Thread: PSU stuffed?

The plot gets thicker, and sounding more expensive. When I unplug the CPU power from the motherboard, the motherboard stays running and the PSU doesn’t trip...

Obviously nothing else happens then. What’s the chances of a CPU or related component on the motherboard failing? Or could it be the power supply could still be toast and doesn’t have enough grunt to get the CPU away?

I don't have time to go through them now, but this guy posted a vid with something similar. It was a simple fix. Not saying yours is the same. He knows his stuff though. If you have some time have a look.

Originally Posted by Homestar

.....

Obviously nothing else happens then. What’s the chances of a CPU or related component on the motherboard failing? Or could it be the power supply could still be toast and doesn’t have enough grunt to get the CPU away?

I'd say 'less likely' .. to be CPU.

Had a quick squizz, it seems you have the ASUS X99 Pro/USB3.1 motherboard.
If so, you have Q Code on the board.
Helps to assess errors.
The Q Code on your board is D0.
Looking at the table for the X99 Pro/USB3.1 board, CPU issues should show some values in the Q code readout.
According to ASUS, 00 error = 'not used', so what appears to be 00, must be D0(stupid segmented LEDs make it hard to decipher clearly)
This D0 code = CPU initialization error.

Could be that PSU has lost the ability to supply CPU power.
The other voltage rails may be working fine, just that when the CPU draws more and more power, maybe the 3v rail in PSU has shot itself, and can't provide .. hence CPU initialization error.

Something worth a quick look, can't do harm to try, is on the board there is a set of pins(3) marked CPU_OV. Down towards the RHS lower corner, half way between the Q code readout and the corner, and about an inch up from the lower edge.
By default, it should be set to pins 1 and 2. If not, maybe that could be the issue(who built the PC?). If it's still set to default, maybe try to set it to overvolt, ie. set the jumper to pins 2 and 3.
If it makes a difference(even just a single start, and shutdown event) .. then I'd say for sure PSU's CPU 3v rail is gone.

If there was a way you could test the CPU rail power, you'd know for sure.
That is, pull the CPU power cable, and attach a 3v load that allows you to vary the power draw to load the PSU's CPU rail more and more.

There's a jumper trick where you can pull the PSU, use a jumper pin to get it to start off the mainboard and test the CPU rail under load.
(can never remember the pins to jump tho).

Originally Posted by johntins

I don't have time to go through them now, but this guy posted a vid with something similar. It was a simple fix. Not saying yours is the same. He knows his stuff though. If you have some time have a look.

Originally Posted by johntins

Thanks, but that’s not the same issue - mine won’t even bring the fans on for more than the same time anything else happens. If I got the CPU and motherboard running, I’d have half a chance.

Originally Posted by AK83

I'd say 'less likely' .. to be CPU.

Had a quick squizz, it seems you have the ASUS X99 Pro/USB3.1 motherboard.
If so, you have Q Code on the board.
Helps to assess errors.
The Q Code on your board is D0.
Looking at the table for the X99 Pro/USB3.1 board, CPU issues should show some values in the Q code readout.
According to ASUS, 00 error = 'not used', so what appears to be 00, must be D0(stupid segmented LEDs make it hard to decipher clearly)
This D0 code = CPU initialization error.

Could be that PSU has lost the ability to supply CPU power.
The other voltage rails may be working fine, just that when the CPU draws more and more power, maybe the 3v rail in PSU has shot itself, and can't provide .. hence CPU initialization error.

Something worth a quick look, can't do harm to try, is on the board there is a set of pins(3) marked CPU_OV. Down towards the RHS lower corner, half way between the Q code readout and the corner, and about an inch up from the lower edge.
By default, it should be set to pins 1 and 2. If not, maybe that could be the issue(who built the PC?). If it's still set to default, maybe try to set it to overvolt, ie. set the jumper to pins 2 and 3.
If it makes a difference(even just a single start, and shutdown event) .. then I'd say for sure PSU's CPU 3v rail is gone.

If there was a way you could test the CPU rail power, you'd know for sure.
That is, pull the CPU power cable, and attach a 3v load that allows you to vary the power draw to load the PSU's CPU rail more and more.

There's a jumper trick where you can pull the PSU, use a jumper pin to get it to start off the mainboard and test the CPU rail under load.
(can never remember the pins to jump tho).

Thanks for that - Yes it’s the x99 pro board - jumper was between 1 and 2. I changed it to 2 and 3 and it powered up long enough to start windows booting - much longer than any other boot I’ve managed and I thought it was actually going to work for a second, but then failed to chooch. It did show the CPU is doing what something and the also the solid state drive that windows is on is also working. Does that confirm your 3 volt rail theory/ stuffed PSU from your point of view? Based on it trying to boot, I’m much happier knowing that the other bits seem to be working still.

I don’t have anything I can load it with unfortunately. Oh, I built the computer some years ago, when all this stuff was considered fairly good. My Son used to use it as a gaming rig, but after that I inherited it and it now does not much apart from some occasional video editing which it works quite well doing, so if I can get it to live again, I’ll hopefully get a few more years from it.

If it is the power supply, I’ll be a bit ticked to be honest as that was supposed to be a decent brand, and when I cleaned it out, noticed they are using Nippon Chemi-con caps, which cheap units certainly don’t use. Might be better off just getting a cheap arse one if they only last a few years like this - and it was overkill to the max - 1000 watt so it’s not like it got a flogging or anything - was on special and cheaper than the 600 watt version as it had just been superseded.

The real late model PC's hardly need any power,, most will run happily on a 500w jobby.. buy a 2nd hand $50 dollar one and test...
the real buying point is the CPU power connector pin number.....


and my last spin stop issue was a bad start button.... used the reset and all good...

Originally Posted by Homestar

Thanks, but that’s not the same issue - mine won’t even bring the fans on for more than the same time anything else happens. If I got the CPU and motherboard running, I’d have half a chance.

.
Sorry, Gav. I wasn't suggesting that that vid was the same issue, I was pointing you to him because I'm sure I saw him deal with a similar one, but I couldn't tale the time to find it. He has many vids on line. However, they are long.

Yeah, if she runs on the boosted voltage, then it's 99.9% PSU problem.
(still a slight chance some weird stuff happening on the CPU .. wouldn't know enough about something like that tho)

You haven't said what peripherals you're running, but I doubt you'd need 1000W for most standard PCs.
eg. how many HDDs do you have connected at any one time, incl USB or other externally connected.

I'm currently on an AMD FX CPU(monster power hungry CPU), but more importantly 5 HDDs internal, and 1 USB ext plus two SSDs(one 2.5" SATA, and one M.2) .. I worked out 800W was all I really needed .. ended up with a Corsair 850W.

Recently built son his Ryzen 7, only used a 650W PSU .. runs no worries .. just the 1Tb M.2 SSD. Silverstone 550W model .. ~ $100 (give or take 5). Fan is a little louder than I'd have liked, but son isn't fussed by it.
If you were running dual GPUs and lots of HDDs, then yeah .. 1000W probably makes sense.

Also: looks like yours has liquid cooling? Never used one no idea how much power they draw, and on which rail .. so may make a difference.