Weird problem - maybe PSU dying?

lulubel

Junior Member
Dec 9, 2014
1
0
0
My PC is 2 years old, built with a Corsair GS500 PSU and Asus P8Z77 mobo. I've never had any problems until a couple of months ago, I started to get intermittent graphics driver related errors that froze the screen and crashed my browser. I did some research and it seemed to be something to do with Java, so I made sure the drivers and Java were all up to date and it still did it, but not so often.

Also 2 months ago, my PC failed to boot for the first time. It powered on, went through its checks, 3 red lights - CPU, RAM and GPU (I think) - there was a brief whirring from the graphics card fan, then it started the checks again, and kept cycling through them until I did a hard shut down. On the second attempt it started fine. A few weeks later, it did it again.

My first thought was the hard drive was on the way out because that was the point it seemed to fail to boot, but I didn't do anything about it.

A couple of weeks later, it did it again. Only this time, I'd logged into Windows and just left it for a few minutes before using it. It crashed and started cycling through its checks.

On Sunday morning, I got up, opened the windows as usual. It was cold, and I heard the plastic window on the PC case creaking as the cold air hit it. PC wouldn't start, and I tried several times before stomping off to get my laptop and try and figure out what was wrong.

I suspected it might be shorting out. The house is very dusty and there's lots of pet hair. Even with filters on the intake fans, I have to open it up and clean it out every 2 or 3 months. I took the whole thing apart, took the mobo out of the case to check behind it, then tried booting it with the graphics card removed. There wasn't much dust and hair inside and it still wouldn't boot.

Then I went with my gut (no logical reason for this) and got a 2.5kW electric heater and put it beside it, almost touching the case. It was on the side where the side case fan would be if I'd fitted one. I heard the case creaking again as the heater got going.

An hour later I tried to start it, and it booted fine.

It will run with the heater beside it, but if I try to take it away, it crashes and goes back to cycling through its checks. If I turn the heater down it keeps going for a while, then crashes. I tried moving the heater to the front of it, so the front intake fans (3) would draw the warm air in, but it crashed in a few minutes. I have the odd situation of the mobo running at 25C and the CPU at 50C because the CPU fan is drawing in the warm air!

It did crash once with the heater running, and that was when I plugged a USB stick in one of the front ports. I checked it a few times with both front ports, and it crashed every time, but I didn't think to try the rear ports at the time. When I thought of that a while later, it didn't crash on the rear or front ones, and hasn't done it since.

I've read about condensation on warm hardware in cold conditions (the air can be very damp here), but I don't see how it can be that because it should be well dried out by now and its still crashing.

I wonder if the PSU may be on its way out. We do get some power fluctuations here. It has a surge protector on it, but nothing to protect against drops in power, and I know that can kill PSUs because I've lost one to a power dip before.

I know this is long, but I wanted to go into as much detail as possible because this is the weirdest thing I've ever come across - a computer that wants to be hot!

Any thoughts would be welcome.
 

westom

Senior member
Apr 25, 2009
517
0
71
It will run with the heater beside it, but if I try to take it away, it crashes and goes back to cycling through its checks. If I turn the heater down it keeps going for a while, then crashes.

Heat is a diagnostic tool. For example, a computer may work fine in a 70 degree room. But fail in a 100 degree room. Many then blame heat for damage. Nonsense. Heat has identified a defective semiconductor that fails today only at 100 degrees. And will start failing maybe a year later at 70 degrees.

Generally, hotter locates defective semiconductors. Cool tends to find defective connections (cold solder joints, bad connector, etc)

To say more means you must provide some hard numbers. Your symptoms are typical of a problem that causes the power controller to respond. Why is it responding? By requesting instructions, get a digital meter, and do one full minute of labor. The posted numbers will tell others what exactly is happening - what the power controller is doing. Without those numbers, the better informed can only stay silent.

Power controller is only one component of a power system. PSU is but another component. Nobody can say anything more (honestly) without those numbers.

You can also try selectively heating individual system parts with a hair dryer on high. Highest hairdryer heat is normal and ideal temperatures for every computer part. In your case, selective heating may locate the physical region of that defect.
 
Last edited:

PhIlLy ChEeSe

Senior member
Apr 1, 2013
962
0
0
When you used it before you had issue, how long would it run? Easiest way is to trouble shoot it part by part, if you live near a computer supply place (best buy ETC) buy a PSU and swap it out. Try disconnecting anything not needed to boot such as CD player Etc, more then likely it is an aging issue which means CAPS in the PSU are aging/changing.