Request for Help with Overvolted WD Reds

[jkotzker]

Hey all, just made an account here to ask this question, hoping that’s okay . I am looking for some help.

I have 4 6TB WD Red drives in an enclosure that I’ve been using as a ZFS pool in RAIDZ1, the ZFS equivalent of RAID5. Well, I was recently redoing some cabling and configuration of my homelab, and wouldn’t you know it, I accidentally plugged in the wrong cable and sent 24V right through the enclosure instead of 12V. Now 2 of the drives won’t spin up, the other two come online fine and I can see them connected, but these two drives make a single noise that sounds like an attempt to start spinning and then they cut out, and nothing I do will make them spin up or register in any enclosure or with any cable. Since this is a RAID5 configuration and I’ve lost 2 drives, unless I can get at least one back I’m SOL.

I’ve been doing a lot of searching and reading the last couple of days since this happened, and I’ve come across a lot of threads here and other places on the web which mostly feature the user @fzabkar helping people figure out how to fix or bypass the shorted components of their drive PCBs after making similar mistakes to my own. Unfortunately, my PCB looks different from any of the photos I’ve seen in those threads, and I lack the electrical knowledge to figure out on my own which components I should be testing and how I might fix these drives. I’ve brought the drives to two separate local computer repair shops, where technicians at both have told me PCB damage is not something they deal with. One even took the drives and ran them through their diagnostics just in case I had misdiagnosed the issue, and they weren’t able to get the drives to spin up either.

As far as I can tell, y’all here are my last hope. Especially @fzabkar, if you’re reading this, since from my Googling you seem like you might be the only person on the Internet who actually understands these drives. I don’t need to “fix” them fully, if I can get even one to spin up and read then I can bring my RAID back online and swap out the drives one by one for new undamaged ones. Any help at all you can provide would be much appreciated, and if I can provide any more information please ask and I will. I’ve got a digital multimeter around here somewhere I can take readings with. I tried to include a ton of photos of the drives and PCBs, but apparently new users can only include two images per post, so I’m also including a link to a public Google Drive folder with many more images of both drives and PCBs, I hope that is okay. As far as I can tell the drives are identical, and there is no difference in behavior between the two either.

Thank you in advance to anyone for any help at all you are able to provide!

>>>More Images (Google Drive link)<<<

[fzabkar]

Measure the resistances of the fuses and TVS diodes.

TVS Diode FAQ:

http://www.hddoracle.com/viewtopic.php?f=100&t=86

[jkotzker]

@fzabkar Thank you for replying! I have measured the components you marked as best I could, please tell me if these do not make sense though and I can always try again:

First Drive:

5V TVS Diode: 192.6 kΩ one way, 3.72 kΩ other way
12V TVS Diode: 1.7Ω symmetric
4A fuse closer to 5V TVS Diode: reads as 0Ω either way
4A fuse closer to 12V TVS Diode: reads as 0.L either way

Second drive:

5V TVS Diode: 189.5 kΩ one way, 3.68 kΩ other way
12V TVS Diode: 3.2Ω symmetric
4A fuse closer to 5V TVS Diode: 0.03Ω either way
4A fuse closer to 12V TVS Diode: reads as 0.L either way

[fzabkar]

In both cases the 12V diode is shorted and the 4A fuse is open. You can replace the diode with an SMBJ12A or SMCJ12A (check the dimensions). You can use a 4A smt fuse from Littelfuse and others.

First remove the diode (clip it with flush cutters if you don’t have an iron), then retest the diode pads for shorts. If the short has been cleared, then you could flow a blob of solder over the fuse. The risk is your choice.

[jkotzker]

@fzabkar, I really appreciate the help, I am returning to this now that I have some more time to invest in trying to recover these. When you say the 4A fuse is open, I assume you mean the one closer to the 12V diode that reads as 0.L?

Assuming I want to do the quickest fix possible just to be able to recover the data and replace the drive, could you walk me through what that would look like? I’m pretty much a noob here at this level of hardware. As I understand it:

1. Clip the shorted 12V TVS diode out with flush cutters
2. Retest the diode pads’ resistance with the multimeter and make sure the short is cleared (what should the readings be to make sure of this?)
3. Flow a blob of solder over the fuse (with or without removing the fuse?)

Then, theoretically the drive should spin up and allow me to read it and copy off the data, albeit now with no overvolt protection, is that correct?

Once again, thanks for any help!

[jkotzker]

@fzabkar gentle nudge, couple of outstanding questions when you get the chance. Or anyone else…

[fzabkar]

Sorry, I completely forgot about this thread. Yes, your approach is good. Bridging the fuse is a risk, so it’s your choice. If you use a current-limited supply, that will minimise the risk. It all depends on what equipment you have and your soldering ability.