Faulty case fan cause of system instability?

Do you have a question? Post it now! No Registration Necessary.  Now with pictures!

Threaded View

Recently, my computer started freezing and rebooting. I ran memtest and
found errors within a half hour. After testing various components, I
found that no more memtest errors occured after the removal of one of my
case fans, a 23cm, 4-LED fan made by Globe, model RL4B.

I ran memtest for at least several dozen complete passes and no further
errors were detected. About a month has passed since that time, and no
further unexpected system reboots have occured.

Have you ever heard of a faulty case fan causing memtest errors?

The fan's label states that it uses 0.35 amps on the 12 volt. I thought
that maybe the problem was that my PSU wasn't up to the task, but
I loaded each cores of my Core 2 Duo E7500 with burnMMX while stressing my
GeForce 9600GT with repeated timedemos of Nexuiz. I did that for 20 or
30 minutes and there weren't any errors or problems of any kind, so I
don't think the PSU's 12v is faulty.

Could the problem be some sort of electrical problem caused by the LEDs in the

Just a wild guess.

Re: Faulty case fan cause of system instability?

On Fri, 8 Jan 2010 23:57:32 +0000 (UTC), ShadowTek

Quoted text here. Click to load it

Was the fan "faulty" or are you only assuming it was faulty
since it seems to have caused memory errors?

It is possible (a certainty actually for the typical DC case
fan) for it to put electrical noise on the power rail it
runs from, but the rest of the system components should be
fairly immune to that as a system has quite a few other
sources of electrical noise as well.

I would more likely think that perhaps when you were in the
system, you touched the memory or the board and it shifted
just enough to make the memory slot contacts conduct better,
that there is a problem with them like manufacturing
intolerance, residue, or the memory wasn't fully seated...
or of course a memory module itself has a defect causing
intermittent faults.

Quoted text here. Click to load it

0.35A is a trivial load compared to the rest of the system
even when it's idle, though the fan likely draws less than
0.35A average, the labels tend to list peak startup current

Quoted text here. Click to load it

No, LEDs themselves will not cause a problem though a fancy
controller (which probably isn't present in this case) could
cause some noise.  Only way to be sure is connect the fan
again and see if the problem resurfaces, and if you want to
regain use of the fan you might also try plugging it in
elsewhere instead of same power connector.  For example if
it were plugged into the motherboard fan header try a PSU
connector w/adapter cable if needed.  If it were plugged
into the PSU connecter previously that would be very
unusual, noise from a fan up a  cable near
PSU/inductors/capacitors/back down another PSU cable would
have to be quite a lot of noise... in that situation I would
start to think there is some  defect in the fan but perhaps
a design defect rather than component failure since you
don't mention the fan failing to work.

One other thought -  Is it possible that due to the case fan
being removed the system now runs hotter and gains some
stability?  Some systems do run better when  warm, usually
due to  marginal capacitors with rising ESR as temperature

Re: Faulty case fan cause of system instability?

ShadowTek wrote:
Quoted text here. Click to load it

Fans generate electrical noise from their motors and DC-AC inverters
but not from their LEDs.  The fan current draw is neglibile for the
PSU and, BTW, is usually just half the labelled rating.  Maybe the
Globe fan doesn't have a filter capacitor across it to block the
noise, or the mobo doesn't have enough filtering at its fan connectors
(most I've looked at had none, just empty spaces where such capacitors
could be installed).  Have you tried plugging the CPU fan into the
connector for the case fan (if it's adjustable, test at full speed to
maximize current draw) or the case fan into the CPU fan's connector?
I've seen many mobos where the fan filter capacitors were missing, and
to see if that's the cause of your problem try jamming a 0.1uF ceramic
or 10uF, 25V (or higher voltage) electrolytic in the rear of the fan's
cable connector, between its black ground wire and its red +12V wire
(on 12V fans, red means +12V, not +5V as it does with computer PSUs).
Be sure to get the polarity of any electrolytic cap correct, meaning
the minus side (big stripe running down one side)goes to the black
ground.  If this eliminates memory errors, try to make the attachment
permanent by soldering the leads, but even if you don't, at least
cover it with heat shrink tubing or _good_ cloth or plastic tape so it
can't come loose or short something.

Re: Faulty case fan cause of system instability?

Quoted text here. Click to load it

This fan only has a 4-pin molex connector, so it always runs at full

I tested on one of the PSU's other connectors, but there were still
memtest errors on the other plug as well.

Quoted text here. Click to load it

I don't have any spare caps on-hand. I'd have to stop by the store

Site Timeline