System freezing

[eddietandy]

I have been running a 2600k just fine at 4.2 Ghz, until today when I added a new fan for push/pull. I put in a Gelid FN-FW12BPL-18 as the push fan (CPU_FAN), and moved the stock CM Hyper 212+ fan to the pull side (CHA_FAN1). Both are PWM fans.

When I booted, I got a CPU Fan error, and it showed the Gelid at a really low RPM, ~300. So I go back into the BIOS and it shows it around 800, all good. Reboot and no error, Win 7 runs fine, so I go to check CPU temps at full load. Sure enough it's showing about 4 degrees cooler, but the machine froze and rebooted itself after about 20 minutes. Disabled Q-Fan and ran push/pull at max RPM, then went back into Win 7 to stress test. As soon as the CPU hit 100%, the board reset. Now I'm back in at idle speeds with Q-Fan on, no problem so far.

I'm baffled at what could be wrong. Are the fans maybe pulling too much voltage? Also wondering if the Gelid dipping below 600 RPM is causing an error that freezes the system. I set the CPU fan speed low limit to 400 rpm to try that. Just thinking at this point something is up between the motherboard and the PWM functionality with the Gelid fan.

Any ideas?

 

[Seven]

Disable C3/C6 and report back to us.

 

[eddietandy]

Did this and some IntelBurnTest runs did just fine. Went all the way to the maximum test and no problems. I'll do some folding for a while to see if it locks up.

Of course, I'd certainly like to know the background info on how a new CPU fan and the C3/C6 reporting are linked. The only thing different was the new Gelid PWM fan on the CPU_FAN header, and the existing CM stock PWM fan moved to the CHA_FAN1 header.

BTW thanks for helping :thumbsup:

 

[eddietandy]

Well I thought I was out of the woods, but after more than an hour folding at full throttle, it froze up again. Literally have not had this kind of issue at all since updating to 1253 weeks ago. The only thing different is I now have two PWM fans plugged in. I'll obviously disconnect it and go back to what I had, but I just don't get it...

 

[eddietandy]

I found the option for hybrid sleep in Windows 7 was enabled, so I turned that off. No idea if that's related...

 

[LiuKangBakinPie]

sounds like you got a fan plugged into the wrong place. what did you connect the fans to?

 

[eddietandy]

CPU fan on push side is plugged into CPU_FAN. CPU fan on pull side is plugged into CHA_FAN1. Both are 4 pin PWM fans, and both headers are 4 pin. All other fans are connected to Molex straight from the PSU.

 

[eddietandy]

The box has been running at idle all day without problems, so to this point it has only frozen when running at full load since adding the new fan. And thinking back to how it instantly rebooted when applying full load with fans at max, it makes me think it's a voltage setting in the the BIOS. Read on another forum about disabling 'internal PLL voltage' and setting 'VRM Frequency= 350', and maybe 'voltage offset = .020'

Anyone else using those settings?

 

[bankster55]

yeah dis PLL volt
vrm 350 is only for extreme overclockers 4.8+
look, why dont you flip up the TPU switch on mobo, reboot/shutdown (not restart) a coupla times into windows and you should be at a safe 4.4 24/7
then disable anything to do with sleep\hibernate on
>cntrl panel, power, balanced, advanced - undo ALL sleep crap in there - like USB
You have it all in the mobo already
>dev manager network device,realtek network dev/intel network dev - power management
> soundcard, modem\router - everything is friggin powersaving today

 

[eddietandy]

yeah dis PLL volt
vrm 350 is only for extreme overclockers 4.8+
look, why dont you flip up the TPU switch on mobo, reboot/shutdown (not restart) a coupla times into windows and you should be at a safe 4.4 24/7
then disable anything to do with sleep\hibernate on

Just hit the easy button and don't worry about it, eh?

I've been stable at 4.2 for a while, and don't really want to run at 4.4. Plus I tried the autotune when I first built it, and didn't care for the results.

So, I have had sleep features disabled except the hybrid sleep which I already mentioned. I've disabled PLL overvoltage now and after some good runs of IBT, I'm folding at full throttle to see if it holds. 30 minutes and counting...

 

[eddietandy]

Well, disabling PLL overvoltage certainly seems to have made a difference. Been folding strong for 7.5 hrs. I'm really curious about the technical relationship (if any) between this setting and the number of PWM fans running.