PC acting strange - could it be the PSU?

Gizmo

Do you have "fast boot" enabled in your Bios/UEFI ? if so try disabling it as ive seen this issue with a few Nv Maxwell/Pascal based cards

Philtre

Gizmo said:

Do you have "fast boot" enabled in your Bios/UEFI ? if so try disabling it as ive seen this issue with a few Nv Maxwell/Pascal based cards

Done that. Thanks, for suggestion, though.

Gizmo

Do you have any PCI-e "Gen" setting in your bios? my sister's PC had an issue like this with an old ati 7970 i gave her 
Setting she had were auto/gen2/gen3 any thing other than gen2 caused a blackscreen on boot.

Also do you have a 2nd PCI-e slot to try the 970 in?

Have you tried slacking the Dram CAS/RAS timings?

Philtre

Gizmo said:

Do you have any PCI-e "Gen" setting in your bios? my sister's PC had an issue like this with an old ati 7970 i gave her 
Setting she had were auto/gen2/gen3 any thing other than gen2 caused a blackscreen on boot.

Also do you have a 2nd PCI-e slot to try the 970 in?

Have you tried slacking the Dram CAS/RAS timings?

BIOS settings all OK.

Tried the 2nd PCI-e slot and still the same problem.

Will check the CAS timings.

Thanks for all suggestions.

It's been fine for nearly 3 years, until fairly recently. It's going to be either GFX card, PSU, or Mobo.

crunchman

Could be the RAM. I had a PC years ago where it would run for a few minutes and reboot itself with no warning. Turned out to be a dodgy dimm.

That was a bit easier to diagnose as I'd only recently added the extra RAM.

dano

Philtre said:

Heartfeltdawn said:

Does the machine still trip the mains breaker with the graphics card out?

Don't know, it happens so sporadically. But I know what you mean - perhaps it's pulling too much juice.

 Is it an an MCB or RCD breaker tripping ?

Does it trip the breaker at startup? Is there much else on the same breaker loading it too ? 

May be a failing cap in the psu from live or neutral to earth tripping the RCD.

Philtre

@dano It's a MCB breaker. On that breaker there's the PC and router, and not much else when it trips. Happens once a week maybe. Very sporadic.

darthed1981

I'd be in the car with those suggesting trying a new PSU.  I've had these symptoms a few times with a PSU on it's way out.

As others have suggested, make sure you check and re-seat everything.  It is very easy to knock a memory module or even the CPU loose, or the PSU fan which would cause the PC to overheat and the CPU turn itself off.

You can also get a processor diagnostic tool from Intel which is really useful, https://downloadcenter.intel.com/download/19792/Intel-Processor-Diagnostic-Tool which will help you rule out things.

You can get memory testing tools very easily as well, and speedfan lets you check temperatures and also look at SMART data from your SSD http://download.cnet.com/SpeedFan/3000-2094_4-10067444.html

Gizmo

Humm other things i might be looking at if i were TS'ing this on a bench....

Anyway of installing another clean OS (windows 10 is as mad as a box of frogs,and with M$ pushing all kinds of bullshit driver updates/fixes/and intels latest band aid fixes without proper testing id be tempted to try another clean OS install)  

Check is to see if your 970 bios is upto date (NV inspector can give you the details) as i recall Nv fixed some boot high to idle clocking/speed issues with there drivers a while ago (not to sure if that was just with 1070's though) and this was throwing up some error with cards running older or custom bios.....

How does it run under heavy loads btw ? if it was PSU id expect it to trip out within a few mins of gaming/benching

Other than that id look at testing the 970 in a 2nd machine,if its still doing in on other boards see about an RMA 

Philtre

@Gizmo will check this things soon.

The interesting thing is that the freeze and reboot only happens at the start of using the PC, and only started doing this after I update the BIOS firmware to fix the RAM POST led issue. Once the computer gets going it's OK.

I might roll back the BIOS firmware...

equalsql

Are the heatsinks on the CPU and graphics cards jammed with dust?
You might need to remove the fans and brush or blow out any crap between the fins.
I've brought back quite a few 'dead' PCs doing this.

Philtre

Cleaned the whole thing out with a hoover today. And rebuilt it. Even opened up the PSU and cleaned that out,

Philtre

I think it's the BIOS firmware version playing funny buggers.

I already mentioned that the PC had been running fine until I updated the BIOS firmware to the latest version. I did this to cure a boot RAM led POST error.

Take a look at the BIOS firmware versions for this Z97-A mobo:

https://www.asus.com/uk/Motherboards/Z97A/HelpDesk_BIOS/   (click on "See All Downloads")

I originally had Version 2205 (the oldest version) and all was fine. So I updated to Version 2801 and that's when I got the random reboots with the NVidia GTX970 installed. This BIOS version has a mysterious "Improve system stability" note. So I tried downgrading to Version 2601, which also has a "Improve system stability" note. Still got the random reboots. So I then downgraded to Version 2501 and everything has been fine since yesterday.

Could the mysterious "Improve system stability" BIOS updates be conflicting with something in Windows 10 or the NVidia card? Has Microsoft patched Windows 10 recently with something that might conflict?