Mobo or PSU go boom...?

Biggilus

Biggilus

Biggilus

Associate
Joined
14 Aug 2009
Posts
197
Hi! I'm rubbish when pcs go wrong and yesterday mine did, so I'm here for help please.

Out of nowhere the pc cut out and died mid-raid yesterday. Went to turn it back on and the front lights flickered on for a split second, as did the fans start turning again for a split second then nothing. Repeatedly pressed power button but all dead from this point on. Left it an hour or so and magically she came to life, however on the boot screen i was prompted to do a BIOS recovery (internal clock on BIOS showed 01/01/2009, time 00:00) and I have no idea how to go about this so I F10'd (save+exit) and Windows booted fine. All was working fine. Fired up FF14 left it idling for almost an hour but then it cut out just like before. Turn it back on, same instant flicker of light and fan activity then nothing. Now it's the morning and she's working alright for now (10 mins) but I am waiting for it to die once more.

Does the above point to any component in particular? Dodgy PSU? Mobo on it's last legs? Maybe something else?

On the plus side the wife took pity and gave me the green light to upgrade what was needed..

I'm thinking a new mobo/cpu/ram is on the cards but I am first looking for sage advice here to what the actual problem might be.
 
Given that it works ok for a bit, cuts out for a bit, then starts working ok when left alone my money is on possible temperature issues. What's the cooling like? What hardware you running etc. Does it cut out with the case side off would be a good first test, could be something over-heating and causing the shutdown.
 
Tried with the side case off and a desktop fan blowing into it, and still she cuts out :(. Mobo clock reset back to 1/1/09 again too.

So on the inside:
i5-4690K 3.5GHz (not clocked atm)
Corsair Vengeance Pro Red 16GB PC3-19200 2400 MHz
Radeon (TM) RX 480 Graphics 8Gb
Asus Z97-k
Corsair 750W PSU (Modular) - Low Noise

Had the system for 6 years ish now pretty much trouble free, upgraded to this graphics card a couple of years ago I think too.
 
I agree with @Dave Burton that it's likely a temperature issue. Check that all the fans - particularly the CPU fans - are working. Which CPU cooler do you have? Is it an AIO cooler? If so then check that the pump is still working and that there is still sufficient fluid in the cooler. You could also try refreshing the thermal compound - after 6 years it will likely have dried out.

Do be careful. I have cooked CPUs before.
 
Thanks for the input so far gents.

It does seem to be behaving like a temp issue - idling and browsing for 75 mins no issues, fired up FF14 and within 10 mins she cut out. I monitor the CPU temps and idling sitting at around 40C, ingame 60C so this still seems alright to me?

I have just given a bit of TLC too - cleaned and reapplied thermal paste (the crusty gunk I saw couldn't have been good) and cleaned out what seemed like half a cat with the amount of dust everywhere - the heatsink blades were completely clogged.

So here's hoping that's all it needed, I'm not convinced though... half empty kinda guy

E: With respect to the CPU fan, many times the PC wouldn't boot due to "CPU fan error!" so I would jump into BIOS and see it had the chassis fans RPM listed but the CPU had N/A - which is weird because I could see the CPU fan spinning. I left that setting on ignore - doesn't sound wise but the fan still spins despite the warning saying it isn't. Another way around this I found was multiple system resets and once in a while it would see a fan was indeed working and boot fine.
 
Good job on clearing the dust from the heatsink. That will probably sort things out. But do check that the CPU fans are plugged into the correct headers on the motherboard.
 
Thanks everyone! Saved me splashing out on new gear - good old dusting and pasting and she's stable again. Monitored temps look the same, still 40ish idling and 55-60 in game (not the most demanding of games either). I've had the CPU temp 90+ in other games before for prolonged periods of time (I know I know that's bad) so quite why heat was an issue at 60ish doesn't make sense to me.

If it was the PSU dying - I assume absolutely zero signs of life would be seen from that point on? But what gives it away if the mobo was at fault? Again I didn't think heat because my last rig would alarm very loudly when it got hot (heatsink was misaligned... not my fault on that one though!)

Thanks again :)
 
60c is perfectly fine for the cpu so im not sure how it could be a heat issue.

I would be tempted to update the bios on the motherboard

A psu doesn't have to fail catastrophicy it could be faulty, for example an intermittent fault as the psu heats up something expands inside causing insuficent power delivery.
 
Of course I spoke too soon. Was going strong for 3 hours or so then bam :/. Waited a wee while, booted up again and sat idling for another few hours all good - but within 5 mins of gaming it went down. The most annoying bit about this is the insane BOOM that comes through my speakers whenever it turns off either like this or regular shutting down scares the bejeezus out of me every time lol

If it was the graphics card overheating it wouldn't behave like this would it? Is it sensible to think either PSU, mobo or cpu?
 
Ok downloaded CPUID HWMonitor and it says the mobo is at 114 Celsius....
----------------------------
Temperature 0 114 degC (237 degF) [0x72] (Mainboard)
Temperature 1 38 degC (100 degF) [0x26] (CPU)
Temperature 2 114 degC (237 degF) [0x72] (TMPIN2)
Temperature 3 36 degC (96 degF) [0x24] (TMPIN3)
Temperature 4 30 degC (86 degF) [0x1E] (TMPIN4)
Temperature 5 107 degC (224 degF) [0x6B] (TMPIN5)
Temperature 6 105 degC (221 degF) [0x69] (TMPIN6)
-----------------------------

Ummm? Is that even possible?

E: Can't be true, rebooted and went to BIOS as soon as I could so we're talking max of ten seconds and the BIOS stated it was 30 Celsius.

Also found out CMOS batteries are dirt cheap so yep I'll replace that this weekend, no idea if it will have an impact on the issues mentioned but stranger things have happened I guess
 
Last edited:
Still struggling here and using this moan to bump.
Replaced the CMOS battery so at least now when it does cut out the time and date are reserved.

Been fine with the wife working on work stuff all day, but the moment I enter FF14 it lasts less than five minutes - quite why it lasted around 4 hours during a gaming session yesterday I don't know.
Could the graphics card cause an instant shutdown like this?
Any other suggestions for narrowing down what may be the cause?

E: Happened in Nier also. Was in windowed mode so watching GPU and CPU temps like a hawk. CPU went up to 63C quick enough but stayed there for a few mins, GPU up to 53 and stayed there too until the machine cut out. No last second spikes or anything temp wise.

Also possibly another bit of useless info - when it cuts out it takes a few mins before I can power back on - however it *always* involves me turning off the PSU switch and back on, even if I leave it an hour without doing this I can't restart it, but a quick off then on at the PSU and it will start up (but it still need a few mins after a fresh cut out).
 
Last edited:
Did a few tests with Passmark BurnIn test - memory was fine, but the machine crashed doing the CPU test and also the 3D graphics test on separate occasions.
 
Managed to crash the pc running the Burn In test on both the CPU and 3D graphics. However when I removed the graphics card I coudl run the CPU test again and again with no issues. Pop it back in and crashed after six or so minutes.

How does running just a CPU test cause this? Is the graphics card at fault or is it truly a PSU issue the fact it behaves like this?
 
You must log in or register to reply here.
Back
Top Bottom