[SOLVED] Computer Will Not Stay On.... Running out of Ideas?

Oct 29, 2020
4
0
10
Hey everyone,

Here is the story. My little sister has a custom built PC that she has been using for awhile and she upgraded her CPU and Motherboard this past year. Recently, it has run into a big issue that I am scrambling my brain with. After sitting powered on for anywhere from 40 to 60 hours, the PC will shut itself off. The motherboard lights will remain on however. When attempting to turn the PC on - I get absolutely no response whatsoever. The PC will sit like this forever until I unplug the power. Once I let the PC sit for about 2 or 3 days, I am able to plug the power back in, and boot it up absolutely no issues. It runs fine once again for anywhere from 40-60 hours and bam, the issue happens again.

It does have a Corsair 100 liquid cool on the CPU, however, I did a stress test on the CPUs performance when it was on, and it never went above 40 degrees Celsius. The ram also didn't seem to have any issues and even after swapping from 2 sticks to 1 in all positions on the MOBO, I am unable to get the computer to turn on and stay on. I have also removed the GPU entirely from the MOBO to see if this was the cause of any issues and it refuses to turn on still.

Basic specs of the computer are as follows.

Corsair 100 Liquid Cool
i5-9300HF CPU
ASUS TUF Z390-Plus Gaming Mobo
Msi GTX 970 100me
Corsair CX750m PSU
2x HyperX Fury 8g Cards DDR4

If anyone has any ideas or is interested in solving this strange circumstance. It would be appreciated! Looking for the Sherlock Holmes of Computers haha.
 
Solution
What do you mean by "power drained"? Are you just leaving it turned off, or are you flipping off the switch on the back of the PSU and unplugging it from the wall?

I'd recommend you run an actual stability test, just to rule out stability. Download ASUS Realbench, click on the stress test (Not benchmark) and select half the amount of your installed RAM in the options, then run the test for an hour. See what happens. Check on it periodically to monitor whether it errors out or if the temperature exceeds 80°C. Probably won't, but I'd do it anyhow to be sure.

Then, I'd download Prime95, choose the Small FFT option, disable AVX by unchecking the option next to each of the AVX options. AVX, AVX 2 and AVX 512 if it becomes un-gray after...
What Windows version are you running?

Also, I do not see the i5-9300HF on the CPU support list for that motherboard. Well, it's not listed on the Plus gaming WiFi support page anyhow, which would actually probably show MORE support than the non-WiFi version since it's newer, but the ASUS website must be in a construction phase right now because there are a number of motherboards today not showing up there and the Plus gaming is one of them. Either way, it COULD be a factor, even though it tends to "work" normally most of the time. Probably it isn't relevant and is simply an oversight on the support list, but I've seen other circumstances where a CPU wasn't on the support list and it would "work" but there were weird quirks or behaviors.

Has it done this since the beginning with that CPU and motherboard or was it not doing this before and has just recently begun to do it?

How old is that CX750m, as in, how long has it been in service?

What did you use to "stress test" the CPU? This probably isn't relevant, but, you never know. There may be some process or service that is only scheduled to kick in every two or three days and that scheduled item is causing an error or other problem like a driver conflict.

When you upgraded the CPU and motherboard, did you do a FULL, CLEAN install of windows, from scratch, or did you just continue to use the old windows installation as is?
 
Oct 29, 2020
4
0
10
What Windows version are you running?

Also, I do not see the i5-9300HF on the CPU support list for that motherboard. Well, it's not listed on the Plus gaming WiFi support page anyhow, which would actually probably show MORE support than the non-WiFi version since it's newer, but the ASUS website must be in a construction phase right now because there are a number of motherboards today not showing up there and the Plus gaming is one of them. Either way, it COULD be a factor, even though it tends to "work" normally most of the time. Probably it isn't relevant and is simply an oversight on the support list, but I've seen other circumstances where a CPU wasn't on the support list and it would "work" but there were weird quirks or behaviors.

Has it done this since the beginning with that CPU and motherboard or was it not doing this before and has just recently begun to do it?

How old is that CX750m, as in, how long has it been in service?

What did you use to "stress test" the CPU? This probably isn't relevant, but, you never know. There may be some process or service that is only scheduled to kick in every two or three days and that scheduled item is causing an error or other problem like a driver conflict.

When you upgraded the CPU and motherboard, did you do a FULL, CLEAN install of windows, from scratch, or did you just continue to use the old windows installation as is?

When she upgraded the CPU and the Motherboard, she actually got in touch with a representative and confirmed the compatibility. Now obviously that was a human, human error could have occurred. That would definitely be interesting. But the computer itself was working for almost 6-7 months before these problems started. So I do not believe it to be a compatibility issue. Unless it is just now happening.

With that, as I mentioned, the computer has been running for almost 6-7 months before these issues started and originally the computer shut down and refused to power on. Which prompted me to let the power drain entirely before turning it back on. I double checked everything and just re-seated everything. Once I did that, everything seemed to be fine and it ran again. 2 weeks later the issue happened again, just shut down. Now it appears the computer will run for a solid 3-5 days before it just shuts down and needs to have all power drained before starting up.

I stress tested the CPU by doing what she does on the daily. Initiating a live-stream and running various windows on google chrome, will simulationously opening games. I had GTA V, Skyrim, and a bunch of other misc. games running on the screen. It immediately caused the CPUs temperature to rise 10-12 degrees Celsius and averaged around 40, but it never got much hotter than that except for a random spike to maybe 45 once and awhile. I figured this was an easy way to test it, since its the most intense thing she would ever be doing on the computer at any given time.

The CX750m is definitely pushing its age. I'd say its probably around 5 or 6 years of age at this point. She took it out of an old build our family had. It was tested just before her upgraded build and was working in tip top shape though. It also is still providing power somehow or someway considering it does let everything work for hours on end - even keeping the MOBO lit up even though the CPU itself won't power on.

I do want to get it professionally tested and plan on doing it here soon, but I just have an feeling that's not it... unless that's just me being wrong haha.

100% Full, clean installation of windows from scratch. She upgraded from HDDs to SSDs and got Windows 10 Pro.
 
What do you mean by "power drained"? Are you just leaving it turned off, or are you flipping off the switch on the back of the PSU and unplugging it from the wall?

I'd recommend you run an actual stability test, just to rule out stability. Download ASUS Realbench, click on the stress test (Not benchmark) and select half the amount of your installed RAM in the options, then run the test for an hour. See what happens. Check on it periodically to monitor whether it errors out or if the temperature exceeds 80°C. Probably won't, but I'd do it anyhow to be sure.

Then, I'd download Prime95, choose the Small FFT option, disable AVX by unchecking the option next to each of the AVX options. AVX, AVX 2 and AVX 512 if it becomes un-gray after unchecking the other AVX options. Usually you will need to uncheck one of them before the others become "un-grayed out", if you get my drift.

Then, run it on the Small FFT, not "Smallest FFT", just "Small FFT", option, for 15 minutes. If it is thermally compliant it will remain under 80°C. If it exceeds 80°C at any point then end the test by selecting stop or exit from the drop down file menu. Don't simply click the X in the top right hand corner like you would for most applications because that will not stop the utility from running, it will only reduce it to the system tray. You must use the drop down File menu to stop or exit.

If it does fine on the stability and the thermal compliance test, then I'd move on to testing the power supply. If you have any sort of volt meter, whether analogue or digital, or can borrow one, I can show you how to easily test the power supply to see if it needs to be replaced or not. This definitely sounds like a capacitor issue, although it could be something else. You might want to verify periodically that the fan for the power supply is continuing to run and doesn't shut off at some point causing the PSU to overheat and shut down, OR, better yet, considering the age of that PSU which is about two to three years past it's warranty expiration, just replacing it to eliminate that as an issue especially since problems where the system simply shuts off end up being the power supply about 80% of the time.
 
Solution
Oct 29, 2020
4
0
10
What do you mean by "power drained"? Are you just leaving it turned off, or are you flipping off the switch on the back of the PSU and unplugging it from the wall?

I'd recommend you run an actual stability test, just to rule out stability. Download ASUS Realbench, click on the stress test (Not benchmark) and select half the amount of your installed RAM in the options, then run the test for an hour. See what happens. Check on it periodically to monitor whether it errors out or if the temperature exceeds 80°C. Probably won't, but I'd do it anyhow to be sure.

Then, I'd download Prime95, choose the Small FFT option, disable AVX by unchecking the option next to each of the AVX options. AVX, AVX 2 and AVX 512 if it becomes un-gray after unchecking the other AVX options. Usually you will need to uncheck one of them before the others become "un-grayed out", if you get my drift.

Then, run it on the Small FFT, not "Smallest FFT", just "Small FFT", option, for 15 minutes. If it is thermally compliant it will remain under 80°C. If it exceeds 80°C at any point then end the test by selecting stop or exit from the drop down file menu. Don't simply click the X in the top right hand corner like you would for most applications because that will not stop the utility from running, it will only reduce it to the system tray. You must use the drop down File menu to stop or exit.

If it does fine on the stability and the thermal compliance test, then I'd move on to testing the power supply. If you have any sort of volt meter, whether analogue or digital, or can borrow one, I can show you how to easily test the power supply to see if it needs to be replaced or not. This definitely sounds like a capacitor issue, although it could be something else. You might want to verify periodically that the fan for the power supply is continuing to run and doesn't shut off at some point causing the PSU to overheat and shut down, OR, better yet, considering the age of that PSU which is about two to three years past it's warranty expiration, just replacing it to eliminate that as an issue especially since problems where the system simply shuts off end up being the power supply about 80% of the time.

Power drained = Unplugged from wall and PSU turned off. Letting it simply sit for almost 24 hours.

I definitely will attempt to do the stability test once I can get it to power back on. It just shut down again today which is what prompted me to come ask questions here. If I am unable to get it to power back on, I will definitely attempt the PSU test first. I do have an extra PSU in a different rig that I know is working currently, so I may attempt to pull from that one in the mean time and see if it starts up. Won't have free time to do that until Saturday. I appreciate the input and the test ideas. I know a lot of basics with computers, but not too much of the more in-depths processes. So I really do appreciate it again.

I will keep you and anyone else looking, updated. Hopefully it is just the PSU!
 
Well, it would probably be a LOT easier to test the PSU that's in there now, than it would to swap PSUs, and if you live in a region where there are hardware stores and home centers, or Walmart type stores, you can pick up a volt meter for very cheap. If there's a harbor freight near you you can get one for like 7 bucks:

https://www.harborfreight.com/7-function-digital-multimeter-63759.html

Or Amazon:

https://www.amazon.com/Neiko-40508-...2470955011&rnid=2470954011&rps=1&s=hi&sr=1-14

And for maybe ten bucks more than those you can probably find a much better quality one if it's something you'd like to have on hand, since they are very useful for a LOT of things, like checking the voltage of batteries (Computer CMOS battery, AAA, AA, C, D, 9v, your car battery's at rest and charging voltages, etc.), among other things.

 

TRENDING THREADS