EGVA power supply pin layout change - featuring fried hard drives. Beware when warrantying a PSU.

[sgircys]

After an extremely frustrating day of troubleshooting, I figured I'd share my story on here as a word of caution to anyone else out there who might be in the same, rare situation that I found myself in yesterday.

 

For a bit of backstory - I built a new PC about a year ago which included a new EVGA GQ 1000w Gold power supply. Unfortunately, as soon as I booted up the PC for the first time, I knew there was something wrong with the power supply. The coil whine was horrible; worse than anything I had ever heard from any other PC in the past. I sent it (at my own expense) to EVGA under warranty as it was brand new. As per their instructions, I sent only the power supply unit itself and no cables. They were very clear in their instructions - "Keep all accessories as you will only be receiving a power supply in return." No problem. I set the aside for when I would get the power supply back from them. In the meantime, I re-used my older Corsair power supply as it got the job done. A few weeks later, I received the RMA'd power supply from EVGA, but life got in the way so it sat in the box until yesterday, when my Corsair unit started getting noisy enough to really bug me.

 

I pulled the Corsair out, along with all of its cables as I am very aware you cannot mix power supply cables. Then I opened up the EVGA box and grabbed the cables that go along with it, which I had set aside and labelled previously. I plugged everything in and tried to boot up the PC with no luck. Only a click, which I figured might be an overload protection circuit. I immediately had to double check to see if I mixed any cables somehow, but everything was correct and only the EVGA GQ cables (that came with the power supply) were used. As the first step of troubleshooting, I disconnected the SATA power from my SATA hard drives. And just like that, it booted up completely fine. Once I had isolated that the SATA power was the issue, I decided to check the voltages with a multimeter. To my surprise, they were all completely wrong. 12V where 3V was supposed to be, nothing where 5V was supposed to be, and so on. I tried a different SATA power cable from the same, matching set and it was the exact same.

 

At this point, I called EVGA. To their credit, I was able to speak with someone in a matter of minutes, which can't be said for most manufacturers. After explaining the situation, and the tech pulling up my RMA file, he knew what the problem was. He notified me that "At some point, the pin layout of these power supplies was changed". I was never told this when I received my power supply back from warranty, and clearly my cables were incompatible with the power supply now - with no way of knowing other than by checking with a multimeter. The tech told me that he believed it was only the SATA power that was changed, which would make sense as my PC was able to boot just fine with the SATA power disconnected from the drives. He said he was sending me a new set of cables and that would fix the issue. While that should be the case, what a horrible decision to change a power supply pin layout within the same product (with the only way to know being manufacturing date?) with absolutely no notice. And by following EVGA's protocol of not sending in power supply cables during a warranty claim, you're essentially screwed. I thanked the technician for his help and acknowledged that it wasn't his fault, personally, that this happened and that I'll wait for the new cables to arrive - once again using my old Corsair in the meantime.

 

After removing the EVGA and putting the Corsair back in, once again, the problem really showed itself. All of my SATA drives were gone. They were fried. 22TB of storage gone. I double and triple checked, using a different PC as the test PC with the drives even, but they were dead. Thankfully, I do have cloud backups, but my wife and I did both lose our entire day's work as the most recent backup was from the morning. I did contact EVGA again and spoke to another technician who said he will be speaking to his manager about this tomorrow to see what they can do about this situation. As other people have said, EVGA's customer service is quite good and I do appreciate that. Hopefully they're able to help me by fixing my situation, but this could still be a serious problem for other people.

 

TLDR : EVGA decided to change their SATA power cable pin layout on the GQ power supply and you'd have no way of knowing without checking the pins with a multimeter. And they can, and do fry hard drives.

[emosun]

You can probably fix the drives by changing out the controller boards on them seeing as the actual data on the platters is OK and just the boards would have been fried.

[sgircys]

1 minute ago, emosun said:

You can probably fix the drives by changing out the controller boards on them seeing as the actual data on the platters is OK and just the boards would have been fried.

Yeah, I've been looking into that. I'm able to find a controller board for one of the two drives, but I can't seem to find the one for the other drive anywhere. I'm going to try contacting DonorDrives directly and seeing if they have a lead on a donor board.

[jaslion]

Yeah you definitely have a case here as they have written guidelines on the return process with the intention of you reusing the cables. If the psu you get back is not compatible with the older cables you will get new cables (when it is no longer in stock or something and you get a new unit) they will always ask you to dispose of the cables from what I've read over the years.

 

So this is for sure a big communication failure and also a MASSIVE issue for people that want to buy some extra cables as these psu's are now no longer compatible with their own cables. Imagine asking to buy a new sata cable from evga and they just send you the original but you now have a new model. Poof everything gone and you did nothing wrong.

 

I'd follow this up closely and press for some damages as this is just flat out not on you at all.

[jaslion]

3 minutes ago, sgircys said:

Yeah, I've been looking into that. I'm able to find a controller board for one of the two drives, but I can't seem to find the one for the other drive anywhere. I'm going to try contacting DonorDrives directly and seeing if they have a lead on a donor board.

Btw I only recommend this for getting data back. Once the boards are replaced on drives that worked fine I would no longer consider them safe to use for critical data as they tend to be more fragile after a boardswap. Sometimes it will be like nothing happened but other times you just bought a bit more time.

[sgircys]

2 minutes ago, jaslion said:

Yeah you definitely have a case here as they have written guidelines on the return process with the intention of you reusing the cables. If the psu you get back is not compatible with the older cables you will get new cables (when it is no longer in stock or something and you get a new unit) they will always ask you to dispose of the cables from what I've read over the years.

 

So this is for sure a big communication failure and also a MASSIVE issue for people that want to buy some extra cables as these psu's are now no longer compatible with their own cables. Imagine asking to buy a new sata cable from evga and they just send you the original but you now have a new model. Poof everything gone and you did nothing wrong.

 

I'd follow this up closely and press for some damages as this is just flat out not on you at all.

Yeah, it's a huge problem and they really need to address this so it doesn't happen again. Anyone who purchased a power supply (at least from this series) between 2020 and 2022 (I'm estimating) and has to RMA it, will be sent a new power supply which a changed pin layout and it will fry their SATA drives. Their process needs to be immediately be changed to include sending the cables back. And I'd even say that they should make a branding distinction on the new power supplies and cables so customers at least know there is a backwards compatibility issue going forward.

[jaslion]

12 minutes ago, sgircys said:

Yeah, it's a huge problem and they really need to address this so it doesn't happen again. Anyone who purchased a power supply (at least from this series) between 2020 and 2022 (I'm estimating) and has to RMA it, will be sent a new power supply which a changed pin layout and it will fry their SATA drives. Their process needs to be immediately be changed to include sending the cables back. And I'd even say that they should make a branding distinction on the new power supplies and cables so customers at least know there is a backwards compatibility issue going forward.

Yeah pretty much like what corsair did with type 3 and type 4. They had overlap but made a simple chart stating that psu's with this name and date (or color label) will have these cables be compatible between 3 and 4 and then these cables NOT be compatible. With a warning in box during the switchover.

 

I even have personal experience with them honoring the warranty of a old ax1200 that broke and sending back an ax1600i as the ax1200 and 1200i was not available at that time + there was premium extended warranty on it which they then would transfer to normal warranty if accepting the better unit.

 

Got a BIG BIG BIG warning IN the box ON the box IN the shipping box and STICKERED on the plastic wrapping that the cabling is different and please follow the guidelines on the provided link. As well as in email form. Because simply put you wouldn't have known. Probably would have noticed that the plug wouldn't go in as they switched the square and rounded holes around a bit but still a bit of power and boom it would have plugged in.

 

 

12 minutes ago, sgircys said:

backwards compatibility issue going forward

 

And forwards. Because any older cable kits or replacement sets simply won't be labelled that they aren't compatible with the new version.

 

 

[OddOod]

Welcome to the forums!
Holy heck. That's BAD. Like, they should pay for professional data recovery and replacement drives bad. I love EVGA and hope they survive long term (fantastic customer support in my experience) but wow, not okay. 
Thanks for the write up!

[sgircys]

14 minutes ago, jaslion said:

Got a BIG BIG BIG warning IN the box ON the box IN the shipping box and STICKERED on the plastic wrapping that the cabling is different and please follow the guidelines on the provided link. As well as in email form. Because simply put you wouldn't have known. Probably would have noticed that the plug wouldn't go in as they switched the square and rounded holes around a bit but still a bit of power and boom it would have plugged in.

Yeah, this is exactly what I said when telling other people this story so far. It should be made EXTREMELY clear that they are incompatible and will cause damage. Big stickers in, and on the box. Ideally, they power supply end up the cables wouldn't even fit in the new power supply - use a different plug shape completely. But if you cant change the plug shape, at least have a graphic or little tag at the end of the cable indicating which version of the power supply it is for. But at that point, they'd need to make a distinction between the two, rather than not even changing the part number.

[sgircys]

14 minutes ago, OddOod said:

Welcome to the forums!
Holy heck. That's BAD. Like, they should pay for professional data recovery and replacement drives bad. I love EVGA and hope they survive long term (fantastic customer support in my experience) but wow, not okay. 
Thanks for the write up!

Thanks for the welcome!

 

And I'll be sure to update the post with what I hear back from EVGA. I'm hoping the traditionally good support is able to make things right.

[sgircys]

If anyone has a way to pass this on to the LTT team, in case its something they're interested in covering, I'd be more than happy to assist. 

[sgircys]

Here is the latest update :

 

 

After waiting to hear back from EVGA all day Tuesday, I followed up via email this morning asking what the status was with this issue. I was told that their recommendation was to contact the hard drive manufacturers and try to make a warranty claim there. Unfortunately one hard drive is out of warranty and the other hard drive may be eligible for a warranty claim - but they are both out with a third party data recovery service currently having the controller boards replaced.

I wasn't particularly happy with their "solution" as it seemed like they simply wanted to wash their hands of the situation. My reply to them outlined how this was impractical as I would need to buy new drives to migrate the data to (the data recovery company told me that they recommend not using these drives after they are repaired - only use them to migrate off the data), at an upfront cost to me. Additionally, I am having to pay for the data recovery service, shipping the drives, not to mention all of the lost time and productivity spent troubleshooting this problem.

 

EVGA replied that they "recommend checking on the warranty option first" on the hard drives, and the following:

As for the data recovery and needing other drives first to save the data to, additional costs beyond the drives that were potentially affected would be outside the scope of assistance we can provide. I’ve never encountered a warranty that offers to cover loss of data or the costs related to the recovery of data, and to the letter of our warranty terms, we technically don’t cover any loss or damages incurred by our products either

So all that to say, I'm not exactly happy with how this is being handled, given that this matter is entirely the fault of EVGA and a serious mistake.

 

I'll continue to update as this progresses..

[seon123]

On 3/19/2024 at 3:55 PM, jaslion said:

Yeah pretty much like what corsair did with type 3 and type 4. They had overlap but made a simple chart stating that psu's with this name and date (or color label) will have these cables be compatible between 3 and 4 and then these cables NOT be compatible. With a warning in box during the switchover.

Type 3 and Type 4 cables are actually compatible. The only difference in pinout between Type 3 and Type 4 is in the 24 pin, which has 4 extra pins on Type 4 for Vsense. But you can shove a Type 3 pin cable into a Type 4 PSU, have four empty pins at the side of the connector, bend the retention clip, and everything will be just fine. Corsair did have other model-specific cables, which were not compatible, but Type 3/4 specifically was never an issue.

On 3/19/2024 at 3:24 PM, sgircys said:

TLDR : EVGA decided to change their SATA power cable pin layout on the GQ power supply and you'd have no way of knowing without checking the pins with a multimeter. And they can, and do fry hard drives.

When I started reading, I thought it was another case of a user reusing incompatible cables and frying their HDDs. But that doesn't seem to be the case here. I thought EVGA had mostly sorted out their PSU pinout shenanigans, but apparently not. EVGA's page for modular cable compatibility says the GQ isn't compatible with any of their individually sleeved replacement cables, but I assumed they just didn't make them. If EVGA really changed their pinouts, custom cable makers like @CableMod might want to verify that they also aren't sending out cables that fry HDDs.

 

There is still the chance that you accidentally mixed up and accidentally reused some modular cables, and that the EVGA tech is also confused (EVGA changing pinouts on newer model PSUs vs EVGA changing pinouts on older vs newer GQs). But assuming that isn't it, yikes.

[sgircys]

On 3/21/2024 at 5:08 PM, seon123 said:

If EVGA really changed their pinouts, custom cable makers like @CableMod might want to verify that they also aren't sending out cables that fry HDDs.

 

There is still the chance that you accidentally mixed up and accidentally reused some modular cables, and that the EVGA tech is also confused (EVGA changing pinouts on newer model PSUs vs EVGA changing pinouts on older vs newer GQs). But assuming that isn't it, yikes.

I've actually been in contact with CableMod on reddit and I'm hoping we can prevent this from happening to any of their customers.

 

And I'm positive that cables weren't mixed up. They're labelled according, to match the power supply. This is the only EVGA power supply that I own and the cables (and power supply, obviously) are labelled "EVGA GQ".

[Ubersonic]

It's worth noting that EVGA don't actually make PSUs.

 

Their high end models are made by Superflower (hence being amazing), their second tier are made by Seasonic and their lower models are made by FSP and HEC.

 

I believe the GQ 1KW is an FSP model, but they may have switched it to HEC which would explain a change of pinout but it's crazy they didn't catch that before sending it out to you like what were they thinking O.o

[JosephKnowsNot]

On 3/19/2024 at 9:24 AM, sgircys said:

They were very clear in their instructions - "Keep all accessories as you will only be receiving a power supply in return." No problem.

Honestly, that would be a problem for me. I would expect them to check to make sure the cables were not damaged, not just the PSU itself. I don't consider them an accessory. In my mind, an accessory would be a decorative cover that snaps on or something, cable ties, etc. not electrical components. Actually, even a mounting plate could potentially cause a problem.

 

I really think PSU brands should rethink that policy. A failing PSU could damage the cables as well. It seems like a bad idea to tell the customer to reuse part of a product that is not known by them to be good, even if it is compatible.

[artuc]

The warranty person probably can't authorise a resolution for this for you.

 

Not legal advice but I would send them a response along these lines, lay out all the facts and what action you will take.

 

 

Quote

 

I believe EVGA is liable for the following reasons.

 

Your warranty process caused this situation.

 

You failed to request the original cables and supply new.

 

You have failed to provide the user any warning documentation or labels advising of the difference and danger in using incompatible leads from a prior revision

 

Your product design allowed for this to occur (could be keyed/different connector type used etc.)

 

I have therefore had damage caused to the amount of $$$s to my storage hardware and cost of data recovery.

 

Please find report from xyz data recovery outlining the damage, their invoice and my quote for replacement drives.

 

You have x days to respond to this matter or I will proceed to:

 

contact consumer protection, small claims court, tribunal etc.

 

I hope that this can be resolved to everyone's mutual satisfaction

 

Warm Regards

 

 

 

Something along those lines. Could also probably hit up social media, review sites, tech journos/influencers about this If you don't get any traction but I'd try spelling it out to them first how much they stuffed up and it's on them.

 

 

[MarkPol88]

2 hours ago, artuc said:

Could also probably hit up social media, review sites, tech journos/influencers about this If you don't get any traction

Looks like Rossmann noticed his reddit post:

 

[OddOod]

https://www.tomshardware.com/pc-components/power-supplies/evga-replaces-customers-destroyed-hard-drives-after-power-supply-rma-blunder

[Godlygamer23]

On 3/19/2024 at 10:24 AM, sgircys said:

with no way of knowing other than by checking with a multimeter

The physical orientation of the wires would give it away as well. Some wires would start to run in a different direction - they would travel different paths versus the original cables. Now, this is still not something the average(or maybe any)user would detect because they would assume everything is the same as it's the same model, based on EVGA's own advertising. This was totally preventable on EVGA's part, from start to finish. 

[Poinkachu]

 

So, can update GN too on it, I guess.

[seon123]

58 minutes ago, Poinkachu said:

 

So, can update GN too on it, I guess.

I just realised that at 5:30, they referred to the G5, while OP has a GQ. So either the EVGA rep misspoke, or (the GQ and G5 are very similar, and FSP OEM) both models have the same pinout issue. 

 

On EVGA's website for PSU cable compatibility, the G5 is listed as being compatible with most of EVGA's PSUs, no other mentions. Let's hope this doesn't actually affect the G5