SSD went bang, is it saveable?

[sirgraystar]

rma it for a new one

data recovery is very expensive btw

Would you really expect the manufacturer to accept the RMA after he blew it up? Maybe I'm being overly cynical but I wouldn't expect that kind of damage to be covered by the warranty.

 

[Justin Credible]

A very small chance.

Sorry, but when you hear a "bang" and see sparks and smoke come from an SSD that's had too much voltage applied to it, there's little chance that the NAND has survived.

Yes, any voltage regulation circuitry might be the direct cause of the noise, sparks and smoke, but I doubt the controller IC or NAND will have survived.

I do hope for the sake of the OP that I am wrong, but I spent about 20 years of my working life repairing TVs. Delicate circuitry does not like excessive voltage being applied. Unless SSDs have been designed with protection against over voltage or reverse voltage being applied, then an incorrect power supply feed will cause damage. Sparks and smoke indicate serious damage, rather than just a fuse blowing.

If Justin could remove the SSD PCB from it's case and post a photo, it may be possible to identify what has been damaged (there will almost certainly be visible signs of what burnt out).

Yeah, sparks and smoke flew out of the connection, where the cable meets the SSD. Still smells like TCP haha.

I will get a picture of it up tomorrow so you guys can take a peak.

Thanks for the help all, much obliged!

 

[Dervious]

Yeah, sparks and smoke flew out of the connection, where the cable meets the SSD. Still smells like TCP haha.

I will get a picture of it up tomorrow so you guys can take a peak.

Thanks for the help all, much obliged!

If the incorrect power feed simply shorted across a couple of the ground pins in the SSD, then you may not have damaged any electronic components. What is likely to have happened is that the connector itself and a small part of the PCB will have been damaged. This increases the chances of retrieving your data if the damage can be repaired.

If you can remove the case from the SSD (should be fairly easy) and take a decent photo of both sides of the PCB, then it may be possible to give a better idea of what damage has been caused. Fairly closeup photos would help.

Good luck. I hope the damage is minimal.

 

[panyan]

take a decent photo of both sides of the PCB, then it may be possible to give a better idea of what damage has been caused

+1

 

[Justin Credible]

So I managed to take a few photos, seems a couple of things have burnt out. Also the middle chip (for the lack of a better word) looks a different colour to the rest.

It's quite hard to make out but at the bottom right of the Top view there's a small black spot where it's burnt out, near a chip.

Top View

Bottom View

Cheers for the help!

 

[Dervious]

First impression is that it could have been worse.

Obvious burnt out component next to the PSU connector.

Looks like the IC, just a little lower and to the left has "popped".

If it's gone no further, you might get your data back. It'll need specialist attention though.

 

[Dervious]

Looks like something else may have burnt out, bottom right.

The good news is that there are no visible signs of damaged NAND chips. However, if excessive voltage reached those memory ICs, they will have almost certainly been damaged, and I would have thought that their contents would either be badly corrupted, or become impossible to recover.

If you really need your data back, it is a job for a specialist.

Good luck.

 

[Justin Credible]

Awesome, thanks for helping! Will send it off to a specialist see what he can do.

Thanks again!

 

[Dervious]

Please let us know how you get on. I hope my theory is wrong about those NAND chips.

I know it's been said before, and is now probably obvious to you......

Always backup important data. Have at least one copy separate from the main one.

And if it's any consolation - for a lot of people it's the loss of important data that prompts them to start making backups.

 

[Rroff]

If its that important I'd highly reccomend getting a Synology or QNAP NAS for backup.

From the pics I'd say theres probably a decently high chance the storage chips themselves are intact but your looking at a fairly high cost for recovery from people equipped to do that reliably.

 

[Dervious]

From the pics I'd say theres probably a decently high chance the storage chips themselves are intact but your looking at a fairly high cost for recovery from people equipped to do that reliably.

Sorry, but I disagree.

Looks like excessive voltage has been applied (sounds like it too, from the description of how the damage was caused). Unless OCZ have put some pretty good over voltage protection on the power supply rails to the NAND chips, then the 12v on the 5v supply that damaged what I assume are regulator components may well have reached the NAND. At least there are no visible signs of damage to the NAND, but that doesn't mean that the silicon inside is still healthy.

 

[Justin Credible]

If its that important I'd highly reccomend getting a Synology or QNAP NAS for backup.

From the pics I'd say theres probably a decently high chance the storage chips themselves are intact but your looking at a fairly high cost for recovery from people equipped to do that reliably.

Anyone people you would recommend mate? Bearing in mind I'm UK based.

Cheers

 

[Rroff]

Sorry, but I disagree.

Looks like excessive voltage has been applied (sounds like it too, from the description of how the damage was caused). Unless OCZ have put some pretty good over voltage protection on the power supply rails to the NAND chips, then the 12v on the 5v supply that damaged what I assume are regulator components may well have reached the NAND. At least there are no visible signs of damage to the NAND, but that doesn't mean that the silicon inside is still healthy.

I may be wrong but it looks like the component thats exploded is the noise supression/regulator for +5v and after it there appears to be a diode for over-voltage protection. Impossible really to tell whats happened beyond the obvious but I think theres a fair chance that the NAND is intact.

EDIT: Question would be how long voltage was flowing as the diode is only designed to protect against spikes in voltage and/or whether the +5 lines were the only place that were exposed to 12v.

 

[Dervious]

I may be wrong but it looks like the component thats exploded is the noise supression/regulator for +5v and after it there appears to be a diode for over-voltage protection. Impossible really to tell whats happened beyond the obvious but I think theres a fair chance that the NAND is intact.

EDIT: Question would be how long voltage was flowing as the diode is only designed to protect against spikes in voltage and/or whether the +5 lines were the only place that were exposed to 12v.

Unless the zener diode went short circuit, and stayed short circuit before (hopefully) a fusible link went OC, then it wouldn't take long to damage the NAND chips. Perhaps this SSD had some form of protection against over voltage (I would have thought it would). If that were the case, I'd expect to see a small mark where the fusible link has burnt out, and maybe the same for the diode.

In the top photo, there appears to be three burn marks, the third looks like a pin hole has formed of the top of an IC (possibly part of the power regulating circuitry). Maybe the mark is something else, in which case I think there's more hope that the NAND survived, and that the SSD may even be repairable.

 

[Justin Credible]

I may be wrong but it looks like the component thats exploded is the noise supression/regulator for +5v and after it there appears to be a diode for over-voltage protection. Impossible really to tell whats happened beyond the obvious but I think theres a fair chance that the NAND is intact.

EDIT: Question would be how long voltage was flowing as the diode is only designed to protect against spikes in voltage and/or whether the +5 lines were the only place that were exposed to 12v.

It was exposed to the voltage for about 10 seconds, as the PC wouldn't turn off so I dived for the mains switch.

I will send it off and come back with the results.

Once again, thanks to all, much appreciated!

 

[Subliminal Aura]

//ignore

Pics in thread

 

[Diagro]

Hmmm, if someone can figure out where the traces go you could simple repair them

 

[Rroff]

Unless the zener diode went short circuit, and stayed short circuit before (hopefully) a fusible link went OC, then it wouldn't take long to damage the NAND chips. Perhaps this SSD had some form of protection against over voltage (I would have thought it would). If that were the case, I'd expect to see a small mark where the fusible link has burnt out, and maybe the same for the diode.

In the top photo, there appears to be three burn marks, the third looks like a pin hole has formed of the top of an IC (possibly part of the power regulating circuitry). Maybe the mark is something else, in which case I think there's more hope that the NAND survived, and that the SSD may even be repairable.

BTW not quite sure how my earlier posts came over but they weren't intended to contradict what you were saying and/or tell you about basic micro-electronics (which I assume from your posts you have atleast a working knowledge of). Rather to show I didn't just glance at the image and see the NAND visually intact and say that would be fine then

Hmmm, if someone can figure out where the traces go you could simple repair them

Would not reccomend that - as Dervious mentioned above the power management IC appears to have been damaged and theres likely other damage within the power supply circuitry that may result in stuff malfunctioning and the destruction of the NAND even assuming its intact currently.

From a rudimentary understanding of the power supply circuitry based on what its possible to deduct from the images and what is and what isn't showing damage my instinct is theres a fair chance that damage is contained to the power supply components but theres a pretty big chance the NAND has been damaged also.

 

[Scougar]

If these tracks are worth money (more than the cost of a potential recovery which is likely to be £400+ I would imagine), then send it away to see what they can do with it.

All the best, and please follow up with the outcome

 

[panyan]

if the drive could be repaired im guessing another drive of the exact same model needs to be purchased and the Data chips need to be unsoldered and then swapped between them - then the drive can be accessed and the data retrieved?

intreguing