PC unexpectedly loses power, please help!

Aug 2, 2018
4
0
10
specs.png



I have been experiencing random loss of power with my PC for several weeks that has progressively worsened. Now so much worse that I often cannot use it at all and sometimes it will shut abruptly and go into a cycle of power losing (constantly restarting) and the way to get out of this is to turn power switch off on the PSU.

My PSU is a Corsair AX1200i and the self test shows no faults. I have an independent PSU tester but that would mean unplugging every cable from all PC components and then testing it in the tester and I see no reason to do this as the PSU shows no fault on its self test. I had to unplug all cables from the PSU prior to being able to do that test. The obvious fault for this symptom is usually PSU or bad RAM or driver issue but no windows errors are logged or any BSOD, dump files etc.

I also have 2 AMD R9 295X2 GPUs.

The motherboard and CPU is around 11 years old and limits power of GPUs, I am aware of that and plan to use these GPUs in a far better system in the near future.

I have no other PSU to test. I did buy a new HDD which is the Seagate shown in the image above showing specs. I normally use a 1TB Samsung 840 Evo SSD. I wanted to be sure I could rule out any windows, driver issues or possible malware that wasn't detected as the cause.

I am not sure where else to go from here. The power issue still remains. It appears to be random from what I can tell. It happens when nothing else is running and system is idle or when doing general use such as web browsing. I do know my PSU power cable is fitted with a 10amp fuse and it probably needs less than this for safety sake of components? Not sure though.

I did unplug the two GPU x2 8 pin power cables from both GPUs and just used the CPU integrated graphics for testing and it still had power failure. I did not unplug the GPUs from the motherboard though so some power would still be drawn even while not in use.

I don't have spare PSU, motherboard, CPU and RAM to try but sure if I did it would most likely would work fine.

Does anyone know of any issues with latest non beta AMD Radeon R9 drivers, the adrenalin v 18.51 that may cause this issue, though this problem predates that to version before. I was thinking maybe GPU driver use issue but this also happens while system is idle and no active programs are in use so that was why I had ruled out PSU and components because of lack of BSOD or errors, which left me with windows or malware issue, hence the fresh W7 install on a brand new HDD.

My corsair link was showing power of 246 volts (UK power) though I was only using less than 300W so is probably normal. I think it can stand up to 250V for this PSU. Many always say we have 230V in UK but we actually still have 240V though we once had more than this. Every test and power monitoring device I have shows this and when amp use is high voltage drops as is to be expected, as low as 230V when using 3KW. I think this is all besides the point because my mains power has not changed at all. I havec checked out other threads with similar problems and tried some of those suggestions, including taking out CMOS lithium cell, etc.

Any suggestions and help with this is greatly appreciated. I've never had anything quite like this that went unresolved and was so random without being easily diagnosed for so long. Thank you.
 
Assuming you are having no power brownouts (such as when a high current air conditioner turns on?), then a serviceable known good PSU of sufficient wattage will be the likely first step...especially since your rig is still powering off even with no GPUs installed, which alone would spare your rig of several hundred watts of peak power drawn...

Remember, all failing PSUs, regardless of pedigree/name-brand status, were usually fine just a day earlier :)
 
mdd1963, true it could well still be a failing PSU and without looking more closely at components inside we wouldn't know unless we try a different PSU. I will most likely buy a new PSU then to test this but first wanted to rule out motherboard, CPU or RAM faults. Could it not be any of those? I know it's difficult to know for sure without replacement of parts.

Another thing that concerns me is that the BIOS may be infected with malware and as a result infects the OS even on every clean install. I wish there was an easy error code or BSOD with a file dump. If only it were that easy.

I tried combofix last night and it said that comres.dll was infected so it replaced that with an original copy. I notice that online people say this is not a necessary but optional and is not core part of OS bit notice that my GPU drivers make use of it. I actually have several copies in different directories.

I will most likely either buy a low wattage cheap PSU for testing or a high wattage one for around £100 that is 2000W I think bit loud and not suitable for everyday use. Rather than spend £300+ on a PSU I may not need. At this rate I feel like building a whole new system for £900+ Will have to think about this more.