RMA'ing PSU

[redking2436]

So I started this thread just to get some advice on whether my issue is actually a reason to RMA,

Basically I have a system with a 5900x a 3080ti and a seasonic 1000w prime PSU and randomly spread over many days and weeks it will occasionally just shut down mid-game rather infuriatingly, at this point the system becomes unresponsive to the power button with only remnants of RGB lights still being on, to solve this I have to either turn the switch off on the power supply and wait a while or do the same with the anti surge protector everything is connected to. When it happens, monitor and lamp etc are still on with no input, so I know it's not the power itself or the surge protector. From what I could gather online this seems to be the overcurrent protection activating in the PSU.

Pretty much ruled out it being heat related as the fan in the PSU doesn't turn on very often and when hybrid mode is off and the fan runs constantly the error still occurs, now the tricky bit is this fault is sporadic only turning up at certain times when one day it would refuse to launch any intense game from ark to overwatch at high fps and other days having fault with Subnautica, I haven't been able to reproduce the error but have captured a video of what I need to do to clear it.

Am I wasting people's time trying to RMA this unit as I fear they won't be able to reproduce the problem and don't want to waste their time or my own, since replacing the unit with my old 850w PSU I had no further problems, but I didn't spend £220 on what at the time was regarded as one of the best PSU brands just to have it sit as a paper weight?

Thanks for any advice given.

 

[Tetras]

One option would be to massively downclock the 3080 Ti so you can guarantee it won't trip, then see if the problem persists. I heard of this with the older Seasonic Focus Gold models, but I didn't think it still applied to newer models and am unsure if Prime models were affected.

Seasonic has a pretty good RMA service, as far as I know, so I'd RMA it to them rather than the retailer, they'd also be aware if this issue applied to your unit.

 

[redking2436]

Unfortunately i already have a down clock on the 3080ti and it was still tripping , however despite trying my best today i cant for the life of me get it to reproduce the problem. Tried prime and furmark at the same time, loads of different 3d mark stress tests and subnautica which made me bother opening an rma in the first place. The only difference is the psu is sat on top of the case as i didnt want to recable the thing which leads me to wonder if the psu is overheating rather than over current issues. I have it in a https://www.overclockers.co.uk/phanteks-eclipse-p200-air-mini-itx-case-black-ca-0al-pt.html so there isnt too much cooling for it , may try and swap the rear fan to an intake and the front rad to exhaust

 

[[TBR]MadMark]

Are you sure it's the psu? Exact same symptoms and I even replaced the psu as my pc had shutdown and wouldn't turn on. First power cycle my gpu caught fire and I found my problem. Worth checking if you have a spare good gpu.

 

[redking2436]

Are you sure it's the psu? Exact same symptoms and I even replaced the psu as my pc had shutdown and would turn on. First power cycle my gpu caught fire and I found my problem. Worth checking if you have a spare good gpu.

Thats concerning, hope it wasnt as bad as it sounds, gpu seems fine as the issue is sporadic. It can be good for weeks then suddenly whole system shuts off like a power cut and wont turn on again until i either turn off the plug at the wall or the switch on back of psu for a while. Gets through multiple runs of furmark and benchmarks too.

 

[[TBR]MadMark]

My symptoms were sporadic also, my pc just shut off mid game, I had tried furmark to stress the gpu but as this problem can be mainly 1 of 3 things - mb/psu or gpu if you can I would check before rmaing your psu.

 

[MCFC_ANDY]

am i missing something here you have a 3080ti in a p200 mini, good case but i remember watching and reading about it as i fancied it, but even with a say 2060 it gets toasty, but a 3080ti.

 

[redking2436]

am i missing something here you have a 3080ti in a p200 mini, good case but i remember watching and reading about it as i fancied it, but even with a say 2060 it gets toasty, but a 3080ti.

It's actually a pretty good case, most of the reviews reviewed it with the 2 fans that came with it , but having since added a h115 on the front, noctua on the back and 2 slim fans under the graphics card nothing gets hot, however the psu does sit at the top of the case with no ventilation to outside which i didnt think was too bad as the fan never really kicked in on hybrid mode. Now i have the rear fan as intake and the side fans as intake i am hoping the colder air will keep the psu happy in case it was overheating. Won't know till it trips again :/

 

[MCFC_ANDY]

yep i know its a good case, but i remember seeing reviews the gpu gets toasty, even mid range cards got a little warm, but 3080ti i would have thought would be pushing it, but if your temps are fine, i guess its ok

 

[Vimes]

You would have little issues with returning your PSU to Seasonic once it has been accepted as a return. For me it was free postage to them, a competitor acts as their UK agent. Mine was confirmed at being faulty (750w Ultra Prime) and then replaced with a brand new next gen model. Your experience might be different.

I had a 3080Ti at the time, since refunded for different reasons. The 3080Ti along with my 12700k would randomly reboot and shut the PC down. Running firmark or other intensive benchmark type tests would not generate what I believed to be the issue, transient spikes. Even tho the PSU, along with the system as a whole, did not pull close to the maximum sustainable wattage that the PSU was capable of delivering it was these types of micro second spikes which could cause a OCP type failure. At least that is how I thought of it.
So many reports, and the odd video, will lead to feel that is one of the factors leading to the up and coming ATX 3.x specification and the capability of those PSU's in being better able to deal with these types of events.
You would feel that a 1000w PSU could well be able to deal wit ha 3080Ti, I have read of one review having to stop using a Seasonic 1000w in favour of a Corsair one with their 6900XT due to it tripping the OCP protection because of spikes in voltages. Those spikes were measured with a scope and not able to be shown from software measurement.

As GPU's seem to be demanding more and more current I would have thought that the regulation would have applied to those in how they create these spikes. It seems to be the need for the PSU's tho in how to deal with it, hence ATX 3.x.

For you it might be an indicator of what is happening, for me Seasonic were great to deal with and replaced the faulty PSU very promptly.

 

[redking2436]

You would have little issues with returning your PSU to Seasonic once it has been accepted as a return. For me it was free postage to them, a competitor acts as their UK agent. Mine was confirmed at being faulty (750w Ultra Prime) and then replaced with a brand new next gen model. Your experience might be different.

I had a 3080Ti at the time, since refunded for different reasons. The 3080Ti along with my 12700k would randomly reboot and shut the PC down. Running firmark or other intensive benchmark type tests would not generate what I believed to be the issue, transient spikes. Even tho the PSU, along with the system as a whole, did not pull close to the maximum sustainable wattage that the PSU was capable of delivering it was these types of micro second spikes which could cause a OCP type failure. At least that is how I thought of it.
So many reports, and the odd video, will lead to feel that is one of the factors leading to the up and coming ATX 3.x specification and the capability of those PSU's in being better able to deal with these types of events.
You would feel that a 1000w PSU could well be able to deal wit ha 3080Ti, I have read of one review having to stop using a Seasonic 1000w in favour of a Corsair one with their 6900XT due to it tripping the OCP protection because of spikes in voltages. Those spikes were measured with a scope and not able to be shown from software measurement.

As GPU's seem to be demanding more and more current I would have thought that the regulation would have applied to those in how they create these spikes. It seems to be the need for the PSU's tho in how to deal with it, hence ATX 3.x.

For you it might be an indicator of what is happening, for me Seasonic were great to deal with and replaced the faulty PSU very promptly.

I contacted seasonic after tetra's suggestion , they basically had me do the same things as ocuk which is fair enough , if they can't prove it's a faulty unit it's a waste of their time but its a tricky situation for me as even though i have video evidence i cant seem to recrate the event now, maybe redoing the cables for the 4th time after swapping out psu did something or its not got hot enough due to the ventilation change. Going to just wait and see what happens though, considering a new case, if it is an overheating issue with things getting more power hungry and hot because of it.

 

[Vimes]

Just make sure you aren't using cable extensions either, just for the purpose of elimination. As well as, separate GPU cables from the PSU. Kinda obvious, to those who know.

 

[Bottle]

I'm currently having issues with BeQuiet Dark Power 12 1000W PSU. I'm on my second PSU and it has the same low hum all the time. In a silent room with a pretty silent PC, it's the loudest noise. I can even hear it when I put headphones on. I can't work out if this is a bad batch of B-grade PSUs sent to my source or if it's normal and I should just suck it up. Or should I try yet another one from a different source and hope third time lucky.

 

[redking2436]

I'm currently having issues with BeQuiet Dark Power 12 1000W PSU. I'm on my second PSU and it has the same low hum all the time. In a silent room with a pretty silent PC, it's the loudest noise. I can even hear it when I put headphones on. I can't work out if this is a bad batch of B-grade PSUs sent to my source or if it's normal and I should just suck it up. Or should I try yet another one from a different source and hope third time lucky.

have you tried it outside the case to make sure it's not a vibration ?

 

[Bottle]

have you tried it outside the case to make sure it's not a vibration ?

Yes. Originally the supplier wanted me to check that I wasn't confusing it with the GPU, so I removed the PSU so I could unambiguously put my ear to it. Definitely the PSU. It's not a high picthed coil whine, but a lower pitched buzz, almost rumble. The fan is turning all the time and I get the impression it's in fact fan noise (which is ironic being a BeQuiet product). The tricky thing is, I need my PC so can't wait for a new PSU to arrive, I need to commit to a new PSU, receive and fit it and then RMA the current one. At £260, it's a bit unnerving.

 

[Bottle]

Here's a link to a recording of the noise. https://file.io/GCKgTknclPI0

 

[Bottle]

Having replaced the BeQuiet 1000W PSU with a third one, I can report that the low hum sound is still present and so must be normal. What I conclude from this is that BeQuiet PSUs are anything but quiet. When you read in the reviews that, "I had to put my ear right up to the PSU to hear anything", don't believe it. With a quiet system, you can hear the hum from your desk and even with headphones on. However, I suppose if your room has a reasonable amount of ambient noise, you won't notice the PSU. I've resigned myself to living with it and won't be RMAing this one and experimenting further.

 

[Vimes]

@Bottle

I tried to download your file, but it seems to have been deleted.
Nose is subjective, to a point. I do know that my Corsair RM1000i, 2021 edition, is absolutely quiet.