CPU fan failed, runs in text mode, but restarts once in graphics mode

[tparkhill]

My PC is a stock HP Pavilion 500-164 (AMD A8-6500 APU, 8GB DDR3 RAM, 2TB HDD), running Win8.1. It has no add-in cards, and nothing connected except keyboard, mouse & monitor.

CPU fan failed, but I didn't realize it for a while...

I replaced the fan and thermal paste and it now boots w/o "fan failed" error, but only runs if using a text-based OS (BIOS diags, Linux command line, DOS). Once I boot Windows or Linux graphical interface the system immediately restarts.

BIOS-based memory test and a video-RAM test both ran without errors. I replaced PSU with a known-good PSU, too, but still get same reboot behavior.

Is the CPU bad or is it the motherboard? I suspect it's the CPU, but I'd appreciate your opinions before I drop $50-75 on a replacement.

-----

Additional info:

When Windows tries to load it fails with either of two blue-screen (BSOD) error messages:

WHEA_UNCORRECTABLE_ERROR, or MACHINE_CHECK_EXCEPTION

 

[clutchc]

If the fan on the CPU cooler stopped working, it would seem to indicate that the fan motor died. Get a new cooler or fan. Example: http://www.ebay.com/itm/NEW-Genuine-AMD-A8-7600-FM2-Heat-Sink-Fan-RED-EDITION-/332244644560?hash=item4d5b52f2d0:g:R7QAAOSw8lpZLaKe

 

[tparkhill]

Thanks, clutchc, but I've already replaced the fan.

It's either CPU or mobo (or both) that got cooked. I'm just trying to determine which is the more likely culprit. Unfortunately, I don't have another system w an FM2 CPU/mobo to swap-test.

 

[clutchc]

Thanks, clutchc, but I've already replaced the fan.

It's either CPU or mobo (or both) that got cooked. I'm just trying to determine which is the more likely culprit. Unfortunately, I don't have another system w an FM2 CPU/mobo to swap-test.

Then you may have to find a PC repair shop and take one or both in for testing. Or the whole machine. They can swap out components for testing and may have other testing equipment on hand. But finding a reputable repair shop isn't always easy.

Both of those errors you receive can indicate a hardware problem, as you already suspect. On modern processors, there are safeguards that throttle speed when the temp gets too high, eventually shutting down if that doesn't bring temps back in line. Motherboards may not have such safeguards, but usually aren't affected by CPU over temp unless something physically burns. Do you see any scorch marks anywhere? Any caps around the CPU socket that are extremely swollen or split?

If you are relying on the APU's iGPU for graphics, then the fact that it can't handle graphics video stress would make me lean towards the CPU (APU) being the culprit. Unfortunately, one damaged component can sometimes damage another.

 

[tparkhill]

Thanks, again, Clutchc. No need for a repair shop, I'm good with doing the work myself. I replaced the CPU and that improved things, but I had to add a video card before the system would run 100% stable. So, it looke like the motherboard needs to go, too. Sigh... Thanks for your help!

 

[clutchc]

Before you replace the motherboard, why not make sure the "instability" isn't related to another source. Might be software conflicts, malware, bloatware, etc.
If the machine is a pre-built, the first thing I would do is uninstall any crap that came pre-loaded that you don't need or want. Or atleast disable them from starting at boot. Run them on an as-need basis only. Better yet, get rid of them.
Next (even if the machine is one you built yourself), I would run CCleaner and do the Clean and Registry portions both. CCleaner has served me well for many years. It cleans up a multitude of conflicts and gets rid of lots of crap hogging resources. And it is 100% safe.
https://www.piriform.com/ccleaner

 

[tparkhill]

Thanks, but it cannot be software. The problems occur even when running Linux, which is loaded from a USB drive while the internal HDD is physically disconnected.