Question Desktop PC fails to boot after sleep, even after multiple part replacements

[Roundbaby]

Hi,

My brother has a Windows 10 Desktop PC with the following specs:

MSI RTX 4060Ti Ventus 3x
MSI B550 Gamer Plus Motherboard (Previous Motherboard: MSI B365M Mortar)
AMD Ryzen 7 5800X (Previous CPU: Intel i7 9700KF)
32GB GSKILL Ripjaw 3200MHz
Thermaltake 600W Smart Series
Random Wifi Card
1TB NVME M.2 Samsung
1TB SSD

He put the computer to sleep overnight with the previous components. In the morning it didn't boot and there was no response or lights from the motherboard. We immediately assumed it was an issue with the PSU, so we did the paperclip test and used my multimeter to check the voltages. All the voltages checked out and it was able to drive the case fans, so we assumed the PSU was good. The logical solution was that either the motherboard or CPU was bad. Long story short, we ended up replacing both with the new components in the above list. After going through all the work of rebuilding it, we are getting the exact same behavior -- no motherboard lights, no boot, PSU ostensibly fine. Our last ditch effort is to replace the PSU, but we wanted to get a second opinion on this since it will be quite painful to re-cable-manage the entire PC. Can the PSU report "power good" and reasonable voltages on each and every pin of the 24-pin motherboard connector and be nonfunctional? Please let us know, because we've been pulling our hair out about this!

Thanks,
-Jacob

 

[Ralston18]

Was Windows reinstalled after the upgrades or did you just transfer in the boot drive?

How were the PSU voltages measured?

E.g.:

https://www.lifewire.com/how-to-manually-test-a-power-supply-with-a-multimeter-2626158

Is the M.2 the boot drive?

Will the system boot into safe mode?

Have you tried disconnecting the SSD?

 

[Roundbaby]

Was Windows reinstalled after the upgrades or did you just transfer in the boot drive?

How were the PSU voltages measured?

E.g.:

https://www.lifewire.com/how-to-manually-test-a-power-supply-with-a-multimeter-2626158

Is the M.2 the boot drive?

Will the system boot into safe mode?

Have you tried disconnecting the SSD?

As I described in the thread, the computer does not POST -- the motherboard lights do not even turn on. No fans spin, completely dead PC. It would be impossible to reinstall windows or even get into the BIOS. I am very familiar with a multimeter as I am an electrical engineer by trade. I tested the voltages in the pins of the motherboard power connector after shorting together pins 16 and 17 together and have tested each regulated voltage (3.3, 5, 12, -12) to be within spec. The M.2 is not the boot drive, and the system will not boot into safe mode. Could you tell me why you think disconnecting the SSD will have any effect on the motherboard receiving power?

 

[stonecarver]

As I described in the thread, the computer does not POST -- the motherboard lights do not even turn on. No fans spin, completely dead PC.

Being as this is basically two systems that won't even show signs of life.

So have to think of reasons for that no matter how stupid it might sound.

Is the wall outlet working ? Yes you said you tested power supply but I have no idea if you tested it on the same outlet. Trying to rule stuff out.

Have you tried to jump motherboard pins for power and not use the case wiring for power button.

An easy test is use his old board, power supply. and memory slap it all together on a table and jump the power pins with a small screw driver. Can it get to bios.

 

[DSzymborski]

Do you have a known good PSU to test with? Multimeter isn't really that useful a tool here other than confirming extremely basic functionality, and the PSU is cheap group-regulated junk anyway, so ought to be replaced under any and all circumstances.

 

[Ralston18]

SSD being boot (C and M.2 is then being used for data - correct?

I thought it was the opposite.... My error.

In any case, removing non critical components (e.g. a second drive) is one way of eliminating potential culprits or contributors to some problem.

Strange that the PSU voltages tested in spec (reference being the above cited Lifewire link) but the motherboard(s) being completely powerless.

Where in the electrical paths/schematics are there any "single point of failure" that would result in two motherboards/systems failing to receive any power at all?

Wall outlet, power cable to PSU, the PSU itself (which does not explain the measured voltages) would meet that criteria.

Maybe a break or short.

Per @stonecarver and @DSzymborski posts I suggest going back and doublechecking everything.

And I will toss in the suggestion to install a new CMOS battery.

 

[Roundbaby]

Being as this is basically two systems that won't even show signs of life.

So have to think of reasons for that no matter how stupid it might sound.

Is the wall outlet working ? Yes you said you tested power supply but I have no idea if you tested it on the same outlet. Trying to rule stuff out.

Have you tried to jump motherboard pins for power and not use the case wiring for power button.

An easy test is use his old board, power supply. and memory slap it all together on a table and jump the power pins with a small screw driver. Can it get to bios.

Hi,

Yes we have tried 2 different wall outlets and 2 different power cables that we know are rated for the current we should be drawing. We have also tried jumping the reset-sw motherboard jumpers with a piece of solder, still nothing. Our main theory at this point is that the power supply is able to produce reasonable voltages under little/no load (the paperclip test), but maybe it can't produce enough current to drive the entire system at boot. This would result in the motherboard not even attempting to startup, but I would at least also expect motherboard status lights BEFORE startup (like RGB and whatnot) which we are not getting.

 

[Roundbaby]

Do you have a known good PSU to test with? Multimeter isn't really that useful a tool here other than confirming extremely basic functionality, and the PSU is cheap group-regulated junk anyway, so ought to be replaced under any and all circumstances.

Hi,

yes these are some of my thoughts exactly. we've already spent around $350 on replacement parts already, so we're trying to get other's opinions before we shell out another $80 for yet another new part. This will likely be our next course of action. I'm just worried that if replacing the PSU will have no effect as well, the only possible solution at that point would be a cursed spirit haunting the PC and we'll need to get an exorcist.

 

[Roundbaby]

SSD being boot (C and M.2 is then being used for data - correct?

I thought it was the opposite.... My error.

In any case, removing non critical components (e.g. a second drive) is one way of eliminating potential culprits or contributors to some problem.

Strange that the PSU voltages tested in spec (reference being the above cited Lifewire link) but the motherboard(s) being completely powerless.

Where in the electrical paths/schematics are there any "single point of failure" that would result in two motherboards/systems failing to receive any power at all?

Wall outlet, power cable to PSU, the PSU itself (which does not explain the measured voltages) would meet that criteria.

Maybe a break or short.

Per @stonecarver and @DSzymborski posts I suggest going back and doublechecking everything.

And I will toss in the suggestion to install a new CMOS battery.

Hi,

Yes you are correct, the M.2 is data and the SSD is the boot drive. I should have included that detail in the spec list. You can see one of my above responses to see that we tried multiple wall outlets and power cables that are rated for the current we are drawing, and still no avail. This essentially just leaves the PSU. I can only imagine maybe the issue is the PSU can no longer supply the current step required for a startup so it doesn't even try. It might be able to drive a few fans and produce the right voltages under no load, but it's driving power has deteriorated to the point that it cannot perform a startup. Still, we would expect some motherboard lights or status lights after switching on the power switch on the PSU if this was the case, since that is just drawing marginal power.

 

[Ralston18]

For some reason this makes me wonder:

" We have also tried jumping the reset-sw motherboard jumpers with a piece of solder"

Solder with no heat applied - correct?

Versus just using a bare copper wire.....

= = = =

When the PSU voltages were tested: The PSU was completely disconnected from the host computer and all the measured voltages (3, 5, and 12) were within tolerance correct?

Yet upon connecting the PSU to the motherboard etc. the PSU is completely (or appears completely) dead.

The remaining option, as I understand the situation, is to swap in a known working test PSU using only the cables that come with that PSU.

For reference purposes check the voltages on the swapped in test PSU.

Also before swapping in the test PSU inspect the motherboards for signs of damage: swollen capacitors, blackened or browned areas.

I would start with the original motherboard and temporarily set aside any immediate cable management efforts. However, be sure to inspect all cables for signs of damage, kinks, pinches, bare conductor showing, melted insulation, etc..

Objective being to simply get the original motherboard up and running if possible.

Failing that then try the second motherboard with the swapped in PSU.

Overall a very methodical, careful, and thorough "do over" keeping it all as simple as possible.

Hopefully the PC will work again or a specifc culprit will be found.

 

[Roundbaby]

For some reason this makes me wonder:

" We have also tried jumping the reset-sw motherboard jumpers with a piece of solder"

Solder with no heat applied - correct?

Versus just using a bare copper wire.....

= = = =

When the PSU voltages were tested: The PSU was completely disconnected from the host computer and all the measured voltages (3, 5, and 12) were within tolerance correct?

Yet upon connecting the PSU to the motherboard etc. the PSU is completely (or appears completely) dead.

The remaining option, as I understand the situation, is to swap in a known working test PSU using only the cables that come with that PSU.

For reference purposes check the voltages on the swapped in test PSU.

Also before swapping in the test PSU inspect the motherboards for signs of damage: swollen capacitors, blackened or browned areas.

I would start with the original motherboard and temporarily set aside any immediate cable management efforts. However, be sure to inspect all cables for signs of damage, kinks, pinches, bare conductor showing, melted insulation, etc..

Objective being to simply get the original motherboard up and running if possible.

Failing that then try the second motherboard with the swapped in PSU.

Overall a very methodical, careful, and thorough "do over" keeping it all as simple as possible.

Hopefully the PC will work again or a specifc culprit will be found.

Hi,

Yes, just a cold piece of solder because we didn't have any jumpers lying around. All voltages were within tolerance, but we still had some cables connected to the host PC such as the case fans and SSD power. We unplugged the CPU and motherboard power and tested just the voltages in the motherboard 24 pin connector. All voltages were within 2% tolerance. Once reconnecting all connectors and reapplying power, nothing lights up and there's no way to start up the PC. One thing that we haven't tried (out of fear of possibly bricking the rest of the system) was to apply the paperclip to pins 16-17 WHILE the motherboard is plugged in to force a power up. We don't think this is safe so we're tabling it for now. I did a thorough visual inspection of the old motherboard when we first removed it and couldn't find any signs of damage, but it was fairly dusty, so I mainly checked the large components and sockets. I haven't checked all of the cabling since the cable management is very tight and zip tied, so there could be a kink somewhere I can't see. I suppose once I get my hands on a new PSU and will have to redo all the cabling anyways, any potential kinks on the old wires will be immediately apparent.

 

[Ralston18]

Cables, zip ties, connections:

If possible, take a couple of photographs beforehand and post here via imgur (www.imgur.com).

Someone might spot something.

 

[Roundbaby]

Problem solved, and it ended up being user error. The motherboard WAS defective, but we had slightly misconnected the front panel jumpers on the new motherboard. Still surprised that there wasn't even at least motherboard lights to show sign of life when power was switched on the PSU. Thanks for everyone's help on the matter!