[SOLVED] Random shut offs, now won't boot but fans are running

[Justinsanity]

GPU: RX 480
PSU: Seasonic S12 II Bronze 620W
Mobo: 970A-UD3P

Randomly, my computer would shut off without warning, but the fans and power light would stay on. It happened both while under a full load and while idling. I would have to shut off the PSU and turn it back on to restart my PC. It became more and more frequent but now it won't boot. Pressing the power button lights up the case and turns on the fans but that's it. No boot beep or anything.

I have tried:
-Different RAM sticks
-New PSU cable
-Different surge protectors
-Plugging directly into the wall outlet
-Checking all internal cables
-Unplugging PSU and holding the power button for 30 seconds

I unfortunately don't have another PSU on hand. It sounds a lot like I need to replace my PSU, but I read it could also be the motherboard. What do you guys recommend?

 

[vov4ik_il]

I would run some stress tests on one component at a time (to rule out CPU/VRM/GPU/RAM) and then if nothing fails on all together (to test the PSU).

 

[drivinfast247]

I'm surprised it's even working at all without a CPU.

 

[MadsModsat]

To be honest, I would suggest that you verify or rule out the PSU as the possible cause for the problems, before proceeding with various stress tests.

If a PSU fails completely, it can potentially cause permanent damage to other hardware components in your PC in the process. This can become very expensive.

It does sound like the PSU could be a very likely cause for the problems, and I think it should be prioritized to be investigated further.

When you stresstest the PC, it puts a lot of load on the PSU. The more you push a potentially failing PSU, the higher the risk of catastrophic failure, which again leads to increased risk of severe damage to the entire PC, in a worst case scenario.

Anytime you keep a system with a possibly failing PSU powered on, is an oportunity for the PSU to fail completely - and stresstesting can be the one thing, that finally provokes catastrophic failure.

However, if the PSU turns out to have no connection to the stability issues you are struggeling with, and can be confirmed as being in 100% working order, only then would stresstesting individual components to narrow down the faulty piece of hardware be a good idea, in my opinion.

What you should take into consideration when troubleshooting, and the PSU is one of the possible faulty components, is that compared to other hardware, the PSU has a lot more potential to cause a lot of damage to almost all components when it fails, if you're unlucky.

If the CPU is faulty, or if the GPU is dying, or a RAM module has died, the risk of additional damage is far less likely than the risk posed by the PSU.

It would be a shame to damage your graphics card or another expensive component, only to find out later, that the PSU was faulty, and the GPU suffered damage for no good reason.

In my opinion, you should only proceed with stresstesting, if you feel confident that the PSU is good, or if you think the increased risk of stressing the possibly failing PSU is worth it, all things considered

 

[mdd1963]

If /when you swap the PSU, be sure to not mix/reuse old modular cables with new PSU...ever

 

[Justinsanity]

Thanks for everyone's input. I don't have the opportunity to stress test anything as the computer won't even boot. I suppose the only way of knowing its the PSU is by swapping it for a quick test. I'm not sure when I'll get the opportunity to do that but I'll keep it in mind.

Hesitant to ask but,,, I do think I have a really old PSU laying around... would it be beneficial to see if I can get a boot with it? I wouldn't plug it into the GPU, would it even boot without powering the graphics card?

 

[vov4ik_il]

I wouldn't plug it into the GPU, would it even boot without powering the graphics card?

If you take the GPU out and have onboard graphics, it would... but what will it verify?

 

[Justinsanity]

If you take the GPU out and have onboard graphics, it would... but what will it verify?

I would hope it would point to which component is broken - PSU or motherboard.

With the (possibly broken) PSU, the motherboard makes absolutely no beeps whatsoever when I press the power button. I also tried it without the GPU installed and initially there were no beeps either.

I tested my old PSU also without the GPU installed and it made a long beep with 3 short ones, indicating no video card present.

At this point I was convinced it was just because of a faulty PSU because my other one was making no beeps. But just as a final test I tried it again with the possibly broken PSU and it also did the beeps indicating no GPU (as I hadn’t put the card back) Would a faulty motherboard do these beeps?

Are there any other methods to test what exactly is causing the problem?

 

[MadsModsat]

There are other possibilities than a broken PSU of course. But from your descrition, it just sounds like the exact sequence of events you see when a PSU is slowly failing. Everything from random BSODs and other small issues to begin with, and then everything escalating from there on, until it is almost unable to boot. It seems like an open and shut case

But that can also be problematic in itself, the way it seems so obvious. So it seems other possible causes for the PC problems should be looked at more closely as well.

But I'm curious what backup PSU you used for testing, what make and model is it?

You could be in an unfortunate position, where both PSU are not of sufficient performance or quality to power your system, which won't help much in narrowing down possible causes. That fact the the old PSU gives enough power for the speaker to beep after you tried the older one, doesn't reveal a lot on its own, it sometimes happen, although added up to everything else, will help make sense of it all at some point, I'm sure.

RAM can also cause all sorts of problems, but not only have you tried different RAM, the symptoms you describe are a bit different in my mind. Maybe not to begin with, but now that there's no real sign of life, I'd probably look elsewhere for the culprit. In my experience, faulty RAM often lets the PC power on, but would cause endless bootloops or similar instead.

Motherboard could be the problem as well, but I can't really think of what part of a failing motherboard would result in the issues you describe, in the same sequence of events. Could be related to power delivery.

I actually recieved a factory new 9900K CPU which was faulty a few months ago, but it was only the second time in 20 years I came across a factory new CPU which was faulty. It is not unheard of, but I'd say it is a rare occurence. But obviously the CPU was fine until recently, and the symptoms you describe doesn't really point to the CPU being faulty, in my personal opinion.

A failing GPU often shows signs of failing by some graphical glitches in the beginning. I'm not saying that always happens, but in my personal experience, the GPU shows signs of failing both visually and maybe preventing the system from powering on at a later stage.
But it could be the next step to troubleshoot.

Unfortunately I'm focusing a bit too much on the PSU as the most likely culprit, so I don't have any particularly useful suggestions right now, but I'll try and see if something comes to mind.

I'm sure someone will recognize the scenario you are describing, who could offer much better advice than I can think of right now

 

[Justinsanity]

I'm sure someone will recognize the scenario you are describing, who could offer much better advice than I can think of right now

Thanks for your detailed response. The old PSU I tried is quite low on wattage, which is why I didn't have my GPU plugged in. I decided to order a new PSU, and I'm also in the process of upgrading nearly every component on my PC. I will report back here if the new PSU fixes my old system for anyone else that has this issue in the future.

 

[MadsModsat]

It is important, that I can only comment from my own experience - I'm not a certified expert in the area. So I might have missed something obvious, but starting over with a new PSU is very useful, and if it turns out it was not the cause of the problem, it would at least help narrow down what the actual failing component is more likely to be.
If it is the GPU for example, it would be difficult to know if the PC crashed during stresstesting due to the old PSU which was a posssible problem, or the GPU on it own.
Now you can chose your troubleshooting steps with the knowledge in mind, that it is more likely the GPU than the new PSU, if you end up in a situation like that. It was a bit of a clumsy explanation, I hope it maskes somekind of sense.

But I'll keep my fingers crossed that the new PSU will turn things back to working order.

An update about the result would be great, threads were a solution is found and the thread is updated with the right answer, is much more useful to others who has the same problem and happens to come across your thread in the future, than a thread which is marked as solved, but no solution is listed

 

[Justinsanity]

An update about the result would be great, threads were a solution is found and the thread is updated with the right answer, is much more useful to others who has the same problem and happens to come across your thread in the future, than a thread which is marked as solved, but no solution is listed

Installed the new PSU today and I am currently writing this response on the PC that used to not boot. Seems to be working fine right now so I'm thinking my old PSU was busted. I will update this thread if I run into the problem again!