12v rail drop, lockups, hard reboot, etc

[DeathAngelLST]

So AIDA64 reports a drop in the 12v rail to as low as 11.668v when running both AIDA's stability test and MSI Kombustor with anti-aliasing enabled.

I'm getting random lockups, followed by a reboot, while doing video encoding with avisynth, sometimes after a few minutes, sometimes after a couple hours, sometimes when it's almost finished. I also used to get these lockups when playing Battlefield Bad Company 2, forcing me to hard reboot.

Another weird thing is if I plug the pc off the wall for too long - hours -, once I plug it back in sometimes it won't start. I can push the power button dozens of times and nothing happens. Then I would wait for 15min~1h+ and it suddenly powered on. This has been happening for a couple years now, so I rarely plug it off.

Is it time for a new PSU?

 

[redduc900]

The voltage tolerance for the +12VDC rail is ± 5% (+11.400VDC to +12.600VDC).

 

[larrymoencurly]

ATX specs allow +- 5% deviation, or anywhere from 11.40V to 12.60V, so 11.668V is within specs, but you need to verify voltages with a digital voltage meter with at least 3 1/2 digits because the motherboard's monitoring hardware is sometimes way off. OTOH just because ATX specs allow the +12V to vary between 12.60V to 11.40V, you don't really want that big a variation, and it usually indicates something wrong with the PSU.

The PSU not starting after having been unplugged for hours could indicate a problem with the +5Vstandby circuit, such as bad capacitors (but I think reviews show that the HX-520 uses high quality Japanese caps) or a cracked solder joint that closes up and temporarily fixes itself from heat in the PSU.

 

[Adragontattoo]

So AIDA64 reports a drop in the 12v rail to as low as 11.668v when running both AIDA's stability test and MSI Kombustor with anti-aliasing enabled.

I'm getting random lockups, followed by a reboot, while doing video encoding with avisynth, sometimes after a few minutes, sometimes after a couple hours, sometimes when it's almost finished. I also used to get these lockups when playing Battlefield Bad Company 2, forcing me to hard reboot.

Another weird thing is if I plug the pc off the wall for too long - hours -, once I plug it back in sometimes it won't start. I can push the power button dozens of times and nothing happens. Then I would wait for 15min~1h+ and it suddenly powered on. This has been happening for a couple years now, so I rarely plug it off.

Is it time for a new PSU?

The voltage tolerance for the +12VDC rail is ± 5% (+11.400VDC to +12.600VDC).

Also, unless you are checking with a Multimeter, you are just guessing at what the voltage is.

If you want to troubleshoot whether it is a PSU issue, get a Multimeter, plug it in, turn it on and watch what it says your 12v is at under load. Unplug ONE GPU and retest.

The switch itself may be bad, unplug all case LEDs and switches and try an entirely different switch and lead.

What does Windows Event viewer have listed for the error? Have you disabled reboot after BSOD?

Have you tested your RAM?

What about stability testing?

Drop your OC to stock, clear CMOS and recheck these issues.

 

[freeagent]

I was having similar issues. I took a small screw driver and straightend all the pins in the main atx plug. I saw 11.35 as a low before it shut down now it stays over 12v

 

[DeathAngelLST]

If you want to troubleshoot whether it is a PSU issue, get a Multimeter, plug it in, turn it on and watch what it says your 12v is at under load. Unplug ONE GPU and retest.

Only have one video card.

The switch itself may be bad

How so? It powers up fine as long as I don't unplug the pc from the wall for a few hours.

What does Windows Event viewer have listed for the error?

Nothing.

Have you tested your RAM?

Ran it overnight for about 7 hours, memtest shows no errors.

What about stability testing?

12 hours of prime95 stable and AIDA's stability test shows no error messages either.

Drop your OC to stock, clear CMOS and recheck these issues.

Everything is stock, except the vcore which is too low (1.21~1.23v at default bios setting according to AIDA) so I increased it to 1.27ish volt at bios and it sits around 1.248~1.264v (again according to AIDA).

 

[Adragontattoo]

Take a look at the first page of http://www.overclockers.com/forums/showthread.php?t=515316 regarding monitoring and testing. I know you said you arent overclocking but the info still applies.

What about Temps? Under load what are your temps? Specifically CPU.

Also you might want to enable your Blue Screen of Death (as terrible as that sounds) so that you can debug the problematic lockups. This is done by:
--> right clicking on "Computer" from the start menu
--> Select "Properties"
--> Select "Advanced System Settings"
--> Select the "Advanced" Tab
--> Select the "Startup and Recovery"
--> Make sure that "Write an event to the system log" is enabled.

Do you have any cleaner software that runs in the background or at boot?

Something is preventing the system from logging an entry in Event viewer or is cleaning the event log at boot.

Have you checked the various drives for SMART errors and/or tested them?

I had a system that was having all kinds of weird issues (random hard locks, boot oddities, visual corruption, etc.) that I beat on for a few weeks steadily trying to identify and resolve. Different MB, OS HDD, OS, RAM, PSU, CASE, etc... Occasionally I would have an alert stating that one drive had bad blocks detected during boot but once in the OS, I wouldnt see anything from the GUI (running Mint64 13-14) when I checked the disks. At one point, I got lucky and identified the problem HDD by serial and removed it to test separately, and suddenly all the problems went away. For whatever reason, that disk had decided to try to take the entire 22 HDDs, and all other components with it. If I plug that drive in on any system, it immediately pegs at 100% speed, ramps CPU to 100% and IF you can avoid it crashing, requires a reboot to restore normal functionality after you remove it.

 

[DeathAngelLST]

What about Temps? Under load what are your temps? Specifically CPU.

60ºC while running Prime95 - it's not summer yet here.

Also you might want to enable your Blue Screen of Death (as terrible as that sounds) so that you can debug the problematic lockups. This is done by:
--> right clicking on "Computer" from the start menu
--> Select "Properties"
--> Select "Advanced System Settings"
--> Select the "Advanced" Tab
--> Select the "Startup and Recovery"
--> Make sure that "Write an event to the system log" is enabled.

It is enabled, and there are no logs.

Do you have any cleaner software that runs in the background or at boot?

No.

Have you checked the various drives for SMART errors and/or tested them?

All OK so far. I used HD Tune Pro to check for bad blocks on the WD drive, found nothing. Then used Seagate's SeaTools on the 1TB drive, all good. Gonna run it overnight on the 2TB drive, possibly gonna take 12 hours or so.

 

[usyed1]

So AIDA64 reports a drop in the 12v rail to as low as 11.668v when running both AIDA's stability test and MSI Kombustor with anti-aliasing enabled.

I'm getting random lockups, followed by a reboot, while doing video encoding with avisynth, sometimes after a few minutes, sometimes after a couple hours, sometimes when it's almost finished. I also used to get these lockups when playing Battlefield Bad Company 2, forcing me to hard reboot.

Another weird thing is if I plug the pc off the wall for too long - hours -, once I plug it back in sometimes it won't start. I can push the power button dozens of times and nothing happens. Then I would wait for 15min~1h+ and it suddenly powered on. This has been happening for a couple years now, so I rarely plug it off.

Is it time for a new PSU?

I had an antec high voltage gamer psu literally EXPLODE in a god damn explosion that rocked my casing shortly after almost identical findings you are seeing... that was back in 2012 though...