PC dies mid-gaming, then comes back from the dead?

xxtc59

Prominent
Oct 30, 2018
7
0
510
Today, during some gaming, my PC suddenly turned off and then wouldnt power on again (dead silence, no click, no fan attempting to turn), but(!) the standby LED of the Mobo was on and would come on every time I switched on the PSU. Unfortunately no error beeps since this case came without a speaker. I checked for some visible damage on PSU or Mobo, but nothing. I did some basic troubleshooting steps:

- Unplugged power button and tried to boot by shorting the pins on the Mobo, nothing.
- Removed the GPU, I replaced it very recently so it seemed like an obvious culprate, nothing.
- (while GPU removed) reset BIOS, by moving that little jumper, didnt disconnect the battery, PC turns on and boots into Windows like nothing ever happened, everything appears to be working.

I am obviously still very worried since this does seem very strange to me. Is this a symptom for something obvious? Is there anything I can do to check if everything is indeed fine before I plug in the GPU again and pretend like nothing happened?


Relevant specs:
i5 4590 with stock cooler
H97 PRO-GAMER
4x4GB Crucial DDR3-RAM
XFX RX 580
Cooler Master G550M PSU
 
Solution
That's crazy that there is no newer version for that motherboard. Practically every board I've looked at has a more recent bios version out for the fact of addressing the recent Spectre and Meltdown vulnerabilities if nothing else, but that one does not.

I think you would be wise to try that graphics card in somebody's, much newer, motherboard, or one with a bios version that has been released SINCE that card was released, to see if it works in another system or is simply not wanting to work in your system due to a lack of bios support for the card architecture.

You're probably right that it is a bad card if it has been fully seated, power connectors attached, etc.

I would however try to do a hard reset by doing the following...
How old is that G550m PSU? That wasn't a great power supply when it was new, and the first of those units hit the market back in 2013, so that unit could be as much as five years old by now. That would certainly be my first suspect, plus, it's REALLY borderline as to whether that unit is even capable enough for an RX 580 that we'd much rather see a very good 650-750w unit with.

We've recently seen member of this forum, and of the moderation team for that matter, with a VERY good Seasonic Prime Ultra Titanium unit that could not power his RX 580 adequately without tripping the over current protections, and that's one of the best 550w units on the planet. We know it was the protections on the PSU causing the same symptoms you saw, because it happened multiple times on his system and replacing the less than a year old unit with a brand new Seasonic Prime Ultra 750w unit completely resolved his issues AND the other unit is working fine in a less demanding system now, I believe.

Certainly there could be other culprits, but that's what I'd be looking at first.
 

xxtc59

Prominent
Oct 30, 2018
7
0
510


Hmm, interesting. All components are 3 years old pretty much on the dot, although quite heavily used, the PC is always on when I'm home. I kind of dismissed the possibility of it being the PSU first since I had the notoriously power hungry 280X before (using 2x8 pin power compared to the 1x8 pin of this 580) and it never had issues, but if the symptoms match I'll have to look into it. I assume there is no easy way to actually test this without getting another PSU?
 
First, I'd test your unit to see if it is even still putting out what it should be. Run Prime95 version 26.6 or Heaven benchmark WHILE you have HWinfo open and take screenshots of the system voltage sensors. 3v, 5v and 12v. You will probably have to scroll down to find those sensor values.

In cases where it is relevant and you are seeking help, then in order to help you, it's often necessary to SEE what's going on, in the event one of us can pick something out that seems out of place, or other indicators that just can't be communicated via a text only post. In these cases, posting an image of the HWinfo sensors or something else can be extremely helpful. That may not be the case in YOUR thread, but if it is then the information at the following link will show you how to do that:

*How to post images in Tom's hardware forums



Run HWinfo and look at system voltages and other sensor readings.

Monitoring temperatures, core speeds, voltages, clock ratios and other reported sensor data can often help to pick out an issue right off the bat. HWinfo is a good way to get that data and in my experience tends to be more accurate than some of the other utilities available. CPU-Z, GPU-Z and Core Temp all have their uses but HWinfo tends to have it all laid out in a more convenient fashion so you can usually see what one sensor is reporting while looking at another instead of having to flip through various tabs that have specific groupings.

After installation, run the utility and when asked, choose "sensors only". The other window options have some use but in most cases everything you need will be located in the sensors window. If you're taking screenshots to post for troubleshooting, it will most likely require taking three screenshots and scrolling down the sensors window between screenshots in order to capture them all.

It is most helpful if you can take a series of HWinfo screenshots at idle, after a cold boot to the desktop. Open HWinfo and wait for all of the Windows startup processes to complete. Usually about four or five minutes should be plenty. Take screenshots of all the HWinfo sensors.

Next, run something demanding like Prime95 version 26.6 or Heaven benchmark. Take another set of screenshots while either of those is running so we can see what the hardware is doing while under a load.

*Download HWinfo


And, if you have access to a volt meter, or can pick up a cheap one, you can test manually as well.

https://www.youtube.com/watch?v=ac7YMUcMjbw
 

xxtc59

Prominent
Oct 30, 2018
7
0
510


Okay, it might actually be the GPU itself, I plugged it back in and again it did not turn on. When I removed it again it turned on again (it popped up the bios and said something about "power surge detected", although that mightve been me not plugging in the power cable properly, it was only half in at that time, not sure). Anyway here are the HWinfo screens for idle and Prime95 using the iGPU.

IDLE
uQAK0bT.png


PRIME95
TVRtpEy.png

 


I just dealt with a customer who brought in a rig with very similar issues. He was running an older FX 8350, but other than that your systems are comparable. In his rig I couldn't find any voltage dips in windows while benchmarking or in bios. His system like yours would randomly crash sometimes with a black screen of death sometimes with a restart. It ended up being his power supply even though I couldn't find any dips in the the output. As stated your power supply wasn't a great one even when new and you have three years on it now. The wiring gets old and more "brittle" so as you change hardware you may end up with a wire that isn't making full correct contact with the harness anymore. If you had a better psu I would suggest replacing the wire harnesses, however you would be better off just getting a new quality psu. I've personally had a great luck with EVGA Supernova +Gold and +Platinum.
 
So, with only the iGPU in use the voltages look fine. That is, of course, not telling us much of anything other than that the PSU isn't able to be instantly considered rubbish. It doesn't tell us that it IS ok, or not ok, just that it isn't shot completely. Whether it can hold up under the load of your gaming card which adds another 225w (150w for the 8 pin, 75w for the slot) to the draw under load, is a different story. Even just another 75w with no load might be enough to make a borderline unit weird out.

I would try the GPU card in the other x16 slot and see if you have the same issues.
 

xxtc59

Prominent
Oct 30, 2018
7
0
510


Tried it in the 2nd slot, same thing, nothing happens when the power button is pressed.

I also dug out my old (defective) 280X to double check, with that one installed it boots fine (if you can call loads of artifacts on the screen fine) and I was even able to get a screenshot in on idle, but when I installed the drivers to maybe even get a stress test in it didn't boot into windows anymore (but it does still turn on, just a blank screen when the login screen should come on, and loads of artifacts on the boot screen), so cant post that unfortunately.

Lr0OENJ.png

(interestingly no artifacts visible on the screenshot)

I am by no means an expert, but can I safely assume now that the new GPU is doa, or is there still a way it could be the PSU?
 

xxtc59

Prominent
Oct 30, 2018
7
0
510


I am already on the latest version (2503 from March 2016)
 
That's crazy that there is no newer version for that motherboard. Practically every board I've looked at has a more recent bios version out for the fact of addressing the recent Spectre and Meltdown vulnerabilities if nothing else, but that one does not.

I think you would be wise to try that graphics card in somebody's, much newer, motherboard, or one with a bios version that has been released SINCE that card was released, to see if it works in another system or is simply not wanting to work in your system due to a lack of bios support for the card architecture.

You're probably right that it is a bad card if it has been fully seated, power connectors attached, etc.

I would however try to do a hard reset by doing the following.


Power off the unit, switch the PSU off and unplug the PSU cord from either the wall or the power supply.

Remove the motherboard CMOS battery for five minutes. During that five minutes, press the power button for 30 seconds. After the five minutes is up, reinstall the CMOS battery making sure to insert it with the correct side up just as it came out.

Install the RX 580 making sure to firmly seat it until the lock engages, secure it to the I/O end on the case with the retaining screws, plug in the required supplemental power connectors to the card from the PSU.

Now, plug the power supply cable back in, switch the PSU back on and power up the system. See if the system now recognizes the card. If not, then I'd contact the seller or manufacturer about a replacement or RMA.
 
Solution