it. In the meantime, if you can upload a good photo of the PCB from the 5th drive, I could help you to identify its voltage test points. These will serve as a reference for PCBs 1 to 4.

Thank you again so much for the reply.

The 5th hard drive does indeed identify itself correctly to the BIOS. For the few seconds it shows up in Windows before the BSoD, it even shows up as the drive letter I assigned to it back when I first got it. It's just that as soon as anything happens with the drive, the computer blue screens. It does not make any clicking sounds and it does not spin down.

As you requested here is a high res picture of the 5th PCB:

https://imgur.com/vFTKlmC

View: https://i.imgur.com/vFTKlmC.jpg


Thank you so, so much

 

[help_me_please23]

By the way, I got the DOS version of Western Digital Diagnostic Tools (this was not easy to find!) and I whipped out my floppy drive and everything. The DOS WD tool saw the hard drive just fine, even reporting its serial number and firmware version. I set it to scan the disk for errors, and it found none!

I am still humbly begging for your help in getting the drive to stop crashing any time it is plugged into Windows/Linux. I know it's related to the issue with the 4 other drives.

 

[fzabkar]

Sorry for the delay.

I will be uploading some photo clips later today. I am proceeding cautiously because these drives are relatively new and have some chips whose function I don't fully understand (they appear to be custom made for HGST/WD). You have two other active threads in other forums, but I have decided to concentrate on this one.

Your observation in respect of WD's DOS tool (Data LifeGuard ?) is puzzling, but it suggests that the drive can be cloned.

Edit:

Here are the photo clips:
http://www.users.on.net/~fzabkar/HDD/0A90701/

Could you please use your good PCB to measure the indicated voltages at the V1, V2, etc test points, plus the unlabeled test points in the marked squares? Use a screw hole as the ground reference (black probe). Please be careful not to slip with your probes, as these tests need to be done on a powered PCB.

Could you identify the markings on IC1? If you could post these markings for all 5 PCBs, this may help me to identify this chip. I believe it is at the front end the power supply chain.

 

[help_me_please23]

Sorry for the delay.

I will be uploading some photo clips later today. I am proceeding cautiously because these drives are relatively new and have some chips whose function I don't fully understand (they appear to be custom made for HGST/WD). You have two other active threads in other forums, but I have decided to concentrate on this one.

Your observation in respect of WD's DOS tool (Data LifeGuard ?) is puzzling, but it suggests that the drive can be cloned.

Edit:

Here are the photo clips:
http://www.users.on.net/~fzabkar/HDD/0A90701/

Could you please use your good PCB to measure the indicated voltages at the V1, V2, etc test points, plus the unlabeled test points in the marked squares? Use a screw hole as the ground reference (black probe). Please be careful not to slip with your probes, as these tests need to be done on a powered PCB.

Could you identify the markings on IC1? If you could post these markings for all 5 PCBs, this may help me to identify this chip. I believe it is at the front end the power supply chain.

Thank you as always for the reply.

I wasn't able to test my good board. I plugged it into the shucked enclosure's USB board. I plugged in the power and even the USB into my laptop. But no matter what I tested it all came back 0 on my Meterman 5XL. Let me know if I did something wrong.

As for the IC1 markings, here they are:

878JEC
7140HC
ST



87804C
8230N2
ST



8780GC
9180MG
ST



87807C
72603F
ST

Please note that the "ST" on the bottoms are written in a completely different font and appear to be a brand name marking rather than part of the model number.

 

[fzabkar]

"ST" is the logo of STMicroelectronics.

714, 823, 918 and 726 appear to be YWW (Year/Week) date codes. So the chips were manufactured during week 14 of 2017, week 23 of 2018, etc. The important marking which probably relates to the part number appears to be 878. However, I still don't recognise the actual chip. :-(

As for the enclosure, it has a dual MOSFET switch IC which switches the +5V and +12V power to the HDD. This chip is controlled by the USB-SATA bridge IC on the bridge PCB. I suspect that the bridge firmware turns off the SATA power if it doesn't detect a host computer at the USB port. It probably also switches off the power if it doesn't detect a hard drive after a short delay.

You may be able to trick the bridge IC into switching on the power if you short a certain pin to ground while switching on the adapter, and then releasing the short soon after. However, I would need to see a photo of the bridge PCB.

 

[help_me_please23]

I guess I can worry about that later. For the 4 dead PCBs, all I have to do is wire that picofuse to the dead fuses?

 

[fzabkar]

ISTM that you are at the limits of your ability, so you are basically on your own now. I would have preferred to do more testing to determine if there is more damage, but hopefully a replacement fuse is all you will need. Your choice ...

Just remember that you can swap the entire PCB provided that you transfer the serial flash memory IC ("ROM").

 

[help_me_please23]

ISTM that you are at the limits of your ability, so you are basically on your own now. I would have preferred to do more testing to determine if there is more damage, but hopefully a replacement fuse is all you will need. Your choice ...

Just remember that you can swap the entire PCB provided that you transfer the serial flash memory IC ("ROM").

Do you provide that service? I'd be more than happy to pay you to do it.

 

[fzabkar]

I've been retired for a long time. My eyesight and coordination are no longer good enough. Sorry.

 

[help_me_please23]

I just said "screw it" and bridged the two points of the blown fuse and all 4 of my drives powered up and are working again

I'm not going to use them as regular drives; however I was able to get all the data off of them and onto new hard drives. I WILL be using them as "backup" drives from now on though.

The only issue I have left is the 5th drive that crashes Windows every time it is plugged in. Hoping I can recover the data from that somehow.

 

[fzabkar]

Try to clone the 5th drive with HDDSuperClone or ddrescue. Both are Linux tools and can resume after an interruption.

Congratulations, anyway. It seems I was being overly cautious.