Overheating, Bad PSUs and a headache.

[jwescott425]

My computer has a problem and I don't know what wrong! It will lock up (music or sound will begin to stutter and loop) and I can still move my mouse, and then it will eventually fully lock up where I can see the screen, but can't move the mouse or hear any audio. The only way to get out of it is to hard reset the computer. I have a few hypothesis;

Either my motherboard's chipset is cooking (least likely), or my PSU is wimping out (most likely).

System specs:
Mobo: MSI K9N6PGM2-v2
CPU: AMD Athlon II X2 255
RAM: 1 stick of Kingston PC-6400 2GB (with nifty blue heatsink)
Hard drive: Some old 160GB Seagate harddrive I've had for about 6 years
DVD: HPdvd1270 lightScribe burner
Video Card: nVidia GeForce GT 240 1GB EVGA
PSU: Coolmax V-600 (600watt)

Like most motherboards there's no way for me to monitor the temperature of the chip other than physically feeling it and when I feel the heatsink on the chip, while it isn't scalding hot, is considerably hotter than any other component in my computer. The heatsink is the stock aluminum MSI heatsink.

The only reason why I suspect the heat issue at all is because I just moved my computer into my room from our home office. I live in, as Johnny Carson used to say, beautiful Downtown Burbank.. which happens to be as hot as heck during the summer. The past week it has been about 95' F (35' C) and is only get hotter as we progress through July! My room is significantly hotter than the office, because the office is a part of the original construction circa 1920's and just seems to be colder all times of the year.

As for my PSU, I ran HWMonitor to check my 12V readings (yeah yeah I know it isn't exactly accurate but it can give me a ballpark) and it shows that, while running only HWMonitor and Firefox, the voltage would dip down as low as 10.60v. Which I hear is no bueno. Boo.

And I just reformatted my computer because it was well over-due and in hopes that it would possibly clear stuff up.. It has oh-so-very slightly, but not enough.

Any thoughts on this? Should I shell out $10 on a heatsink for my mobo chipset just to be safe? Is my PSU to blame?

 

[C0RR0SIVE]

No name PSU could be pussing out. Get a voltmeter/multimeter and measure the 12v off the psu by hand while the PC is on, be careful not to short anything, and DO NOT open the PSU, just measure off of the yellow and black lines.

Also, check for bulging capacitors on the motherboard.

 

[Slaymate]

Coolmax is a really crappy power supply maker.

 

[jwescott425]

Coolmax is a really crappy power supply maker.

Poor guy! He's trying his best...

So I Googled a guide on how to properly do this, and the first one said to test it with the ATX unplugged. I tried that.. and got really wonky numbers. Further investigation said to test with the computer ON (which makes more sense.. go go Gadget LOGIC) so I plugged in my ATX cable again, fired my computer up and..

12.1v All throughout POST, Startup, talking to you guys on Firefox. I fired up D2 (quite the labor intensive game, I know) and it dipped down to 12v. And like magic my computer is being oh so very stable.

So what I think happened is actually that my ATX cable may have been bumped loose in transit from my office to here. It is a 20+4 connector, and the 4 pronged portion doesn't really lock in to the other 20; there is just a guide rail that helps it sorta stay in place. Is it feasible that something like that would cause the problems I was having? :O

Anywho.. Thanks for the help folks. If it starts acting up again I'll be back

 

[C0RR0SIVE]

It is feasible that it happened. If you dont have enough power getting to your motherboard, that will happen, or if the power isn't clean.

Try downloading and running Prime 95 while checking your voltages.

 

[Digerati]

Like most motherboards there's no way for me to monitor the temperature of the chip other than physically feeling it

You mean "unlike" most motherboards as clearly, most motherboards and CPUs do have sensors that can be monitored by the HW monitor provided with the motherboard (see your MSI utilities disk) or computer maker, or by one of the many 3rd party HW monitors out there like CoreTemp (temps only), SpeedFan and Motherboard Monitor.

Testing any PSU with a multimeter is really not a good or conclusive way to test a power supply. Like ALL power supplies (including PC PSUs, gasoline engines and batteries), the proper way to measure the output is when the supply is under various "realistic loads". A bad power supply will often report good voltages with no load, then drop drastically as soon as any load is applied. To measure properly with a multimeter, you have to have the PSU connected to the motherboard, then jam hard, sharp, highly conductive probes into the connector, deep in the heart of the computer. This is very risky, even with steady hands. One slip and a dozen circuit traces could be damaged - not to mention potentials for ESD damage.

But even if skilled with a meter and have a steady hand, that test is not conclusive because typical multimeters cannot test for ripple and other anomalies. The PSU may "appear" to be suppling correct voltages but excessive ripple is often the cause of unexplained computer problems likes freezes, sudden reboots, no-boots, and sudden shutdowns. So proper, conclusive testing can only be done a qualified technician using an oscilloscope or a power supply analyzer - sophisticated (and expensive) electronic test equipment requiring special training to operate, and a basic knowledge of electronics theory to understand the results. Therefore, conclusively testing a power supply is done in properly equipped electronic repair facilities.

The better method for most "normal" users is to swap in a known good supply (with necessary ESD precautions), take it to a shop where it can be professionally tested by a trained electronics technician (often done free or for a nominal fee), or use a PSU tester, preferably one with a digital readout such as one of these.

I keep one of those testers in my toolbag in the truck. They too are NOT conclusive because they do not test for ripple either. They do incorporate a dummy load however, though it is not large enough to be called "realistic". It is however, much more realistic than the "no" load provided by a multimeter.

Other advantages of these testers is they are keyed so you don't need to know the connector pin-out specifications or number patterns. They are also great for enabling a spare PSU to start up, then used for fan and drive motor testing.

No doubt, if all you have is a meter you can verify your PSU is outputting voltages (or not) - you just cannot tell, or assume the voltages are "clean" and stable under load.

DO NOT open the PSU

I am often amazed how so many people think their computers are harmless. If it plugs into the wall, it can kill! There are deadly voltages inside the PSU. I agree with c0rr0sive - unless you are a certified electronics technician, stay out of the PSU! If you stick anything in there, make it wooden glue stick to hold the fan still while you blast out all the accumulated heat-trapping dust.

 

[C0RR0SIVE]

If you have ever worked on a PSU, you will find, that any 12v+ and ground is all shared inside, unless it's a multi-rail PSU. Hence the reason I said check for the voltages under a heavy load, being Prim95, or something that can bog the processor it self down. The capacitors on the motherboard generally do a decent job at cleaning up the less than desirable current.

 

[jwescott425]

The problem has been resolved. Loose cable and for whatever reason my PSU was set to 220v instead of 115v. Go figure. Thanks gents.

 

[C0RR0SIVE]

That switch doesn't just move on it's own...

 

[GhOsT1321]

The problem has been resolved. Loose cable and for whatever reason my PSU was set to 220v instead of 115v. Go figure. Thanks gents.

If your switch was set to 220v and you turned the computer one, you would know right away that something was wrong. If the PSU was turned on while it was set to this voltage, chances are that you did damage to it.
Are you sure that the switch wasn't in between 220v and 115v, but was actually receiving 115v from the start?