PC overheated, kaput?

MayoForSam

I think my son's PC may have overheated and caused some damage .

These were his specs up to yesterday:

• Phenom II X4 965 BE
• Arctic Cooling Freezer 7 Pro
• MSI 770-C45 motherboard
• 8GB (4x2) Corsair Twin3X4096-1333C9A DDR3
• Crucial 256GB M4SSD
• Samsung SpinPoint F3 500GB
• EVGA GTX 650 Ti Boost SC 1GB GDDR5
• CX430 — 80 PLUS® Bronze Certified Power Supply

Yesterday, he upgraded his graphics card to a PowerColor HD 7870 2GB. Everthing installed fine, no issues.

Then this evening, during a particularly intense game of BF4, it shut off without warning. When I inspected it, I straightaway noticed the case was VERY warm. When I opened it up, the cooler, RAM and graphics card were all very hot to touch (above 50C).

I left it cool down for an hour and tried to power it up again. Nada - not even a flash or a beep. I checked the PSU using the paper clip trick and it seems to be fine. I then removed the RAM and stuck it into another PC (Intel i5) and noticed that one stick was dodgy, the other one was OK. I put the good stick into my son's PC and tried again, brief flash of power but then nothing. I suspect the CPU and/or the motherboard may be fried.

I am wondering how did the PC get so hot without shutting off earlier at a lower temperature?

Any ideas about how I can further isolate the problem - CPU or MB? There's no visible damage to the MB.

Was the PSU a bit underpowered for the HD 7870?

Digital Solitude

430w is enough for that GPU. Seating the CPU in a new motherboard to see if it works then is the best idea. Won't need a fan as all you want to do is see if it starts to boot. Test motherboard and CPU, see what needs replacing.

MayoForSam

Digital Solitude wrote: »

Test motherboard and CPU, see what needs replacing.

I don't have access to a spare AMD motherboard - do you reckon this is probably blown rather than the CPU?

Digital Solitude

Any friends local? Or a PC repair shop might have an FM2 motherboard (I think its FM2).

Worst case scenario you could pick up a dirt cheap CPU and test it, return it and get the replacement. Its hard to say, but likely the CPU.

You've tried it without the GPU I'm guessing?

maki

It's almost certainly not the CPU. CPU failure rates are extremely rare compared to the rest of the components within a PC, and your 50 degree running temperature is hot for a Phenom, but still well within limits.

It sounds like the PSU's overcurrent protection is kicking in, which could be caused by the PSU itself or the motherboard, or even the graphics card. Take out the graphics card and disconnect all drives so you're just testing with the motherboard, CPU, and RAM.

Next step would be a visual inspection of the motherboard, taking care to look out for any bulging or leaking capacitors.

MayoForSam

Thanks, I'll try a few other things tomorrow. Putting in the old graphics card made no difference. Is it just a coincidence that the 7870 went in yesterday? What would cause everything to start overheating so dramatically all of a sudden? One stick of ram is definitely busted, high temperatures would have caused this I presume?

maki

MayoForSam wrote: »

Thanks, I'll try a few other things tomorrow. Putting in the old graphics card made no difference. Is it just a coincidence that the 7870 went in yesterday? What would cause everything to start overheating so dramatically all of a sudden? One stick of ram is definitely busted, high temperatures would have caused this I presume?

Don't worry about heat. 50 degrees really isn't very hot for gaming. Graphics chips are rated for 100 degrees or over. Your CPU has the lowest heat limit at around 62 degrees, and will clock itself down or simply shut off the PC if it hits that.

The 7870 has a 500W minimum power requirement first of all. These are usually conservative, but the CX430 is a budget PSU and you may be pushing it too far.

RAM can just die. There are too many variables at play to narrow down exactly why. Could be just co-incidence.

myrscossie

This happened to me twice on different rigs, upgraded the gpu but not the power supply thinking it was big enough, the first rig had thermaltake 480w which died with a big bang shooting pieces out the back, just needed a new psu and dvd drive to fix, then a seasonic 600w died with alot of sparks coming out the back like a firework killed a LG drive again and 1 2gb stick of ram got lucky both times as I know a friend it happen to lost everything other then the cpu and memory. when a PSU dies it can kill any part connected to it as far as I see the better quality PSU the less chance of it damaging everything as I think my friend had a cheaper PSU.

MayoForSam

Just to follow up on this - the graphics card is OK but the motherboard is burnt out, there is visible damage around the VRM's. I can't test the CPU but it is probably OK.

So my options are (a) try and do a repair on the cheap by just replacing the MB and getting a new stick of RAM (same PSU) or (b) upgrading the motherboard/CPU (probably go Intel rather than AMD), higher wattage PSU and a new pair of RAM modules.

Any thoughts? I might be running a risk of another burn-out if I stick with AM3+, the HD7870 and the 430W PSU.

riclad

Replace the psu with a 600 watt plus unit,
buy a new motherboard ,
buy new ram.
Make sure the psu is the right size, shape to fit inside the pc case .
Are the psu screws holes in the same place as the old one .
examine the cpu, see is there any black scorch marks on it .
the cpu is probably ok.

Digital Solitude

No need for bigger PSU if you jump to Intel ship. Probably hit 500w on AMD might be better, 600w is too much. Thermaltake Hamburg 530w is about €45

MayoForSam

I've actually gone ahead and ordered an i3-4160, motherboard and the 530W Thermaltake PSU - not taking any chances.

Digital Solitude

No risk then so, best of luck with it