A plea to the computer guys

[III./ZG76_Keller]

I've been having an intermittent issue with my PC; I had this problem a few months ago and it seemed to have gone away on its own; but now it’s back.

What happens is that my computer will just turn off and restart, like a quick power failure. I’ve checked the event viewer and it comes up as Kernel Power failure ID 41.
Does anyone know anything that I should be checking?

I’m testing right now to see if it’s an issue with my UPS somehow cutting out, so I’ve switched my CPU from one UPS to the other one.

Anything else come to mind?

The major changes to my PC in the past year were the addition of 4 Gigs of ram and the switch from ATI 5770 to 6950. I upped the power supply from a Corsair 650 to a Corsair 850 as well.

Timeline:
Changed 5770 to 6950,
problem starts about 1 week after the change,
changed power supply, (thought this might fix the issue)
two weeks later added 4 Gigs of ram

[III./ZG76_Keller]

Alright, so it's not the UPS as it just did it again. :cussing:

[Tonka]

Any overclocking or anything going on? Is the new memory you bought set up correctly in the BIOS?

First thing to do is to go to Control Panel -> System, then click on Advanced System Settings on the left hand side.

In the little window that opens, look for the third row, StartUp and Recovery, (System startup, system failure, and debugging information) and click the settings button.

In that window that opens, look for the checkbox 'Automatically Restart', under the System Failure part. Uncheck the 'automatically restart' box, make sure you leave everything else as it was. Click OK, and then OK again in the first window that opened to close it.

From now on, your system will not restart automatically when it encounters a problem. It will display an error message and you'll have to restart it yourself. Make a note of the error message though, as you can use it for further diagnostics.

If you do still get problems, make sure you have your new memory setup with the correct timings in your BIOS. Google the model numbers to find the correct settings. If you are sure it is configured correctly, download and run Memtest86 to scan your RAM chips for errors. If you get errors, contact your supplier for an RMA. If the RAM is error free for at least 2 passes, its time to move onto the next component to test. One pass can take a while, so try to run it overnight or something if you can.

An old system i had suffered from a similar problem to what you describe. The problem for me was a failing Power Supply. By disabling that automatic restart, you should get an error message if another part of the hardware is acting up. If you still get quick uncommanded restart's, i'd be looking at getting the PSU tested, or maybe swapping it out for another one for testing purposes. If you kept the old PSU, try putting it back in to see if you get problems.

Its not very helpful i know, but sometimes the only way to work out which components are causing problems is to swap them out one at a time until the problem stops, then go about getting getting the dodgy component tested or RMA'd.

There should be a few things for you to be going on with there though, hope it helps and doesn't just confuse things for you.

[ATAG_Septic]

Intermittent faults are horrible, I've spent hours proving a particular component's culpability by logical elimination only to prove it can't be a few hours later.

I'm no expert nowadays but it's an "it used to work and now it sometimes doesn't" fault. Which casts suspicion on any contemporary modifications. But, symptoms began a week after modification, which would seem to rule it out. It sounds like overheating or an intermittent connection. It's down to eliminating suspects I suppose.

Sorry it's not very specific old bean and there are much better computerists here but I'm happy to help any way I can.

Septic.

[Doc]

Power supply is getting soft.

[III./ZG76_Keller]

Thanks guys,

No Overclocking

Automatic restart is already turned off (I believe Doc suggested this when the issue first arose)

Did the windows Memory test with no errors detected, will try the one you have recommended.

Memory was purchased to match the original memory, same speed and manufacturer but slightly newer model. Old ram is on channel 1, new ram is on channel 2.

Power Supply is new, old one was doing the same thing. I did a power calculator and I should only be in the 600 watt range.

It's possible that it's a heat issue but nothing is overclocked and my Antec case has 4 fans in it that are running full blast.

I'm worried that this may be a short.

One additional thing to add, I shut my computer off at night when I go to bed and turn it back on when I get home from work. The last two days all the failures happened within the first 90 minutes of the computer being on and then stopped for the remainder of the night. Also these failures happen at random times, from idle state to web browsing to playing a game.

[Doc]

PS new then next thing I would suspect is a power related component such as a cap on your motherboard that is losing voltage.

What typically techs do that are OEM techs is they have another HDD with a bare bones install of the OS and then use a utility like PC doctor and run a burn in test to stress it out for several ours to see if it will reproduce.

If it doesn't then PS is good. May be software related.

In other words you have the worse problem in my opinion. I hate these random issues that don't happen frequently enough to use process of elimination quickly.

It's torture and inhumane to flight sim persons to have it.

[ATAG_Colander]

In my experience..
1.- Power supply running at the upper limit. Get a PS with more Wattage.
2.- Mother board issues. A bad soldering in any component tends to connect when cold and disconnect (or barely connect) when hot.

On a similar note, years ago I had a server that did this but not when I opened to troubleshoot. First I thought that the extra ventilation was keeping it from crashing but after a few days of trial and error I finally found out that the side panel I was removing to work on it was doing a short on the mother board!!!!!

Colander.

[III./ZG76_Keller]

It appears that I MAY have found the issue and it seems to be related to the GPU. If I disconnect my monitor a slight jiggle of the DP cable can reproduce the problem, seems that there might be a short somewhere in my PCI-E connection.

Maybe it's time to replace my MOBO and processor.


:pouting:

[335th_GRAthos]

Then you only need to remove the GPU, clean the contacts with alcohol (the pharmaceutical one! :Grin: ) and put the card back in again. This should solve it I think.

Otherwise, it is memory and you need to run Memtest86+ http://www.memtest.org/

You did not mention how old your motherboard is. There is a crazy mistake on all old motherboards, all capacitators that were ever manufactured ten years ago, were faulty. Then some of them start leaking.
I experiencd this on one motherboard in the last 25years but, you may be that lucky.


~S~

[Wolf]

I had the same issue a while back with a Brand new MB. Turned out it was the new RAM. I got the two sticks replaced and never a dull moment since.


So if you still get the issue after cleaning your GPU or replacing the MB You may find it was the RAM all along.

Best of luck