Computer will not power on.

[Dia]

I've built this computer in the fall of 2013 and everything has been working great up until a couple months. The issue I'm having is that my computer just will not boot sometimes after I shut it down. I've tried to troubleshoot it a couple time and have not found the problem. I'm pretty sure the components in question is the MotherBoard and PSU. These are the things I've tried:

 

1) tried to boot the computer by shorting the power pins on motherboard

2) reseated the ram

3) tried to boot with 1 ram stick (tested all four slots and all four ram modules)

4) removed video card

5) reseated the video card

6) took apart the whole computer and cleaned up any wire clutter if there was any

 

I didn't think the motherboard was an issue because if the computer does turn on, it runs fine. Nothing is wrong once the computer is booted up. I've even ran mem tests and prime 95.

I suspect the PSU is the issue.

7) I did the PSU green wire test and the fan starts to spin fine

 

Conclusion:

ram is good

psu if good

motherboard is good

I've even checked for any bulging capacitors on my motherboard, psu and graphics card.

 

The only thing that does work to boot my computer is to unplug the power supply and leave it unplugged for at least 1-2hrs. I can't think of a reason as to why this works.

 

Specs:

i7 4770k (cooled with a noctua cpu cooler)

msi g45 gaming motherboard

HD 7850

32gb g.skill ram

seasonic 550 g-series gold certified psu

 

 

 

 

[Djole123]

Did you overclock it?

[KSores]

Kick it real hard

 

JK

I had the exact  same issue

I got a new motherboard and it was fine 

 

And for some reason a different port on my graphics card made it better too

[pinksnowmanshandle]

4 minutes ago, KSores said:

Kick it real hard

I would've said did they try turning it off and on again. Though I suppose that works too. Might work better since they're having an issue turning it on anyways. 

Seriously though, does it at least post? Or does it have one of those Q Code displays or whatever?
If you can get it go turn on some how, if you have an overclock try turning that OC off.

[pinksnowmanshandle]

Also, have you tried bridging the pwr sw pins with out using the power button on the case.

[Naeaes]

12 minutes ago, Dia said:

The only thing that does work to boot my computer is to unplug the power supply and leave it unplugged for at least 1-2hrs. I can't think of a reason as to why this works.

I remember dealing with a batch of Asus prebuilts with the same exact issue and the same exact fix. It turned out to be a bad PSU. Well, to be honest, an OK-ish PSU that wasn't fully compatible with Finnish AC power grid. Among the most annoying issues to verify, the stupid things were huge and there were like fourty of them. The customers thought I was crazy/incompetent when I told them to unplug and leave it for the night to temporarily fix it.

 

Anyhoo. Don't rule out the PSU quite yet. The wire trick doesn't put much strain on it. It's not unheard of that a bad PSU can still spin the 0,5W fan but not power the 50W CPU. 

[Dia]

1 hour ago, Djole123 said:

Did you overclock it?

I did before, it was only at 4.0 ghz. I've since reverted that a couple months ago.

1 hour ago, wcreek said:

I would've said did they try turning it off and on again. Though I suppose that works too. Might work better since they're having an issue turning it on anyways. 

Seriously though, does it at least post? Or does it have one of those Q Code displays or whatever?
If you can get it go turn on some how, if you have an overclock try turning that OC off.

I can get it to boot. The only way I can do that is to leave the PSU unplugged for 1-2hrs. After that everything runs fine, even if I leave my computer on for a week.

1 hour ago, wcreek said:

Also, have you tried bridging the pwr sw pins with out using the power button on the case.

Yes I have

53 minutes ago, Naeaes said:

I remember dealing with a batch of Asus prebuilts with the same exact issue and the same exact fix. It turned out to be a bad PSU. Well, to be honest, an OK-ish PSU that wasn't fully compatible with Finnish AC power grid. Among the most annoying issues to verify, the stupid things were huge and there were like fourty of them. The customers thought I was crazy/incompetent when I told them to unplug and leave it for the night to temporarily fix it.

 

Anyhoo. Don't rule out the PSU quite yet. The wire trick doesn't put much strain on it. It's not unheard of that a bad PSU can still spin the 0,5W fan but not power the 50W CPU. 

I don't think my PSU has any compatibility issues. I believe Canadians have the same power standard as Americans. The thing is everything runs 100% fine once I get my computer booted. I've even had it running idle for a week and there were no problems. The problem is either the motherboard/psu and I don't have the means to get another PSU to test if that is actually the issue.

 

Do you know any other tests I'm able to do with the PSU without testing a new one?

[bgibbz]

 

1 hour ago, Dia said:

Do you know any other tests I'm able to do with the PSU without testing a new one?

Do you have acces to another system of the same form factor? if a friend has a system you can try hooking your PSU up and seeing if it will boot.

[Naeaes]

10 hours ago, Dia said:

I did before, it was only at 4.0 ghz. I've since reverted that a couple months ago.

I can get it to boot. The only way I can do that is to leave the PSU unplugged for 1-2hrs. After that everything runs fine, even if I leave my computer on for a week.

Yes I have

I don't think my PSU has any compatibility issues. I believe Canadians have the same power standard as Americans. The thing is everything runs 100% fine once I get my computer booted. I've even had it running idle for a week and there were no problems. The problem is either the motherboard/psu and I don't have the means to get another PSU to test if that is actually the issue.

 

Do you know any other tests I'm able to do with the PSU without testing a new one?

Yeah. Finland has the same 230VAC@50Hz as the rest of the Europe and the network is relatively high in quality too. We figured it had more to do with the waveform. Maybe it wasn't close enough to perfect sine but had interference in it. That stuff can be caused by super random stuff like the weather. Or by a refrigerator or a freezer when they suddenly ramp up. Simple electronic devices like light bulbs and motors don't give a crap if there's interference and any machine with a battery is also gonna be just dandy. But a desktop PC is super allergic to interference. Still, it'd be only by happenstance if this was your issue. And any aftermarket PSU is always going to be more tolerant to jitter than an OEM part. I only mean, AC isn't perfect.

 

Back then I made a simple batch file that appends the time and date into a text file and then reboots the computer. Then I placed that in the startup folder of the computer being tested and removed the login password query. That puts the machine in a continuous bootloop. It enters windows and reboots right away. Then I just left it. Sometimes it took over a week and thousands of reboots but sooner or later any unit I tested, acted up. I was able to tell when exactly the issue had arisen by looking at the timestamps I logged with the batch file. I figured, if it's at a specific time of day each time, it'd help diagnose, but it never was. If you have the time, the same method might work for you too. Once you, think you've fixed the issue, do this to verify.

Those machines had Win 7 in them. Win 10 doesn't have the startup folder in the startup menu anymore but there's a folder in %appdata% that does the same thing. You might want to leave the /t parameter out from the shutdown/r to give yourself the minute to remove the loop batch. I myself had /t 10 in it to make the timeout ten seconds. I can tell you, that's nowhere near enough to navigate to the correct folder and delete the file. It'll also help if you place the batch on the desktop for easy removal and only place a shortcut to it in the startup folder. 

 

Edit: whoa! Sorry for the wall of text.  I don't have the batch file anymore and I can't remember the exact code. But it was really simple, just two lines. If you feel like it, I can help you create your own.

[Devin92]

I have some similar if not the same problem just recently. Once my rig turns on it works fine. But my problem was if I shut it down for a day or so, it wont start up. I took my pc apart and put everything back and the problem went away. Some good ppl form this forum suggest it may be some where short circuiting (which may be justify why putting everything apart and together again will work). When you PC wont boot, do you mean no fan turning no LED light up, or fan running but system wont post? Do you have a HDD? The thing with HDD is at the start up stage it drain more power from PSU and that may explain why it wont start but once started it works fine.

 

Here is the link to my issue may help you.

 

[Inimigor]

If you can get a multimeter/voltimeter try measuring the PSU current and it's variation.

It might be off the tolerance of the computer or oscilating too much, that has happened to my friend's AX860i, the PSU was faulty and it killed the MoBo voltage regulators. Or maybe it could be some capacitors giving up to the ghosts. If you have no warranty left, take it to somewhere that fixes PSU's, TV's and things like that, they can check your PSU for faulty components.

!!! WARNING: DO NOT OPEN THE POWER SUPPLY, IF YOU SHORT SOMETHING YOU CAN DIE !!!!!

Did/Do you have any coil whining on the computer?

If voltages are fine, and you tested all other components and they are fine, it might be the motherboard itself that is faulty.

[Dia]

13 hours ago, Naeaes said:

Yeah. Finland has the same 230VAC@50Hz as the rest of the Europe and the network is relatively high in quality too. We figured it had more to do with the waveform. Maybe it wasn't close enough to perfect sine but had interference in it. That stuff can be caused by super random stuff like the weather. Or by a refrigerator or a freezer when they suddenly ramp up. Simple electronic devices like light bulbs and motors don't give a crap if there's interference and any machine with a battery is also gonna be just dandy. But a desktop PC is super allergic to interference. Still, it'd be only by happenstance if this was your issue. And any aftermarket PSU is always going to be more tolerant to jitter than an OEM part. I only mean, AC isn't perfect.

 

Back then I made a simple batch file that appends the time and date into a text file and then reboots the computer. Then I placed that in the startup folder of the computer being tested and removed the login password query. That puts the machine in a continuous bootloop. It enters windows and reboots right away. Then I just left it. Sometimes it took over a week and thousands of reboots but sooner or later any unit I tested, acted up. I was able to tell when exactly the issue had arisen by looking at the timestamps I logged with the batch file. I figured, if it's at a specific time of day each time, it'd help diagnose, but it never was. If you have the time, the same method might work for you too. Once you, think you've fixed the issue, do this to verify.

Those machines had Win 7 in them. Win 10 doesn't have the startup folder in the startup menu anymore but there's a folder in %appdata% that does the same thing. You might want to leave the /t parameter out from the shutdown/r to give yourself the minute to remove the loop batch. I myself had /t 10 in it to make the timeout ten seconds. I can tell you, that's nowhere near enough to navigate to the correct folder and delete the file. It'll also help if you place the batch on the desktop for easy removal and only place a shortcut to it in the startup folder. 

 

Edit: whoa! Sorry for the wall of text.  I don't have the batch file anymore and I can't remember the exact code. But it was really simple, just two lines. If you feel like it, I can help you create your own.

Thanks for the tip. I'll definitely use your method to test once I think everything is in working condition. I'll hit you up on that offer if I can't write it myself

6 hours ago, Devin92 said:

I have some similar if not the same problem just recently. Once my rig turns on it works fine. But my problem was if I shut it down for a day or so, it wont start up. I took my pc apart and put everything back and the problem went away. Some good ppl form this forum suggest it may be some where short circuiting (which may be justify why putting everything apart and together again will work). When you PC wont boot, do you mean no fan turning no LED light up, or fan running but system wont post? Do you have a HDD? The thing with HDD is at the start up stage it drain more power from PSU and that may explain why it wont start but once started it works fine.

 

Here is the link to my issue may help you.

 

The system does not do anything. No fans spin and no LEDs turn on. I've tried taking apart everything and putting it back together, but that didn't help. I have a SSD and HDD, I'll try unplugging the HDD to test if it'll boot.

 

Thank you!

3 hours ago, Inimigor said:

If you can get a multimeter/voltimeter try measuring the PSU current and it's variation.

It might be off the tolerance of the computer or oscilating too much, that has happened to my friend's AX860i, the PSU was faulty and it killed the MoBo voltage regulators. Or maybe it could be some capacitors giving up to the ghosts. If you have no warranty left, take it to somewhere that fixes PSU's, TV's and things like that, they can check your PSU for faulty components.

!!! WARNING: DO NOT OPEN THE POWER SUPPLY, IF YOU SHORT SOMETHING YOU CAN DIE !!!!!

Did/Do you have any coil whining on the computer?

If voltages are fine, and you tested all other components and they are fine, it might be the motherboard itself that is faulty.

I don't have a multimeter/voltimeter. If all else fails and I've determined it to be the power supply, I'll definitely go take it somewhere to service it. Or I might just replace it : p