Help on an Overvoltage Failure/Repair

[will5001]

Hi All,

Like an idiot I used a different brand of PSU cable in a modular PSU - connected up three drives and pressed "on"

The fans kicked a little and immediately no power. The three drives are all dead (no spin, not visible when connected either internally or externally)

I have three drives - 480GB SSD, 500GB WD, 4 TB WD Blue.

I want to try to recover the 4TB Drive (I might investigate the others) - the data is important but not critical. I can solder, but not well, and I can use a multimeter but I'm no expert!

I know the best option to to send to a professional (and that the data is a nice to have) but given the price of that I am thinking about a DIY solution.

The PSU is a known good PSU (and powering the drive I am using to type this) now that the right cable is being used.

I have investigated the PCB with the multimeter and read around on the forum and have a plan - but wanted to check before doing anything irreparable.

I have attached the photo of the PCB

Diagnosis

Blue I think is the 12v circuit?
Blue 1 TVS?
On the diode test it was 0.182v in one direction and 0L in the other.
Resistance - 3.75 KOhm in the 0L direction, and somewhere between 290 and 306 KOhm in the other

Blue 2- I think this is the Schottky Rectifier??
Shows similar to the TVS(?) as it is in parallel (?) - 294 kOhm in one and 3.71 kOhm in the other.

Blue 3 - sdm fuse (?)
Shows continuity (0 Ohms)

Conclusion - 12v circuit seems fine (would make sense since I don't suspect the PSU of putting out more than 12v)

Orange - 5v Circuit?
Orange 1 - TVS?
Shows 0.040v in one direction and 0.034v in the other (77 Ohm in one and 71 Ohm in the other)

Orange 2 - Schottky Rectifier??
Shows the same as the TVS

Orange 3 - sdm fuse?
No continuity (0L) - when on continuity mode the unit feels like it wants to beep very very briefly and then shows 0L

Conclusion - 5v circuit has blown fuse and furthermore would then short to ground if the fuse weren't blown?

Proposed "Fix" - remove 5v TVS chip (snip or solder?) - replace Orange 3 with a blob of solder (or conductive tape over the top or even solder over it? it's very small and my skill is D+ at best), leave orange 2 in place? or take that out too?

Risks I forsee - without a known good PSU this could create much more damage to the drive itself (if not already damaged).

Questions -
Does this diagnosis make sense?
Could this possibly work??
If it could - any comment on the method?
Is damage to the PSU possible/likely? Would using a cheap external bay to test be better or worse?
If I do this and it doesn't work - does the professional option become any more complicated?

The alternative is to find a new board (2060-800055-002 revp1) and swap the ROM over - the only ones I see on eBay are used and in China and will take forever to get here. A middle ground I between my solution and the professional one is a new board and then taking it to an electronics repair shop to swap the chips - but I can't find where to buy a board.

Attachments

[Lardman]

I can't fault your methodology.

If you've got a blown fuse on the 5v rail stands to reason the first thing to do would be to replace it and hope there wasn't damage further up the line. You're not going to kill a PSU unit either way (unless you have luck like mine ).

[will5001]

Thanks! Assume I should be taking out the TVS diode too?

It appears to be equally conductive both ways. It hasn't failed as a short though like I have seen others do which makes me slightly confused.

[Lardman]

It's about a close to a wire as you're going to get without actually being a wire. I'm not a fan of clipping them off - or just bridging a fuse either, although I know many here do it without problems. I don't feel comfortable letting things run like that, components are cheap or you can borrow them from almost any drive you have kicking around.

[fzabkar]

Remove the 5V TVS diode and measure its resistance out of circuit. Also measure the resistance at the diode pads on the PCB. If the low resistance reading stays with the diode and not the PCB, then you can be reasonably confident.

[will5001]

Thanks - I will remove it and see. I'm right in saying I can power up without a replacement (but just not have the protection)?

What should happen to the other component (number 2 in the photo) can that stay on with the diode removed or does that need to come off too?

Thank you for all the help.

[fzabkar]

Remove the diode and retest, then we can decide what to do.

[Lardman]

Best to double check and make sure you've corrected your cable issue if you're running without protection !

[will5001]

Best to double check and make sure you've corrected your cable issue if you're running without protection !

Thanks - - yes certain that is fixed. I rummaged around and found the right cable and it's successfully powering other drives.

[will5001]

Remove the diode and retest, then we can decide what to do.

Thanks for the help. Yes will find a time after work today.

[will5001]

Found some time on a break to remove the TVS. The numbers stay with the chip when removed from the board and number 2 on the board now has the same numbers as the one on the 12v side. (Only conducting in one direction)

Weirdly the sdm fuse is now showing 25 M Oh rather than open circuit.

[fzabkar]

There is a good chance that bridging the fuse will recover your data. There are more tests you can do, but you would need to show us the rest of the PCB.

[will5001]

There is a good chance that bridging the fuse will recover your data. There are more tests you can do, but you would need to show us the rest of the PCB.

Thank you so much for the help - I've bridged the fuse and the 5v pins now show continuity to the circuit beyond - no shorts to ground.

I've attached a photo of the whole board - the rest of it looks undamaged to my eye.

Attachments

[fzabkar]

Measure the resistance between ground and each of V1, V2, V3, Vx, Vy.

You could also check the preamp resistances by locating the +/-5V test points on the back of the PCB and testing for resistance to ground. Do this with the PCB on and off the HDA.

Preamp_Vcc_Vee.jpg (135.2 KiB) Viewed 15495 times

[sin]

Remove the 5V TVS diode and measure its resistance out of circuit. Also measure the resistance at the diode pads on the PCB. If the low resistance reading stays with the diode and not the PCB, then you can be reasonably confident.

Hail Thevenin

--

[will5001]

V1 - 1.59 k ohm
V2 - 0 ohm
V3 - 0.460 k ohm
Vx- 0.426 M ohm
Vy - 0.425 M ohm

Not sure I can find the right thing on the reverse I have attached the photo

Attachments

[fzabkar]

I think these are the right test points:

A reading of 0 ohms for V2 is a dead short. Is that really 0.0 ohms, or is it OL (overload)?

Are you using the screw hole as your ground reference?

[will5001]

I don't think I'm doing the right thing on the reverse - from the contact point to ground the resistance is different each measurement.

There's definitely a path to ground in the area but my multimeter probes are a bit fat for this work. I had a frustrating experience with a bit of solder wire as an extension to the probe and couldn't get a consistent reading but perhaps around 16 M ohm.

[will5001]

With the continuity tester I'm 70% sure there's a path from V2 direct to ground (either holes or the pins).

I'm guessing that's bad news.

Is it worth powering up and potentially shorting another part of the board or is there now a bigger risk of drive damage?

[fzabkar]

You can scrape away some of the area around the test points to expose the copper.