What to do about a vibrating machine (never *your* vibrating machine)

sixblueboxes · 2013-03-09, 15:42

I have a terrible problem with one of my machines vibrating, quite loudly, when I'm running more than one core. It's a AMD Phenom 9650 Quad-Core machine that I built myself. The vibration is very loud and annoying. I can't seem to find the reason for the vibration; it stops when I grab the chassis, or turn off mprime. I've had problems with vibrating machines in the past, but it was usually with something that was temporary or I could easily turn off, and I just ignored it. This is persistent; I can hear it at night especially when the house is super quiet.

I have six machines in my GIMPS arsenal, and this is the only one doing LL tests. Every other machine makes more noise because the fan is running higher, but they don't vibrate like this one.

Any suggestions?

TimSorbet · 2013-03-09, 16:45

Are the motherboard and power supply securely mounted to the case? If you put something (e.g. a rug) under the part(s) of the case that sit on the floor, does the noise stop?

You might've guessed this, but a likely candidate for the source of the vibration seem (to me) to be a fan (CPU, power supply, or case, in roughly descending order). You might want to reseat/resecure all of these, if possible.

swl551 · 2013-03-09, 17:02

Agree. Fan. Rpm increases with load. Maybe loose p.s. mount.

LaurV · 2013-03-09, 17:19

Vibration can only come from a pump, in a liquid cooled system, or from an unbalanced fan (broken/bent wing/blade, etc). This gets worse in time and cause trouble for other components too, beside of losing of efficiency. I would recommend to identify and change the intruder as soon as possible. This is easy to do, you can touch all the fans with your fingernail, carefully not to let your fingertips inside, try to stop a fan using a brush or an eraser (rubber, don't use hard objects, they will break the blades of the fans). If the side where the air come out is accessible, you can use your finger too (carefully! And don't put your fingers on the blades on the side where the air goes in, those blades are very sharp when they rotate. On the other side of the fan, where the air goes off, the blades "run away" and can't cut your finger). You can identify immediately which part makes vibrations. Try re-arranging it or changing it.

ixfd64 · 2013-03-09, 18:20

I LOL'd at the title.

Aramis Wyler · 2013-03-09, 18:35

Agreed - LOL at title and cpu fan. Running more than one core will crank up the heat, and then the fan. Though the first thing I would do would be to see if the screws are tight on the heatsink, then blow out the fan with air, then check the fan screws.

If that didn't work I'd replace the fan on the heatsink, and make sure I used rubber mounts on it.

Though for what it's worth, I chucked all of my internal fans and went to water for quiet some years ago.

Batalov · 2013-03-09, 21:01

[off-topic]
Hey, finally not a Life-of-Brian but a Fight Club inspired title!
[/off-topic]

sixblueboxes · 2013-03-10, 14:12

Thanks for your thoughts everyone. I'll open up the machine and see what I can see. Maybe while I'm in there I'll put the DVD drive back in.

chris2be8 · 2013-03-10, 16:44

It might be that the resonant frequency of one of the panels matches the speed the fan is turning. Try touching each panel in turn, if the sound stops (or gets quieter) when you touch one panel that's the one thats resonating. You might be able to fix if more firmly, or stick a weight or damper on it to change the resonant frequency.

Checking for unbalanced fans is also a good idea though.

Chris

sixblueboxes · 2013-03-14, 12:09

I haven't had a chance to open up the machine yet, but I have noticed that if I touch the power supply, the sound gets quieter, but if I touch any other part it does not. Also, if I run four cores instead of three, it's not nearly as loud.

Still getting plenty warm in there though:

Code:

CPU FAN Speed:    5273 RPM  (min =  600 RPM)
CHASSIS FAN Speed:1101 RPM  (min =  600 RPM)
CPU Temperature:  +141.8Â°F  (high = +140.0Â°F, crit = +203.0Â°F)
MB Temperature:   +118.4Â°F  (high = +113.0Â°F, crit = +167.0Â°F)

swl551 · 2013-03-14, 13:09

Quote:

Originally Posted by sixblueboxes

I haven't had a chance to open up the machine yet, but I have noticed that if I touch the power supply, the sound gets quieter, but if I touch any other part it does not. Also, if I run four cores instead of three, it's not nearly as loud.

Still getting plenty warm in there though:

Code:

CPU FAN Speed:    5273 RPM  (min =  600 RPM)
CHASSIS FAN Speed:1101 RPM  (min =  600 RPM)
CPU Temperature:  +141.8Â°F  (high = +140.0Â°F, crit = +203.0Â°F)
MB Temperature:   +118.4Â°F  (high = +113.0Â°F, crit = +167.0Â°F)

141 F is cool for an active CPU, but your case temp is high for such a low CPU temp. Unless you running GPU factoring in the same case.

2013-03-09, 15:42	#1
sixblueboxes Feb 2013 Charlotte, North Carolina 31 Posts	*What to do about a vibrating machine (never your* vibrating machine)** I have a terrible problem with one of my machines vibrating, quite loudly, when I'm running more than one core. It's a AMD Phenom 9650 Quad-Core machine that I built myself. The vibration is very loud and annoying. I can't seem to find the reason for the vibration; it stops when I grab the chassis, or turn off mprime. I've had problems with vibrating machines in the past, but it was usually with something that was temporary or I could easily turn off, and I just ignored it. This is persistent; I can hear it at night especially when the house is super quiet. I have six machines in my GIMPS arsenal, and this is the only one doing LL tests. Every other machine makes more noise because the fan is running higher, but they don't vibrate like this one. Any suggestions?

2013-03-09, 16:45	#2
TimSorbet Account Deleted "Tim Sorbera" Aug 2006 San Antonio, TX USA 11×389 Posts	Are the motherboard and power supply securely mounted to the case? If you put something (e.g. a rug) under the part(s) of the case that sit on the floor, does the noise stop? You might've guessed this, but a likely candidate for the source of the vibration seem (to me) to be a fan (CPU, power supply, or case, in roughly descending order). You might want to reseat/resecure all of these, if possible. Last fiddled with by TimSorbet on 2013-03-09 at 16:46

2013-03-09, 17:19	#4
LaurV Romulan Interpreter "name field" Jun 2011 Thailand 2×11×467 Posts	Vibration can only come from a pump, in a liquid cooled system, or from an unbalanced fan (broken/bent wing/blade, etc). This gets worse in time and cause trouble for other components too, beside of losing of efficiency. I would recommend to identify and change the intruder as soon as possible. This is easy to do, you can touch all the fans with your fingernail, carefully not to let your fingertips inside, try to stop a fan using a brush or an eraser (rubber, don't use hard objects, they will break the blades of the fans). If the side where the air come out is accessible, you can use your finger too (carefully! And don't put your fingers on the blades on the side where the air goes in, those blades are very sharp when they rotate. On the other side of the fan, where the air goes off, the blades "run away" and can't cut your finger). You can identify immediately which part makes vibrations. Try re-arranging it or changing it. Last fiddled with by LaurV on 2013-03-09 at 17:20

2013-03-14, 12:09	#10
sixblueboxes Feb 2013 Charlotte, North Carolina 37₈ Posts	I haven't had a chance to open up the machine yet, but I have noticed that if I touch the power supply, the sound gets quieter, but if I touch any other part it does not. Also, if I run four cores instead of three, it's not nearly as loud. Still getting plenty warm in there though: Code: CPU FAN Speed: 5273 RPM (min = 600 RPM) CHASSIS FAN Speed:1101 RPM (min = 600 RPM) CPU Temperature: +141.8Â°F (high = +140.0Â°F, crit = +203.0Â°F) MB Temperature: +118.4Â°F (high = +113.0Â°F, crit = +167.0Â°F)

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
2.5Pflops CPU + GPU machine	xilman	Science & Technology	3	2010-10-29 00:28
Embedded machine	millsie	Software	9	2010-02-23 22:45
offline machine	junky	NFSNET Discussion	8	2004-03-25 08:32
new machine	junky	NFSNET Discussion	11	2004-02-21 03:39
Two programs on same machine?	Unregistered	Software	14	2004-02-15 16:36

2013-03-09, 17:02	#3
swl551 Aug 2012 New Hampshire 2³·101 Posts	Agree. Fan. Rpm increases with load. Maybe loose p.s. mount.

2013-03-09, 18:20	#5
ixfd64 Bemusing Prompter "Danny" Dec 2002 California 9BE₁₆ Posts	I LOL'd at the title.

2013-03-09, 18:35	#6
Aramis Wyler "Bill Staffen" Jan 2013 Pittsburgh, PA, USA 424₁₀ Posts	Agreed - LOL at title and cpu fan. Running more than one core will crank up the heat, and then the fan. Though the first thing I would do would be to see if the screws are tight on the heatsink, then blow out the fan with air, then check the fan screws. If that didn't work I'd replace the fan on the heatsink, and make sure I used rubber mounts on it. Though for what it's worth, I chucked all of my internal fans and went to water for quiet some years ago.

2013-03-09, 21:01	#7
Batalov "Serge" Mar 2008 Phi(4,2^7658614+1)/2 3²·1,117 Posts	[off-topic] Hey, finally not a Life-of-Brian but a Fight Club inspired title! [/off-topic]

2013-03-10, 14:12	#8
sixblueboxes Feb 2013 Charlotte, North Carolina 31 Posts	Thanks for your thoughts everyone. I'll open up the machine and see what I can see. Maybe while I'm in there I'll put the DVD drive back in.

2013-03-10, 16:44	#9
chris2be8 Sep 2009 100101111001₂ Posts	It might be that the resonant frequency of one of the panels matches the speed the fan is turning. Try touching each panel in turn, if the sound stops (or gets quieter) when you touch one panel that's the one thats resonating. You might be able to fix if more firmly, or stick a weight or damper on it to change the resonant frequency. Checking for unbalanced fans is also a good idea though. Chris