Question Did my GFCI brick my PC

[CMOS89]

Hi All,

After lurking for sometime viewing other posts similar to my situation I have decided to create an account and see if anyone can give me some thoughts on my issue.

Sometime ago I moved my small form factor PC into my daughters room, the purpose of the move was to provide some extra heat during these cold winter months where I live(USA) while running some folding@home to support a good cause. Initially the pc worked without issue for several weeks than one day I noticed it was off one morning and the GFCI outlet the PC was plugged into was tripped(red light on the outlet) Not thinking much about it I reset the outlet and turned the PC back on. As a bit of a backstory this GFCI is about 1 year old, we had some renovations done that involved this room about a year ago and the previous two prong outlet was replaced with a GFCI as a result(as far as I know its still on an ungrounded circuit, 1950s home) So another few weeks go by and this same thing occurs again, I reset the outlet and start up the PC with no issue. This happens maybe once or twice more over the course of about a month except on the most recent incident, after resetting the outlet and PCs power button the computer fails to post. PSU turns on and LEDS come on for all components however the Motherboard shows DRAM light on the EZ debug indicator. Below are PC specs and what I tried so far.

MOBA: MSI B560i
CPU: Intel I5-11500
RAM: Gskill DDR4 3200MHz
GPU: RTX 3070Ti
PSU: Corsair SF600
Storage: 1tb Samsung 970 EVO M.2

1. Reseated the RAM
2. Removed all but one stick
3. Cleared CMOS

None of this worked so I contacted MSI thinking it was just a faulty MOBA, still under warranty. The board was sent off and I received it back just recently with the statement from MSI that it was "repaired" As of yesterday I put everything back together thinking this was the fix and now I get a CPU error and still no post, here is what I have tried so far after the first RMA

1. Checked CPU, removed from socket and confirmed no bent pins or improper install. All looked good.
2. Removed and repasted AIO Cooler(deepcool 240)
3. Double and triple checked AIO wiring
4. Installed stock Intel CPU cooler
5. Checked m.2 boot drive mounting
6. Removed CMOS battery and waited 5-10 minutes
7. Checked voltage of CMOS battery with multimeter(was at 3.0volts)

None of this resulted in a PC that posted so as I am contemplating another RMA should I instead be considering that my computer was damaged by some fault event occurring while plugged into this GFCI outlet?

 

[Lutfij]

Welcome to the forums, newcomer!

Did you try powering the system without the discrete GPU? You have a K suffix processor which would allow you to power off the iGPU for display while the discrete GPU is out of the system.

 

[CMOS89]

Welcome to the forums, newcomer!

Did you try powering the system without the discrete GPU? You have a K suffix processor which would allow you to power off the iGPU for display while the discrete GPU is out of the system.

Hi Thank you for the warm welcome! I have not tried this but will give it a shot this evening and report back. Also, I made a mistake in my description of my build CPU is in fact an Intel i5-11500. This build used to have an 11600k but since the MOBA is not capable of overclocking I gave it to a friend to use instead. The 11500 has an iGPU still though so I should be able to test this.

 

[zx128k]

GFCI outlet was tripped. What would that imply? GFCI outlets exist to protect people from electrical shock.

If an appliance is working properly, all electricity that the appliance uses will flow from hot to neutral. A GFCI monitors the amount of current flowing from hot to neutral. If there is any imbalance, it trips the circuit.

There can be this sort of problem with a short to ground. If some of the current goes to ground there is an imbalance between hot and neutral.

I would try another PSU and another outlet. PSU failures can damage components.

 

[CMOS89]

GFCI outlet was tripped. What would that imply? GFCI outlets exist to protect people from electrical shock.

If an appliance is working properly, all electricity that the appliance uses will flow from hot to neutral. A GFCI monitors the amount of current flowing from hot to neutral. If there is any imbalance, it trips the circuit.

There can be this sort of problem with a short to ground. If some of the current goes to ground there is an imbalance between hot and neutral.

I would try another PSU and another outlet. PSU failures can damage components.

It implies to me exactly what you said, the outlet, assuming it is working correctly, detected a difference between the hot and neutral and tripped. If the PSU is the issue could it have damaged the GFCI?

The reason I ask is because while the board was out for replacement I put a different dinosaur PC in its place on the same outlet. After about a week it also tripped, no damage to that build as it worked fine after but is it possible the outlet is the issue here too?

 

[ex_bubblehead]

There are 2 places the fault may lie:

1: Ungrounded GFCI outlet/s
2: Leakage between Hot and Ground in the power supply (which will trip a GFCI).

GFCI breakers are not standard breakers. They trip at the slightest leak between hot and ground. They are there to protect you from dangerous voltage/current and NOT any equipment that may be plugged in. It is not recommended at all to be plugging equipment expecting a ground to be connected into an ungrounded GFCI socket.

 

[CMOS89]

There are 2 places the fault may lie:

1: Ungrounded GFCI outlet/s
2: Leakage between Hot and Ground in the power supply (which will trip a GFCI).

GFCI breakers are not standard breakers. They trip at the slightest leak between hot and ground. They are there to protect you from dangerous voltage/current and NOT any equipment that may be plugged in. It is not recommended at all to be plugging equipment expecting a ground to be connected into an ungrounded GFCI socket.

Hi,

Yes, in hindsight I should have had the outlet grounded before putting the PC in. My only question is what would you recommend I try next?

Everything since the MOBA RMA has been tried on a different outlet in my house, same PSU however. The PSU was tested when the system went down originally(pre-RMA), I did the fan test and I booted another system with it just to check and it worked, although I only had that system on for less than 5 minutes. Would leakage in the PSU only show up over extended PC uptimes?

 

[CMOS89]

Update:

I was able to spend sometime over the weekend trying several of the above suggestions and unfortunately none of these worked..I am still receiving a CPU error and no post. Below is what I have tried.

1. Attempted booting without the GPU since the CPU has a iGPU, still no post.
2. Removed the motherboard and checked for any shorts - The case I am using is a SSUPD Meshlicious. I did not notice the MOBA shorting out to anything.
3. Tried different RAM, Tried a different brand of DDR4 RAM that I had laying around.
4. Tried a different PSU, have a Corsair RM 550x that is compatible with the SF series cables.

All of the above attempts at booting were done on a different outlet that is grounded and still no post. Is it time to RMA the Motherboard again?

 

[zx128k]

The fault could have started a intermittent and tripped the GFCI outlet. Now you could have a fault anywhere within the system. A faulty SSD for example will stop posting. Its the fact the system doesn't work as well which component is causing the issue? CPU/motherboard/RAM etc.

 

[CMOS89]

The fault could have started a intermittent and tripped the GFCI outlet. Now you could have a fault anywhere within the system. A faulty SSD for example will stop posting. Its the fact the system doesn't work as well which component is causing the issue? CPU/motherboard/RAM etc.

So far I have tried and ruled out so many things, I even tried a different CPU and still no joy. Tried another PSU and nothing. Different Ram and the same result. CPU light immediately after pressing the power button and no post. I will give it a shot removing the M.2 drive and see if that makes a difference. Otherwise I believe I have ruled out pretty much all the other components and still the motherboard will not post, cant even get into the BIOS.

I am prepared to RMA the board again but prior to doing so I found a guide on how to do some basic checks on the motherboard with a multimeter, I figured it can't hurt to check it myself some before potentially sending it off for repair.

https://electrouniversity.com/how-to-test-a-motherboard-with-a-multimeter/

 

[zx128k]

Could be anything. I would try CMOS reset and BIOS flash. Like you state the motherboard could be a brick. The issue is if the motherboard is gone, why has this happened twice? Would be good to know, could be just blind bad luck but there could be a good reason as well.

 

[CMOS89]

Could be anything. I would try CMOS reset and BIOS flash. Like you state the motherboard could be a brick. The issue is if the motherboard is gone, why has this happened twice? Would be good to know, could be just blind bad luck but there could be a good reason as well.

I have only sent the board off for RMA one time, Prior to the RMA the DRAM error light was showing up on the Motherboard upon hitting the power button. Since receiving this error I have been unable to get the system to POST, not even once. Even after receiving the board back from RMA I have not been able to get a successful post and every attempt to power on the PC and get the system to post has failed, CPU error light is now showing in all attempts since getting the board back. I will try your suggestions regarding the SSD, CMOS, and BIOS flash.

Fingers crossed that these work, if they don't is it even worth paying to send it off for RMA again? I am afraid they will just send it back like the first RMA and not give me any information as to what was wrong and what they fixed other than "repaired". This is what they said after the first RMA and yet the system still wont post.

 

[zx128k]

Most people would swap components to find the fault. Try every other component in another system. If you tried a second cpu and the rest of the components work in another system. PSU works with another system. Then that would leave the motherboard. If you can't get the BIOS screen with just one DIMM of RAM at jedec thats not a good sign.

You state different RAM and CPU. Different PSU and removed all drives. Kind of narrows things down. Did you use liquid metal for thermal paste? Did you try removing the motherboard from the case to test? Its heading toward a RMA for the motherboard. I would most likely have RMA'ed by now.