Why did resetting CMOS fix my computer?

Toggle sidebar Toggle sidebar
W

westphalia

Honorable
Jan 27, 2014
3
0
10,510
Hello, wondering if someone can help me understand what went wrong with my computer, although it is now fixed.

The other day while playing a game (and only while playing this game) the fans on my computer would suddenly rev up drastically and the computer would reboot. I tried it again in an attempt to diagnose the problem and on the third or fourth time the computer began not to boot at all. The HDD LED would blink on and off but no boot.

After researching a bit I thought maybe dust or debris was the problem, as I have two cats. I cleaned the inside of the case with compressed air but there wasn't really much dust to begin with and that didn't fix the problem. Removed the GPU, still didn't work. Removed the heatsink, reapplied the thermal paste, put it back on, still no dice. At this point I noticed the computer's fans would begin to spin for about one second and then stop, then repeat this over and over, even without me pressing the power button (i.e. only the back switch was on). Next I tried unplugging all power cables from the motherboard and resetting them, and also reset the CMOS. Now the computer works fine. Ideally I would have done these two steps separately so as to better isolate the problem, but alas. I'm glad it was relatively easy to fix, but I would like to understand why this worked for future reference. Any ideas?
 
Solution
T

There probably is but it would involve some test equipment because it would need to be tested under load so it would probably be easier to just replace the power supply. On the other hand, I wouldn't do anything if it is working OK now. Maybe look deeper if the problem crops up again.

You might want to check the power coming from the wall outlet. Under or over or fluxuating voltage can be hard on power supplies. But...
Sort by date Sort by votes
T

thx1138v2

Splendid
Jun 18, 2011
2,172
97
22,040
Everything at this point is just a guess but I'm amazed at how many times just unplugging everything and plugging it back in fixes problems. I'm sure engineers may have more technical explanations but what works, well, works. Things heat up and cool off and expand and contract and, as you mentioned, dust settles into connections. With the size of the traces in today's circuit boards and connectors (more connections in smaller spaces) it doesn't take much to semi-bridge a couple of conductors just sufficiently to make make it malfunction. There is also vibration as long as disks and fans are spinning.

It may have been the BIOS reset also. Very few people think about it but we are constantly being bombarded with charged particles of different varieties. The most ghost-like is the neutrino and they very rarely interact with matter. But very rarely is not never. So it can happen. So I'm really not joking when I say sun spots can be the cause. Here's an interesting perspective:
http://www.swpc.noaa.gov/wsa-enlil/
 
Upvote 0 Downvote
W

westphalia

Honorable
Jan 27, 2014
3
0
10,510
@ihog - nope, the computer is about six months old and I've never overclocked anything. I don't think I've touched the BIOS settings since initially setting them up.

@thx1138v2 - absolutely fascinating. I guess it's possible the source of my problem will remain a mystery. One thing I couldn't understand was why the fans would rev up like that when the computer did not appear to be overheating and has never run very hot even under load.
 
Upvote 0 Downvote
T

thx1138v2

Splendid
Jun 18, 2011
2,172
97
22,040

You didn't give the specs on the machine but the temperature sensors report through either traces on the mobo or wires from a three wire fan connector and they report to the BIOS initially. So it could still be either - something across one of the reporting connections which would make the reading appear higher than it actually was or something in the BIOS that somehow got diddled to tell it to shut down. For example a setting that tells it the temperature at which to shut down got set to zero, in which case any temperature reading, even the ambient temperature, would shut it down.

You didn't say whether or not you got to the BIOS screen and beep before it shut down but the way you described it I don't think you did. You might want to pull the PSU and give it a good blowing out with compressed air while holding it upside down. Another possibility is the PSU might be going bad if it is either cheap or getting old..
 
Upvote 0 Downvote
W

westphalia

Honorable
Jan 27, 2014
3
0
10,510
That sounds likely to me. I wasn't getting to the BIOS screen at all, no beep, no nothing. The PSU was my next thought on what might be broken. It's only six months old and from a reputable manufacturer (Corsair) but they do break. Is there any way to test a PSU short of having it fail / testing it in another machine?
 
Upvote 0 Downvote
T

thx1138v2

Splendid
Jun 18, 2011
2,172
97
22,040

There probably is but it would involve some test equipment because it would need to be tested under load so it would probably be easier to just replace the power supply. On the other hand, I wouldn't do anything if it is working OK now. Maybe look deeper if the problem crops up again.

You might want to check the power coming from the wall outlet. Under or over or fluxuating voltage can be hard on power supplies. But again you need at least a volt meter for that. A UPS that has conditioning circuity can help some of that if it is a problem but the cheap UPS's won't have any conditioning circuitry.
 
Upvote 0 Downvote
Solution
You must log in or register to reply here.

TRENDING THREADS

Latest posts

Moderators online

Share this page

COMPANY
RESOURCES
FOLLOW
Top Bottom