Author Options:

weird PC power problem? Answered

I'm experiencing a weird problem with my PC.
recently (about 2 months ago) i have upgraded my video card to a Sapphire 7970 and about 2 weeks ago installed a second monitor to my system (dual D2342P , added one first, bought new desk, added second one).

in the past 2 weeks i have had no problems at all.
in the past 2 DAYS my PC has randomly shutdown abut 6 times (most recent 1hr before this was posted).
the first few were relatively close together, about 2 - 4 hr apart.
the most recent ones have been almost exactly 12 hrs apart. (~11 am - ~11pm) .

yesterday (Monday) i replaced all 7 case fans and cleaned the filters. (they needed doing. running for almost 6 years)

currently have disconnected second monitor connected by D-Sub and CPU temp has dropped by 12C (72C - 60C)

if needed i will post the AIDA summary.

Motherboard: ASUS Commando
CPU: Intel Q6600
Ram: 4x 2GB sticks (2x Micron, 2x Generic)
OS: Windows 7 Ultimate (genuine)

At time of posting CPU temp  is 69C



Best Answer 6 years ago

Sounds like your CPU is overheating. I had this problem a few months ago, thought it was a graphics card issue because it only happened in games, turns out the extra load of gaming raised the CPU temp. To solve it I simple removed my CPU, cleaned all the old thermal paste off and then applied new paste and reseated the heat sink/fan. I've been crash free ever since but you need to do this NOW. My brother started having similar shutdown problems and he waited to long and it cooked both is CPU and mobo.

Also, make sure that your fans are setup to maximize flow (some should vent while others feed), and space out any PCI cards and front bay items (DVD drives, HDDs) if possible.

this is my case Xaser III 1000D.

the fan layout is
Front 2: intake
Top: Intake
Side: Top, intake (onto CPU)
Side: Bottom, Exhaust (Graphics card)
Rear: Exhaust

at the moment the side is off.

i will have to dig for some Thermal Paste.
on next shutdown, i will go for the thermal paste.


6 years ago

extra info,

current Mobo, CPU & Micron Ram are both ~6 years old with about 90% uptime +- 5%

I can't reply to your case fan layout, but it should be as follows for optimal airflow.

Front 2: Intake
Side Top: Intake
Side Bottom: Intake
Top: Exhaust
Rear 2: Exhaust

This way the air moves from front to back with no disruptions.  It is always best to have a 'pressurized' case, 1 more intake fan than exhaust.

Make sure you always run with the side on the tower, it keeps the airflow clean and consistent in the tower.  Without it, the air will just flow out the side without reaching your CPU.

The front and sides should have filters that will keep the dust from getting to your CPU heatsink. Running one in, one out on the side does nothing for cooling. The air simply goes from one fan into the other without supplying air to anything.

With regards to thermal paste, I highly recommend getting decent paste, not the cheapest. Even toothpaste is better than some of the cheapest paste. Many people have great success with Arctic silver 5.  Make sure you clean the old paste off the CPU and heatsink before you apply new.

Updating the bios probably won't help anything.  I never recommend updating the bios 'just cause'.  Only update if you need a feature that is described in the revisions.

One thing I forgot to mention, your GPU fan exhausts out the back of the PCIE slot. It needs air coming at it to be effective. By running the bottom side fan as exhaust, the 2 fans (case and GPU) are fighting for air and neither one is getting the air it needs.

side question.
would updating the BIOS help?
it's still running "stock" (2007)

the latest is 2009

no shutdown for ~16hrs

(i know, i just taunted Murphy)

Check the motherboard for bad capacitors. The capacitors leak/bulge
at the top when they go bad.

Run a memory test. If the test fails, replace memory.
If memory test fails with new memory,Replace CPU.

There is one component most people overlook and drastically affects your entire system. What PSU do you have in the system?

You have a fairly power hungry PC there and if the PSU isn't large enough to support it all then you will have all sorts of issues that can look like problems being caused by other systems. After about 5 years of use a PSU starts to decline in its abilities to offer good stable power. When you added the new video card you where probably sitting on the edge of the PSU's abilities and it is finally giving out on you. You will want a 650W or larger PSU for your system.

Make sure you get a good name brand like a Corsair. If your spending less then $50 then its not a good PSU. 

950W should be fine assuming it's not old, as mpilchfamily mentioned.

Not all 950W PSU are created equally. From what i can find it is on the cheap side as far as price comparison of known good 900+W units. So the company has likely used some cheaper parts in it that could be causing the problem.

Like i said after about 5 years PSU start to decline. Often companies like Hytec shave money on the cost by using cheap generic capacitors. These do not stand up well over time and can either fail or not do an adequate job of filtering the voltages. So you may be getting allot of ripple on the lines causing the instability in your system.

Regardless of the build quality of the unit, it's the age of it and the problems you're having that indicate you need to replace the PSU. You don't want to be putting a nice new high end GPU on a 5 year old PSU. The problems your having can damage the card or the entire system.

you got in about 5 minutes before me.
the PSU is only 2 months old ( probably sitting on the shelf for 6)

it's about 2 months old. ( 1 week younger than the graphics card. lol)

A power hungry graphics card running two monitors will suck power like a Hoover sucks dust. I wouldn't use anything smaller than a 750 watt and would opt for a 1000 watt PSU if at all possible.

Shutdowns like that can be caused by a driver conflict. You did just install a new driver with the new card so that could be all the cause of the problem.

Running a second monitor will tax the system more but I've never measured my system to find out if the temp goes higher.  I run 2 monitors and sometime 3 on my office system.

Check that your memory is seated well and is working right.  I had an issue recently with random shutdowns that would only occur when running a certain software with another software.  I thought it was a software conflict but found out that one of my memory cards was not seated right and the problem didn't show up until caused by the software.