Troubleshooting freeze-ups on Acer Veriton X4610G How can I pinpoint the random freezes?

@Larryodie Thank you very much for your posts and effort.

I'll try uncoupling as many USB devices as possible, but mouse and keyboard are essential. The only one I could decouple is the HP LaserJet printer.

I investigated SPECCY, but it seems that it can't do automatic logging. Unfortunately I can not predict when the system will freeze, so without automatic logging of voltages I will always miss to capture voltages at (or right before) freeze moment.

The PC is not under heavy load when freezing and temperatures remain nice and low.

Thanks for your effort.

@Larryodie

I intend to try SPECCY as soon as I get the time and will lookout for voltage fluctuations.

Because identical freeze-ups also occur when I've booted Linux, I am sure that it is not a Windows malware / virus issue, except when it has infected UEFI/BIOS, Intel Management Extensions or the Printer Firmware. I don't have any indication of that yet. Nonetheless I have run a deep boot-up scan, without any findings.

I had flicked on the BIOS non writable jumper to prevent unwanted BIOS changes. I have de-activated Intel ME as far as possible and don't use the onboard Intel network interface (that can be accessed autonomously by Intel ME). I had a printer firmware lock configuration in place, but that was in an unlocked state... I can't remember if I did that myself?

Freeze-ups occur totally at random. I can't predict them. I've got a hunch that they occur more often after a fresh start on a cold machine. I've seen them occur when browsing (Firefox) and scrolling down. I've seen them when Idle. I've also seen them after long running unattended tasks. I did manage a long run in Windows recovery mode, with a lower resolution and most drivers unloaded.

My hunch is that it could be graphics related? As onboard graphics shares/takes memory with/from main memory, would Passmark MemTest86 also have tested that memory or would it have skipped it?Does graphics always claim the same physical pysical memory?, then there could still be untested DIMM memory. I've never blown out a CPU yet, how likely would it be that the CPU/On board graphics would be failing? Otherwise components on the motherboard could be at fault. How could one test for that?

What could cause memory corruption, that is not detected by Passmark MemTest86 (that runs in a very simple low res DOS like environment)?

Thanks again for your reply.

I like your thinking. I'd also want to be sure before buying a new one.

I tried running HWiNFO. Unluckily the voltage sensor readings from the IT8728F chip appear to be largely messed up. Only the CMOS batt voltage shows correctly and I can see a +3.3 volt measure under the wrong label furthermore all other readings make no sense and can't be correlated with the BIOS measures. Only the processor voltage is correct. HWiNFO claims to be the only tool that reads this values, but needs additional information from Acer in order to make sense of the data. HWiNFO can do logging as well. What voltage measures are the best indicator for PSU problems?

Thanks for your reply. I can see a fan and it is blowing when on.

I shut it down multiple times a week, but sometimes it can be asleep for a day.

It has been on for a lot of hours when memory testing without freezing.

Strangely I has run without freezing for a few days now. The weather here has changed from below 0 C to above. Although in the house its always above 16 C.

I had similar problems for a while in august to. But last week it kept freezing up.

(I unplugged it from the wall socket also BTW.)

Mem testing ran in low res BTW. I'm not sure if it tested the memory that is in use by the onboard GPU?