PC constant freezes after CPU fan swap

[rh71]

WIN10 64bit
Q6600 CPU with some standard Rosewill fan
Gigabyte EP45 mobo
4GB G.Skill RAM
Radeon 6800
Corsair 550W PSU

I put the PC together from old parts and it had been running for my boys for months, but the CPU fan kept shifting because the plastic feet inserted into the mobo were partially broken. HWMonitor showed temps of all 4 cores pegging 98C (200F) constantly the last week - I know that's bad and the only symptoms at the time were slow-responding apps / loading. Never froze, but at times HWMonitor would only show 2 of the 4 cores (not sure why). So I finally ordered a CoolerMaster Hyper T4 that installed without a hitch along with some Arctic Silver which I've also used many times in the past (I did remove one RAM module so I had room to do it then re-seated the RAM back in).

Upon boot-up, HWMonitor showed all 4 cores at a cool mid 40s celsius. Perfect. Then minutes later the screen froze and the PC emitted some squeal through the monitor speaker. No response from the mouse or keyboard. Many reboots resulted in the same thing, only minutes of usage before freezing. Even happens in safe mode. The light on the mouse turns off so I know that's the point it happens and it's not a display issue. I read CPUs either work or they don't so I know the CPU was not detrimentally effected from the heat.

I've since re-installed WIN10 completely (that ran for almost an hour without freezing) but when finally loaded to the desktop, it's still happening on a fresh boot with nearly nothing else connected. Tried a different mouse, keyboard, and even without the wifi card in so it's none of those. I've also reseated both 2GB RAM modules. Would you suspect the RAM first? PSU issue?

 

[DAPUNISHER]

Try each stick one at a time and in more than one slot, if you have not already. Clear CMOS, and swap psu since that is still a suspect. But, my first guess, because of the timing, is that the mobo may have been damaged.

 

[LTC8K6]

The squeal through the speaker is likely an alarm. Find out what alarm. Probably related to heat or fan speed.

 

[VirtualLarry]

HWMonitor showed temps of all 4 cores pegging 98C (200F) constantly the last week - I know that's bad

so I know the CPU was not detrimentally effected from the heat.

Hmm...?

 

[VirtualLarry]

You're sure that you plugged in the new CPU fan's power/RPM fan connector onto the board? Did you reset or re-calibrate the "CPU Smart Fan", or set it to "Disabled" to run full speed?

Sure that you're got the new heatsink properly attached? I mean, you admit you used the old heatsink, without any sort of proper attachment. It doesn't sound like you're that careful, or care, TBH. You just want this cheap hunk of junk parts to work. It may be time to put some or all of them out to pasture, or at least, for the gold recyclers.

 

[rh71]

Hmm...?

It either works or it doesn't, right? So since it's booting, how else can the CPU be bad? I was worried it would fry for good.

You're sure that you plugged in the new CPU fan's power/RPM fan connector onto the board? Did you reset or re-calibrate the "CPU Smart Fan", or set it to "Disabled" to run full speed?
Sure that you're got the new heatsink properly attached? I mean, you admit you used the old heatsink, without any sort of proper attachment. It doesn't sound like you're that careful, or care, TBH. You just want this cheap hunk of junk parts to work. It may be time to put some or all of them out to pasture, or at least, for the gold recyclers.

New CPU fan is spinning, I have the cover off. The old HSF was clipped fine at the top 2, but not the bottom 2 and as such, it felt flush enough for me not to go into emergency OMG mode for a kids' PC that they barely used. The new HSF has a pressure-clip and it's surely on and ain't movin'. Temps are good at 40Cs so that isn't the problem.

Update.. so after running windows' memory test (mdsched) that came back fine, the PC has stayed on long enough for me to download/install the latest windows updates (still going). Though twice instead of freezing it has rebooted itself unexpectedly (setting turned off in system properties). Temps remain fine. I don't expect it to stay compliant long because well, I haven't done anything different.

 

[VirtualLarry]

Swapped the PSU yet? That would be my next thing to try.

BTW, during the Windows Updates, and the freezing or rebooting, did the BIOS alarm sound at all? That's generally either a fan alarm (not spinning, or not spinning fast enough, which can be caused by mis-calibration of Smart Fan settings in BIOS, as well as a bad/flaky fan, or most likely, forgotten to be connected, or a temp thing, but you say that temps are good now with the new cooler.

Edit: The alarm could also be the dGPU, not getting enough power. Some of them have a little piezo beeper on them, to signal error conditions. If the PSU was failing or worn out, it could cause freezes and hangs and reboots of the CPU, as well as dGPU errors.

My next course of action, would be to swap in a known-good PSU, followed by a lower-powered dGPU.

 

[rh71]

UPDATE:

The random freeze is still happening. Things I've replaced since the start of this thread:
WIN10 to WIN7
new PSU
new RAM

So all that's left is the GPU, mobo, or the Q6600 (which I thought either worked or won't boot at all). It seems to be heat-related because it only happens after it's been on a while (watching Youtube or gaming). And if we immediately power it back on, it freezes again very shortly. Thing is, HW Monitor doesn't show any heat issues and all components are a pretty constant 40-50C.

 

[VirtualLarry]

I\d like to help, but I'm really not sure what's the problem here, other than, maybe the board or GPU have caps that are getting "old" (leaking), or VRM components that are getting "worn" (overheated and degraded from usage).

At this point, I might try a new Q6600 CPU. (Should be ample used supply on ebay, but getting one that wasn't overclocked may be a big crap-shoot.)

If you replace the CPU, board, AND GPU, then you've effectively replaced the entire rig. Which might not necessarily be a wise choice, financially, unless the parts cost you nothing.

*somewhere*, I have a Q8200, and a G31/33-ish micro-ATX S775 board.

I also have a "collector's edition" Q6600 BNIB from MC, and a DFI X48 S775 board, again, BNIB. (DDR2). I don't really want to get rid of them, though, they might go into a technology museum some day.

Edit: Have you checked the caps? Maybe with a magnifying glass?

Might not hurt, to try to (gently, carefully) remove the northbridge, and southbridge heatsinks, and re-paste them with something like MX-4. (Not AS5.)

NB/SB overheating can cause crashes too, on older / used rigs.

Edit: I've got a Ryzen 2200G, 8GB DDR4-3000, 512GB M.2 NVMe SSD, and AM4 B350 micro-ATX board (open box, other stuff is NIB), in my FS/FT thread. If you're interested in upgrading at this point, it would help me out right now if someone bought some of my stuff.