![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
I'm trying to debug a machine I built a year and a half ago and worked fine until very recently, and I'm running into some really strange problems.
First, my machine:
CPU: i5-2500k
Cooler: Scythe Mugen 2
Mobo: Asus P8P67m PRO B0 (currently on BIOS version 3602)
PSU: Antec NeoPower 650
Case: Antec Sonata Proto
RAM: 2x4GB G.Skill
Intel SSD (120GB)
Samsung HD (2TB)
My machine started to reboot periodically -- like once every 1-5 days. Generally while under no real load besides backing up online. I read through the event viewer logs, and I saw that one of my drives was timing out, so I thought this might be a BIOS issue. I flashed to BIOS version 3602 twice, got the system up and running again, and still noticed occasional reboots. I then reset all of my ASUS-derived OCs to run completely stock. That seemed to help, and I stopped seeing some minor graphics corruption I'd see occasionally, but then the crashes started the infrequent crashing started happening again.
Since I have a B0 P67 board, I was wondering if I had bad SATA2 connectors, so I moved my drives from the onboard SATA2 to SATA3 and my CDROM from SATA2 to the on-board Marvell controller SATA3.
This is when things got really crazy. I reboot, and I see errors for "CPU over temperature error!" and "CPU over voltage error!" I went into my BIOS, and the readings were totally bogus. 512F chip, 24V power rails. If I powered down and powered back up, the system wouldn't boot at all (no fans, no beeps, no lights) until I cleared CMOS. Once again I saw the temperature / voltage problems, I cleared BIOS, and I got into Windows. I saw that the drive problems were still happening so I upgraded my Intel Rapid Storage Technology driver. Rebooted, and the events went away.
Then while watching TV, I watched my machine reboot itself after ~13-15 min of idle time, about 4 times. I then played some Diablo 3, and it crashed in the middle. I tried to boot up in the morning today, and the fans wouldn't even so much as spin until I cleared CMOS. When I booted back up, I saw the same voltage errors. And when I saved BIOS and rebooted, I once again no longer POSTed until I reset CMOS. I've tried this about 5 times today, and each time I have to reset CMOS.
So right now, I unplugged all my SATA data and power connections, rebooted into the BIOS, and saw no cpu over temperature or over voltage errors. However, when I went into the BIOS my readings are totally wacky:
- CPU temperature fluctuates between N/A, 12C, and 106C
- Motherboard is either a reasonable 26C or N/A
- CPU voltage is 1.26V or 0.002V
- 3.3V is either 2V, 3.2V
- 5V is 5V or 0V
- 12V is 12V or 24V
- CPU_FAN, PWR_FAN, CHA_FAN1, and CHA_FAN2 are all jumping all over the place between actual values and "N/A"
One last thing -- when I unplug/plug my power cable into my PSU, I hear very faint snapping/cracking sounds that sound like there might be some arcing. I don't know if this has been happening for a while or if I just noticed it now.
In summary, I went from infrequent crashes to a dead machine. What do you think is wrong with my system? I'm guessing it's either the PSU, the mobo, or the CPU -- but I don't know which and I don't have spares for any but the PSU to switch out. Because of my huge ass HSF, if I want to replace the PSU I'll need to take out my entire motherboard....
Thanks for any help!
First, my machine:
CPU: i5-2500k
Cooler: Scythe Mugen 2
Mobo: Asus P8P67m PRO B0 (currently on BIOS version 3602)
PSU: Antec NeoPower 650
Case: Antec Sonata Proto
RAM: 2x4GB G.Skill
Intel SSD (120GB)
Samsung HD (2TB)
My machine started to reboot periodically -- like once every 1-5 days. Generally while under no real load besides backing up online. I read through the event viewer logs, and I saw that one of my drives was timing out, so I thought this might be a BIOS issue. I flashed to BIOS version 3602 twice, got the system up and running again, and still noticed occasional reboots. I then reset all of my ASUS-derived OCs to run completely stock. That seemed to help, and I stopped seeing some minor graphics corruption I'd see occasionally, but then the crashes started the infrequent crashing started happening again.
Since I have a B0 P67 board, I was wondering if I had bad SATA2 connectors, so I moved my drives from the onboard SATA2 to SATA3 and my CDROM from SATA2 to the on-board Marvell controller SATA3.
This is when things got really crazy. I reboot, and I see errors for "CPU over temperature error!" and "CPU over voltage error!" I went into my BIOS, and the readings were totally bogus. 512F chip, 24V power rails. If I powered down and powered back up, the system wouldn't boot at all (no fans, no beeps, no lights) until I cleared CMOS. Once again I saw the temperature / voltage problems, I cleared BIOS, and I got into Windows. I saw that the drive problems were still happening so I upgraded my Intel Rapid Storage Technology driver. Rebooted, and the events went away.
Then while watching TV, I watched my machine reboot itself after ~13-15 min of idle time, about 4 times. I then played some Diablo 3, and it crashed in the middle. I tried to boot up in the morning today, and the fans wouldn't even so much as spin until I cleared CMOS. When I booted back up, I saw the same voltage errors. And when I saved BIOS and rebooted, I once again no longer POSTed until I reset CMOS. I've tried this about 5 times today, and each time I have to reset CMOS.
So right now, I unplugged all my SATA data and power connections, rebooted into the BIOS, and saw no cpu over temperature or over voltage errors. However, when I went into the BIOS my readings are totally wacky:
- CPU temperature fluctuates between N/A, 12C, and 106C
- Motherboard is either a reasonable 26C or N/A
- CPU voltage is 1.26V or 0.002V
- 3.3V is either 2V, 3.2V
- 5V is 5V or 0V
- 12V is 12V or 24V
- CPU_FAN, PWR_FAN, CHA_FAN1, and CHA_FAN2 are all jumping all over the place between actual values and "N/A"
One last thing -- when I unplug/plug my power cable into my PSU, I hear very faint snapping/cracking sounds that sound like there might be some arcing. I don't know if this has been happening for a while or if I just noticed it now.
In summary, I went from infrequent crashes to a dead machine. What do you think is wrong with my system? I'm guessing it's either the PSU, the mobo, or the CPU -- but I don't know which and I don't have spares for any but the PSU to switch out. Because of my huge ass HSF, if I want to replace the PSU I'll need to take out my entire motherboard....
Thanks for any help!
Last edited:
