Friend's Win7 64-bit PC won't restart properly

[VirtualLarry]

Trying to figure this one out. Friend has a ASRock AM2+ mobo, 4x4GB of DDR2, an AM3 Athlon II X4 640 CPU, a GT610 GPU, a 120GB SSD, and Windows 7 64-bit SP1.

Recently, his GF said his PC was "freezing", and what he told me was, he sometimes goes to restart, and it just sits there at a black screen, with the fans running. The ASRock BIOS full-screen logo doesn't come on his screen (NO POST?), and he has to hit the RESET button, then it boots up.

I'm wondering, if it's power-supply (was replaced only 1-2 years ago, if that, with a new-model TR2-430 ThermalTake, with a 5-year warranty), motherboard (board does have solid caps where it matters, but it's 10 years old, this seems a likely suspect), or drives (his SSD is a few years old, and he's got a WD 500GB HDD in there as well). Or maybe CMOS battery is low?

I've ordered a new mobo to drop in (a Gigabyte 78LMT-USB3 R2), and have some DDR3 16GB RAM and a bigger SSD, so we can do an upgrade replacement on some items.

Still, other than wholesale replacement of a bunch of parts, it would be nice to figure out exactly what's failing.

Suggestions on where to start? CMOS battery? PSU? Motherboard?

From:
https://superuser.com/questions/199261/windows-7-wont-restart-properly
Enable "PME Wake-up event" in BIOS, or remove CMOS battery temporarily, are the suggestions given there.

From:
https://superuser.com/questions/530418/my-pc-doesnt-restart-properly
Suggests that there may be a system Service hanging, after the video output is shut off, and that prevents the PC from restarting.

Could be true, in my friend's case, I suppose. He did have some Windows Updates, and something about a hardware driver update, maybe a video driver update delivered via Windows Update, because he hasn't kept up with updating his video drivers.

--

I think to start off with, I'll try removing (and testing the voltage, if I remember to bring my meter over) his CMOS battery, and resetting the settings, and then (or maybe before), updating his video drivers to the newest NV drivers. (What with all of the Meltdown / Spectre patches happening lately, that might be affecting things too. Actually, I think that there were some issues with the patches with older AMD-based PCs, but I thought that those reports were talking about Athlon XP-era PCs, and not Athlon II / AM3 CPUs, which are quite a bit more modern and recent.)

Edit: I have a more and more suspicious feeling, that this issue is in fact a software issue, due to the fact that it's possible that drivers / services aren't shutting down properly and restarting, due to the fact that he needed to update all his drivers with post-meltdown versions, since the OS-level patches interfered with drivers. At the very least, his NV drivers.

 

[Ketchup]

Hard to say without seeing the Windows event viewer. If you take a look at it and it only shows improper shutdown, or notes no problem at all, it's probably the power supply.

Also, just as a note, the power supply in the desktop I use as a server has been doing this for quite a while (Windows is done and has cut off the display and the hard drive, but power to the fans stayed on). Finally got a new power supply for it recently and haven't had another issue. Windows never noted an improper shutdown because, in it's mind, there wasn't one.

 

[VirtualLarry]

Thanks for the tips. I'll check the Event Viewer, next time I'm over there (or if he lets me remote in later today to check things out).

I was subconsciously leaning towards it being a PSU problem, but I was consciously dismissing it because I had replaced the PSU only a year or two ago, and it has a 5-year warranty. OTOH, it IS a ThermalTake.

I mean, the mobo is 10 years old, but it's held up thus far, and he's not having any other issues with it that I know of, so probably it's not at fault?

 

[aigomorla]

if its hanging at post its:

1. CPU
2. Voltage
3. Ram

You can rule out videocard, because if its displaying its not video.

But first i would try to give the CPU more voltage and see if that fixes something.
Then i would use a PSU tester to see if the psu is holding 12V 5V 3.3V rails on idle, since again, its a posting issue.

Then lastly, i would run memtest on the ram in another machine to rule out ram.

 

[VirtualLarry]

It's not necessarily a POST hang, it may be hanging on Windows restart (before it give the BIOS the actual ACPI command to restart), in a system service or driver.

He says that hitting RESET, it then comes up. So, TBH, that doesn't exactly sound like a POST hang, unless it also happens from a cold boot,or he has to hit RESET multiple times. I'll inquire more, when I can get a hold of him.

OTOH, it could be a POST hang, due to his cheap chinese DDR2. (Maybe the CPU's IMC need a voltage boost?)

A POST diag card or mobo diag LED would be useful here.

 

[aigomorla]

just try to raise voltage here and there on a few things, and see if it fixes it.

If its a really old PC, it could be hardware degradation.

 

[VirtualLarry]

It's not overclocked, nor has it been.

 

[aigomorla]

It's not overclocked, nor has it been.

hw degradation is still hw degradation.

who says you need to be overclocked to see or experience it?

circuits are like highways, the more they are traveled on, they get worn out.
Compound the factor of that with heat and it accelerates.
Things get worn out and die, Caps become less efficient and don't work as intended.

If you can go into bios, i would just bump up voltage up 2 notchs from default and see if that fixes anything.
What else do you honestly have to lose and its a quick and simple test to see if it is hw degradation.

 

[VirtualLarry]

If you can go into bios, i would just bump up voltage up 2 notchs from default and see if that fixes anything.

I'll try that next time that I'm over there. I can't really do that via TeamViewer (access BIOS settings), and I don't trust my friend to muck about in the BIOS over the phone, he has enough trouble navigating the NVidia sit to download drivers via their selector-widget.