pc randomly shuts down

[d00m1ord]

so one of the computers in my house keeps randomly shutting down. it can be on and working and then suddenly it just turns off. I have looked in windows event viewer and found nothing and it doesn't blue screen it just dies.

 

specs are

 

i5 4430 with stock cooler

8 gb corsair xms3 ram

gigabyte GA-H81N motherboard

wd blue 500gb ssd

FSP270-60LE PSU ( only has a 20 pin connector for the mobo not a 24 )

 

I suspect the power supply but im not certain hence why im asking for help =)

 

thanks in advance

[Bombastinator]

Could be hardware or software.

It’s an old machine so checking temps to see if the TIM went hard and cracked is a big one.   PSU, running a few rounds of memtest probably couldn’t hurt either.

[d00m1ord]

5 minutes ago, Bombastinator said:

Could be hardware or software.

It’s an old machine so checking temps to see if the TIM went hard and cracked is a big one.   PSU, running a few rounds of memtest probably couldn’t hurt either.

Ok is there any way to tell between hardware and software issue?

 

ill run memtest on it should I also stress test cpu or not?

 

Thermal paste seemed good when I took the cooler off but it won’t hurt to reapply it

 

thanks for the suggestions

[Bombastinator]

22 minutes ago, d00m1ord said:

Ok is there any way to tell between hardware and software issue?

 

ill run memtest on it should I also stress test cpu or not?

 

Thermal paste seemed good when I took the cooler off but it won’t hurt to reapply it

 

thanks for the suggestions

Did you reapply when you took the cooler off?  It has to be done every time you do that.  The paste needs to be really completely completely cleaned off each time as well.

 

you won’t need to do a stress test if it’s heating up so bad it crashes under normal use.  It should be way too hot to start with.  What OS is it running? 

[d00m1ord]

8 hours ago, Bombastinator said:

Did you reapply when you took the cooler off?  It has to be done every time you do that.  The paste needs to be really completely completely cleaned off each time as well.

 

you won’t need to do a stress test if it’s heating up so bad it crashes under normal use.  It should be way too hot to start with.  What OS is it running? 

Haven’t reapplied paste yet but I was planning to just need to get some paste 

 

it’s running Windows 10 most recent update

[Bombastinator]

14 minutes ago, d00m1ord said:

Haven’t reapplied paste yet but I was planning to just need to get some paste 

 

it’s running Windows 10 most recent update

If you removed the cooler but didn’t redo the full and complete repasting ritual you may have your problem right there.  Not reapplying paste does sometimes work so it’s not a surety.

[d00m1ord]

1 hour ago, Bombastinator said:

If you removed the cooler but didn’t redo the full and complete repasting ritual you may have your problem right there.  Not reapplying paste does sometimes work so it’s not a surety.

I already had the problem before I removed the cooler so thermal paste isn’t the cause

 

also I’m having trouble running memtest. I made the bootable usb but the pc doesn’t see it as a boot device what have I done wrong?

[Robchil]

if it is more systems in the same house i would suspect the powersource, and put a UPS inbetween to take off the drops. 

 

EDIT: I'm blind tho.. one of the computers:

 

PSU's degenerate over time, when the just shuts down it looks like lack of power. 

 

 

 

[Bombastinator]

4 hours ago, d00m1ord said:

I already had the problem before I removed the cooler so thermal paste isn’t the cause

 

also I’m having trouble running memtest. I made the bootable usb but the pc doesn’t see it as a boot device what have I done wrong?

“The” Bootable USB?

[d00m1ord]

1 hour ago, Bombastinator said:

“The” Bootable USB?

Yes? Downloaded memtest and followed the instructions to put it on a usb stick and followed instructions to boot from it and it wasn’t detected

[Bombastinator]

6 minutes ago, d00m1ord said:

Yes? Downloaded memtest and followed the instructions to put it on a usb stick and followed instructions to boot from it and it wasn’t detected

Ah.  I just run it.  It’s part of win10 I think.  There may be something about this run it from a bootable drive thing.

[d00m1ord]

20 minutes ago, Bombastinator said:

Ah.  I just run it.  It’s part of win10 I think.  There may be something about this run it from a bootable drive thing.

Do you mean the windows memory diagnostic built into windows?

[Bombastinator]

4 minutes ago, d00m1ord said:

Do you mean the windows memory diagnostic built into windows?

I don’t know.  There used to be one.  My knowledge or win10 is limited.  I’ve got it now, but I was a former user of 8.1 and win10 seems to change slightly with every iteration of which there have been many.

[d00m1ord]

12 minutes ago, Bombastinator said:

I don’t know.  There used to be one.  My knowledge or win10 is limited.  I’ve got it now, but I was a former user of 8.1 and win10 seems to change slightly with every iteration of which there have been many.

Well I ran the memory tester built into windows 10 and it found nothing I thought you meant the program memtest that you have to download and make a bootable usb from

[Bombastinator]

1 hour ago, d00m1ord said:

Well I ran the memory tester built into windows 10 and it found nothing I thought you meant the program memtest that you have to download and make a bootable usb from

The program runs better.  It’s frequently suggested over the built in one.  The built in one has a habit of throwing false negatives.  The program is a bit better about it.   Not getting a bootable win10 USB is a different problem and one I know little about.  It’s a problem I’ve seen posted about and people have answered it so there are people here who know how to deal with it.  I’m just not one of them.  
 

it’s entirely possible memory is not the problem.  It’s not even super high likelyhood.  It’s mostly that testing for it is cheap and easy... or it’s supposed to be..

 

PSU I think is most probable.  PSUs are difficult to get and expensive right now because of supply chain stuff.  So lower likelyhood but free options gain some precedence because if you go through the PITA to replace the PSU in this environment and it ISN’T the problem it’s kind of annoying.

[d00m1ord]

32 minutes ago, Bombastinator said:

The program runs better.  It’s frequently suggested over the built in one.  The built in one has a habit of throwing false negatives.  The program is a bit better about it.   Not getting a bootable win10 USB is a different problem and one I know little about.  It’s a problem I’ve seen posted about and people have answered it so there are people here who know how to deal with it.  I’m just not one of them.  
 

it’s entirely possible memory is not the problem.  It’s not even super high likelyhood.  It’s mostly that testing for it is cheap and easy... or it’s supposed to be..

 

PSU I think is most probable.  PSUs are difficult to get and expensive right now because of supply chain stuff.  So lower likelyhood but free options gain some precedence because if you go through the PITA to replace the PSU in this environment and it ISN’T the problem it’s kind of annoying.

So I’ve run a virus scan, malware scan, registry check, stress test, memory test, I even had the computer just with some programs open to simulate typical use and it ran perfect for about 6 hours then 4 restarts in a row. It turns off cpu fan stops spinning and then it boots right back up and it did that 4 times in about 1 min before I just shut it off.

 

took the cover off the case and it felt warm inside but not from the cpu cooler since that has a huge vent directly above it but it felt warm around the PSU and it felt quite warm to the touch. Now the PSU only has a small 40mm fan on one end of it and there is not a single bit of ventilation near it.

 

could it be that the PSU can’t get rid of its heat due to airflow and is getting to hot and shutting down?

[Bombastinator]

42 minutes ago, d00m1ord said:

So I’ve run a virus scan, malware scan, registry check, stress test, memory test, I even had the computer just with some programs open to simulate typical use and it ran perfect for about 6 hours then 4 restarts in a row. It turns off cpu fan stops spinning and then it boots right back up and it did that 4 times in about 1 min before I just shut it off.

 

took the cover off the case and it felt warm inside but not from the cpu cooler since that has a huge vent directly above it but it felt warm around the PSU and it felt quite warm to the touch. Now the PSU only has a small 40mm fan on one end of it and there is not a single bit of ventilation near it.

 

could it be that the PSU can’t get rid of its heat due to airflow and is getting to hot and shutting down?

Checking temps by hand isn’t very accurate.  CPUs, GPUs and motherboards all have multiple temp sensors on them and you can check temps pretty accurately that way.

[d00m1ord]

9 minutes ago, Bombastinator said:

Checking temps by hand isn’t very accurate.  CPUs, GPUs and motherboards all have multiple temp sensors on them and you can check temps pretty accurately that way.

The temp sensors on the mobo and cpu won’t tell me the temp of the PSU though but I did measure the temp using an industrial thermometer I use for work and know to be accurate and I measured the temp of the air coming off and the temp of the PSU casing and both were 40 c  strikes me as hot. Is it not?

[Bombastinator]

11 minutes ago, d00m1ord said:

The temp sensors on the mobo and cpu won’t tell me the temp of the PSU though but I did measure the temp using an industrial thermometer I use for work and know to be accurate and I measured the temp of the air coming off and the temp of the PSU casing and both were 40 c  strikes me as hot. Is it not?

Not horrifically I think.  I don’t have a good base concept for that though.

 

How old is the PSU?  What make and model is it?

 

If the PSU is old there’s an argument for replacing it anyway in general principal.  Old PSUs aren’t as safe as the more modern ones and can go bad even faster, so anything more than about 5 years old is arguably old enough to change out.  There’s a widely variable exception for prebuilts.

[d00m1ord]

17 minutes ago, Bombastinator said:

Not horrifically I think.  I don’t have a good base concept for that though.

 

How old is the PSU?  What make and model is it?

 

If the PSU is old there’s an argument for replacing it anyway in general principal.  Old PSUs aren’t as safe as the more modern ones and can go bad even faster, so anything more than about 5 years old is arguably old enough to change out.  There’s a widely variable exception for prebuilts.

See original post for model and I would say it’s probably at least 4 yrs old

[Bombastinator]

So I see.  With a 20 pin connector I’d say more like 10.  Fair argument for just replacing it even if it’s not the problem.  It needs to be done soon if not now.

[d00m1ord]

28 minutes ago, Bombastinator said:

So I see.  With a 20 pin connector I’d say more like 10.  Fair argument for just replacing it even if it’s not the problem.  It needs to be done soon if not now.

It’s definitely not 10 because I brought a new one 4-5 yrs ago because the one that came in the case died on me so I had to find a replacement so I’ve only had it for 4-5 yrs and i got it new but it may well need replacing it’s just the issue of sourcing a PSU that fits the case

[Bombastinator]

But 20 pin?  Not 20+4 but 20

[d00m1ord]

16 minutes ago, Bombastinator said:

But 20 pin?  Not 20+4 but 20

Yes 20 pin and a 4 pin for the cpu not 20+4 or 24

[Bombastinator]

52 minutes ago, d00m1ord said:

Yes 20 pin and a 4 pin for the cpu not 20+4 or 24

Huh.  I was seeing 20+4 back in 2006 and it was considered a kind of throwback thing even then.