Hi everyone!
I've got a Corsair HX750 Gold PSU, which is the only remaining component from my first ever PC build. It has been running flawlessly since I purchased it in spring 2013.
The problem started a few days ago. I got my first M.2 NVMe SSD, and started using it as my boot drive.
I installed the drive on monday, and just a few days later, on friday, I had the first crash. The screens went dark, I heard a bang over the headphones, and the computer rebooted only to fail to POST. The motherboard LEDs and buzzer beeps indicated a "VGA not detected" failure. My GPU is only a couple months old and never gave me any issues, but I trusted the motherboard's instinct and took the GPU out.
I then turned on the PC with the monitor conected to the motherboard's IO, only to be presented with the same "VGA not detected" code. This really confused me, as I had already removed the GPU. I then decided to remove the newly added NVMe SSD, and the computer made it to the BIOS fine this time.
I then added the NVMe back, but Windows wouldn't boot, so I was forced to reinstall Windows 10. I then added seated the GPU back in its slot, and everything seemed to be working as good as new for a while.
Fast forward to sunday night: The same crash scenario happened. The monitors went black at the same time, loud bang over the headphones, and I kept hearing the audio from the YouTube video I was watching, until it cut off a few seconds later. The computer was unresponsive that entire time, since not even the volume knobs on the keyboard would work.
The computer restarted on its own and this time booted up just fine.
I paid a visit to the BIOS config, to see if I had missed anything important during the NVMe setup. It is there where I noticed that there was something odd with the voltages being shown:
+12V was alright, within spec.
+5V was okay too, nothing too weird about it.
The +3.3V reading was displayed on RED on the BIOS, I remember seeing values as low as 2.95 Volts.
I've always been told to take software voltage measurements with a grain of salt, but these being the main rail voltages off the BIOS, they seemed legit.
I then went back to Windows and kept using the computer as usual.
The next day, as I was browing the internet, the computer crashed in the same exact fashion again.
When it rebooted, I did a quick test to figure out what was going on: I ran OCCT on max PSU stress mode, and ran a benchmark on the SSD at the same time. I saw the 3.3V rail reading drop to 3.08 Volts, and after a dozen seconds of stress, it crashed in the same manner.
When the computer rebooted, I ran the same test again. This time, the system crashed as soon as I clicked "start" on the OCCT stress test.
After doing some investigation, I've learnt that while SATA SSDs run off +5 volts, NVMe drives use the +3.3V rail. Could it be undervolting the drive and thus causing the system to crash?
I am starting to wonder whether my PSU had this fault all along, but I never noticed it until I actually put a serious load on the +3.3 rail.
The idle voltage on the 3.3V rail usually stays between 3.1 and 3.2 Volts.
Computer specs:
CPU: i7 6700k
Motherboard: ASUS Strix Z270H Gaming
GPU: Sapphire 5600XT
PSU: Corsair HX750 Gold (2013)
SSD: Samsung 970 EVO Plus NVMe 500GB
PD: Windows Event Viewer shows nothing, which is what initially led me to believe that this is most probably a hardware problem.
Hi everyone! I've got a Corsair HX750 Gold PSU, which is the only remaining component from my first ever PC build. It has been running flawlessly since I purchased it in spring 2013. The problem started a few days ago. I got my first M.2 NVMe SSD, and started using it as my boot drive. I installed the drive on monday, and just a few days later, on friday, I had the first crash. The screens went dark, I heard a bang over the headphones, and the computer rebooted only to fail to POST. The motherboard LEDs and buzzer beeps indicated a "VGA not detected" failure. My GPU is only a couple months old and never gave me any issues, but I trusted the motherboard's instinct and took the GPU out. I then turned on the PC with the monitor conected to the motherboard's IO, only to be presented with the same "VGA not detected" code. This really confused me, as I had already removed the GPU. I then decided to remove the newly added NVMe SSD, and the computer made it to the BIOS fine this time. I then added the NVMe back, but Windows wouldn't boot, so I was forced to reinstall Windows 10. I then added seated the GPU back in its slot, and everything seemed to be working as good as new for a while. Fast forward to sunday night: The same crash scenario happened. The monitors went black at the same time, loud bang over the headphones, and I kept hearing the audio from the YouTube video I was watching, until it cut off a few seconds later. The computer was unresponsive that entire time, since not even the volume knobs on the keyboard would work. The computer restarted on its own and this time booted up just fine. I paid a visit to the BIOS config, to see if I had missed anything important during the NVMe setup. It is there where I noticed that there was something odd with the voltages being shown: +12V was alright, within spec. +5V was okay too, nothing too weird about it. The +3.3V reading was displayed on RED on the BIOS, I remember seeing values as low as 2.95 Volts. I've always been told to take software voltage measurements with a grain of salt, but these being the main rail voltages off the BIOS, they seemed legit. I then went back to Windows and kept using the computer as usual. The next day, as I was browing the internet, the computer crashed in the same exact fashion again. When it rebooted, I did a quick test to figure out what was going on: I ran OCCT on max PSU stress mode, and ran a benchmark on the SSD at the same time. I saw the 3.3V rail reading drop to 3.08 Volts, and after a dozen seconds of stress, it crashed in the same manner. When the computer rebooted, I ran the same test again. This time, the system crashed as soon as I clicked "start" on the OCCT stress test. After doing some investigation, I've learnt that while SATA SSDs run off +5 volts, NVMe drives use the +3.3V rail. Could it be undervolting the drive and thus causing the system to crash? I am starting to wonder whether my PSU had this fault all along, but I never noticed it until I actually put a serious load on the +3.3 rail. The idle voltage on the 3.3V rail usually stays between 3.1 and 3.2 Volts. Computer specs: CPU: i7 6700k Motherboard: ASUS Strix Z270H Gaming GPU: Sapphire 5600XT PSU: Corsair HX750 Gold (2013) SSD: Samsung 970 EVO Plus NVMe 500GB PD: Windows Event Viewer shows nothing, which is what initially led me to believe that this is most probably a hardware problem.