- Oct 5, 2002
- 456
- 12
- 18,785
I have been having a mountain of issues with my system. It started when I tried to replace a drive and it would no longer turn on.
After stripping it down to essentials, it would turn on but would display no video, so I plugged my GPU back in and I would get video.
However, after this I was getting all sorts of random problems. The motherboard would not post about 10-20% of the time, I would sometimes get all sorts of odd POST error messages that made no sense (such as claiming the CPU had been changed when it had not been touched in years, or case intrusion detection went off when neither the motherboard nor case has that feature).
It also almost never manages to post if I have any SATA drives connected, even if it's just an optical drive.
I can get it to boot from USB however, so I used that first to run memtest86+, no problems after 12+ hours and three cycles. Then I booted the usb drive into Linux and ran 36 hours of Prime95, no issues, and then added two hours of Furmark on top of that, no issues. So I figured if memtest86+ showed no errors, and the stress of Ptime95 along with furmark gave no errors, this means the CPU, GPU, RAM, and PSU must be ok and likely the motherboard has gone bad, especially since it has been on for almost a week now running off that Linux USB. Once it POSTs it seems to run fine, but it's getting it to complete POST in the first place that is the problem.
As I was taking it apart and disconnecting everything to get it ready to rebuild with a new motherboard however, since everything was now disconnected I figured it wouldn't hurt to plug it into a new PSU tester I got during this mess, and to my surprise it fails:
View: https://i.imgur.com/B2V9aqx.jpg
The thing is, all the voltage rails appear to be fine according to that readout. Yet, the "Power Good" signal is what is failing, which doesn't really make sense if all the voltages check out. I am confused now if the PSU could be the culprit, the motherboard, or both? The PSU is also still under warranty, would this be enough to count for a warranty replacement?
After stripping it down to essentials, it would turn on but would display no video, so I plugged my GPU back in and I would get video.
However, after this I was getting all sorts of random problems. The motherboard would not post about 10-20% of the time, I would sometimes get all sorts of odd POST error messages that made no sense (such as claiming the CPU had been changed when it had not been touched in years, or case intrusion detection went off when neither the motherboard nor case has that feature).
It also almost never manages to post if I have any SATA drives connected, even if it's just an optical drive.
I can get it to boot from USB however, so I used that first to run memtest86+, no problems after 12+ hours and three cycles. Then I booted the usb drive into Linux and ran 36 hours of Prime95, no issues, and then added two hours of Furmark on top of that, no issues. So I figured if memtest86+ showed no errors, and the stress of Ptime95 along with furmark gave no errors, this means the CPU, GPU, RAM, and PSU must be ok and likely the motherboard has gone bad, especially since it has been on for almost a week now running off that Linux USB. Once it POSTs it seems to run fine, but it's getting it to complete POST in the first place that is the problem.
As I was taking it apart and disconnecting everything to get it ready to rebuild with a new motherboard however, since everything was now disconnected I figured it wouldn't hurt to plug it into a new PSU tester I got during this mess, and to my surprise it fails:
View: https://i.imgur.com/B2V9aqx.jpg
The thing is, all the voltage rails appear to be fine according to that readout. Yet, the "Power Good" signal is what is failing, which doesn't really make sense if all the voltages check out. I am confused now if the PSU could be the culprit, the motherboard, or both? The PSU is also still under warranty, would this be enough to count for a warranty replacement?
Facebook
Twitter
Instagram