Urgent: PC wont turn on with PSU connected to MOBO

[STSMiner]

@Naifamar
try, but if it doesnt help, see if you can borrow another psu, just to able to exclude it,
as it basically only leaves the board as cause.

I was experiencing the same issue the OP of this thread is having, took me two days to track down the issue.
It was not until I stumbled on that video I posted (post #11) that thought ... "Hmm ! Maybe that's the issue !"

Yes, I tested with the CMOS battery out, PC booted just fine, but not with the CMOS battery in.

New battery for the CMOS solved the issue in the end, this is why I suggested in post #2 to replace the CMOS battery with a new one to rule that out.

Sometimes it's the little things that get "overlooked" !

 

[nomdeplume]

Today I tried to power on and it worked again one time, then not powering, I think this should be faulty battery ? Will buy new today

Make sure to follow your mobo procedure on shorting pins and amount of time full procedure requires. It's a pain but I'd consider retraining RAM as well.

I spent a solid week isolating an issue recently that required 15 minutes to reset after every failure.

 

[LabRat 891]

I was experiencing the same issue the OP of this thread is having, took me two days to track down the issue.
It was not until I stumbled on that video I posted (post #11) that thought ... "Hmm ! Maybe that's the issue !"

Yes, I tested with the CMOS battery out, PC booted just fine, but not with the CMOS battery in.

New battery for the CMOS solved the issue in the end, this is why I suggested in post #2 to replace the CMOS battery with a new one to rule that out.

Sometimes it's the little things that get "overlooked" !

a 'dead' (or defective CMOS batt.) can and will cause issues that a merely "absent" CMOS batt. will not. [My favorite: 'whacky' RTC behavior]

However, OP's symptoms are more in-line w/ my experiences w/ a short.

Last time I had something similar happen, I was foolishly trying out a server proprietary PCIe riser.
Turned out, Supermicro had re-routed power and ground pins, causing a short.

I've also seen NICs cause this when they've catastrophically failed, and all sorts of AICs when dust managed to get into the slot.

- Reseat cards, giving ea. slot the good 'ole (S)NES Blowjob
- Remove CMOS batt, reset CMOS.
- Check for shorts (bench test mobo, min. HW installed)
- Test using another PSU
- Re-seat CPU, do *not* overtighten ['have caused no POST from a warped socket/board]

 

[Fourstaff]

How do I check for short? With multimeter?

Check the motherboard if its touching something it shouldn't. Loose screws if the board is still mounted, or the board is not touching anything conductive if you are laying it on the table.

Today I tried to power on and it worked again one time, then not powering, I think this should be faulty battery ? Will buy new today

This does sound like a PSU failing problem though.

 

[Naifamar]

I was experiencing the same issue the OP of this thread is having, took me two days to track down the issue.
It was not until I stumbled on that video I posted (post #11) that thought ... "Hmm ! Maybe that's the issue !"

Yes, I tested with the CMOS battery out, PC booted just fine, but not with the CMOS battery in.

New battery for the CMOS solved the issue in the end, this is why I suggested in post #2 to replace the CMOS battery with a new one to rule that out.

Sometimes it's the little things that get "overlooked" !

Replaced CMOS battery, not working, it also worked today morning, came back, not working, its like casino

a 'dead' (or defective CMOS batt.) can and will cause issues that a merely "absent" CMOS batt. will not. [My favorite: 'whacky' RTC behavior]

However, OP's symptoms are more in-line w/ my experiences w/ a short.

Last time I had something similar happen, I was foolishly trying out a server proprietary PCIe riser.
Turned out, Supermicro had re-routed power and ground pins, causing a short.

I've also seen NICs cause this when they've catastrophically failed, and all sorts of AICs when dust managed to get into the slot.

- Reseat cards, giving ea. slot the good 'ole (S)NES Blowjob
- Remove CMOS batt, reset CMOS.
- Check for shorts (bench test mobo, min. HW installed)
- Test using another PSU
- Re-seat CPU, do *not* overtighten ['have caused no POST from a warped socket/board]

Its working like 1:50 of all time. Today it started working everytime I short pins, now its not working at all, Idk its like casino, what’s happening

Check the motherboard if its touching something it shouldn't. Loose screws if the board is still mounted, or the board is not touching anything conductive if you are laying it on the table.


This does sound like a PSU failing problem though.

I tried to test 24 PIN with multimeter, every pin seems to give normal voltage,

 

[Fourstaff]

I tried to test 24 PIN with multimeter, every pin seems to give normal voltage,

The PSU can probably provide power for the multimeter but may not have enough power to turn the rig on. I had a similar problem before (PC randomly start/not start as it wished), it went away after I replaced the PSU. You may be facing a similar issue. This is a cat and mouse diagnostic game though, buying a new PSU is quite a bit of money and it may not fix the problem.

 

[Naifamar]

The PSU can probably provide power for the multimeter but may not have enough power to turn the rig on. I had a similar problem before (PC randomly start/not start as it wished), it went away after I replaced the PSU. You may be facing a similar issue. This is a cat and mouse diagnostic game though, buying a new PSU is quite a bit of money and it may not fix the problem.

Would you advise going to repair shop so they can plug new psu and check? It would be easier than buying psu and returning it back?

 

[Fourstaff]

Would you advise going to repair shop so they can plug new psu and check? It would be easier than buying psu and returning it back?

If you have a convenient repair shop nearby, I guess you can give it a try. But if they charge you $50, you might as well get a new PSU. Just make sure its a good make with enough wattage (Seasonic, most Corsairs, Silverstone, Superflower, some CoolerMasters, Antec)

 

[Naifamar]

If you have a convenient repair shop nearby, I guess you can give it a try. But if they charge you $50, you might as well get a new PSU. Just make sure its a good make with enough wattage (Seasonic, most Corsairs, Silverstone, Superflower, some CoolerMasters, Antec)

Thanks, before going to shop wanted to ask you, I plugged power out for 1 hour, and PC started, then it restarted after 5 seconds. I now more and more convinced about bad PSU, what do you think of PC restarting after 5 seconds?

 

[nomdeplume]

I think we could all agree further booting with a suspected faulty PSU is not advisable.

 

[STSMiner]

Replaced CMOS battery, not working, it also worked today morning, came back, not working, its like casino


Its working like 1:50 of all time. Today it started working everytime I short pins, now its not working at all, Idk its like casino, what’s happening


I tried to test 24 PIN with multimeter, every pin seems to give normal voltage,

OK, now that the CMOS battery (cheap route first) has been ruled out properly it's now time to replace the PSU and all the cables.

Don't use the old cables with the new PSU, swap every cable out with the new cables that comes with the new PSU.
Side note .. This still might not solve the problem, it could be something else that we've all overlooked.