Sign in with
Sign up | Sign in
Your question

Unstable fan speed

Last response: in Systems
Share
April 5, 2012 11:03:28 AM

Hi guys,

I have lately added a new PSU 80+ bronze to my PC, and since then I am having a little issue with my case LCD panel resulting in unstable fans speed and the bottoms not functioning properly.

(The connectors of the panel are very sensitive)

The issue is that CPU Core i5-2300 fan speed varies from 1200 to 1300rpm temp from 28c ro 35.

GPU 800 to 900rpm 27c to 33c.

I am just asking if the unstable speed may damage the fans then the CPU and GPU, knowing that the PC works for more than 10 hours a day or is it just a trivial problem and it is not very urgent, especially that I know that all the issue comes from the LCD panel.

Thanks for your help.

More about : unstable fan speed

a c 78 B Homebuilt system
April 5, 2012 2:15:42 PM

These fans aren't going to get damaged with +/- 100 RPMs. The problem is indeed trivial.

That being said, which PSU was it that you just inserted into the computer?
April 5, 2012 2:21:31 PM

900 rpm \ 60 secs is 15 revs per second

I also wouldn't worry about plus or minus 100... The sensor is just probably not too sensitive and may vary 1 rev per second
Related resources
a b B Homebuilt system
April 5, 2012 3:41:50 PM

You do realize that the whole point of automatic fan speed control is that the speed SHOULD change when the device temperature changes? When your CPU (or GPU) is working harder and producing more heat its temperature rises, and this SHOULD cause the fan speed to increase to limit the temperature rise. When the workload decreases, the fan should slow down as the temp drops.
April 5, 2012 7:49:11 PM

Raiddinn said:


That being said, which PSU was it that you just inserted into the computer?


OCZ 650w zs series.
a c 78 B Homebuilt system
April 5, 2012 8:15:42 PM

Unstable PSU power delivery causes some crazy things to happen with fans, but with that PSU you probably don't have to worry too much about that.

Just wanted to verify that didn't have to do with it, and it doesn't sound like it does.
April 5, 2012 8:26:55 PM

Paperdoc said:
You do realize that the whole point of automatic fan speed control is that the speed SHOULD change when the device temperature changes? When your CPU (or GPU) is working harder and producing more heat its temperature rises, and this SHOULD cause the fan speed to increase to limit the temperature rise. When the workload decreases, the fan should slow down as the temp drops.


Consequently, the automatic fan speed works in my computer since at start-up it reaches 2000 Rpm than it falls to 1200-1300rpm.
==>But after it keeps changing every 5 sec(I have counted it).
a c 78 B Homebuilt system
April 5, 2012 9:10:41 PM

IIRC, CPU power saving stuff doesn't work inside the BIOS. That means while you are in there, the CPU is at 100% load the entire time until it gets into Windows where the fan controllers and stuff are loaded.

That would mean the fans have to work a lot harder to keep the CPU cool before the OS is loaded.

That should account for what you are seeing.
a b B Homebuilt system
April 6, 2012 4:16:50 PM

Fan motors need almost full voltage (12 VDC) for starting up. Supplying them too low a voltage will not start them. So, most automatic control systems supply the full voltage first, THEN lower it to what is required. This works out well for the normal Start-up process for the entire system. When the BIOS first begins its initiation, it sets fan power to full voltage, then goes about the rest of its many initiation and POST processes. After those are done, it can get around to actually starting to use the fan controller routines it has, and these will begin doing control. That's when your fans drop to a lower speed, because at that point the temperatures of the CPU and mobo are still low - after all, they've just turned on.

AFTER all that has happened, you seem to say that the fan speeds change up and down by small amounts (maybe 100 rpm) every five seconds or so. I do NOT think that is a problem. It is possible that the actual supply voltage from the PSU is changing, and you could examine that with a voltage display. However, it think it much more likely that the changes you observe are simply the fan controller doing it normal job. Remember that this is a digital device, and hence all its actions are in small discrete steps, and not a continuous range. How small is "small"? Small enough to give control as good as needed, but probably big enough to see if you watch. For example, MAYBE yours uses 32 possible voltage outputs, which would correspond roughly to steps of 100 rpm for a fan that tops out at 3000 rpm. (That is NOT exactly how your works - I just picked some numbers to illustrate the point.)

Also bear in mind that a controller works on DEVIATION of measured process value (temperature here) from Setpoint, and it has a Deadband. That is, within a small Deviation from Setpoint it makes NO change; when the Deviation gets over some small value, it does change to the next possible output state. It sounds like your system's temperatures drift up and down so that, roughly every 5 seconds, the Deviation is enough to cause fan speed to change. As long as the speed is going up and down and the temperature is stable, the fans and their controller are behaving properly.
January 17, 2014 5:52:58 AM

I will bump this.

I have similar problem..on certain level of active cooling fan acts really annoying, it goes up and down (by 100 rpm) in 6 second loops. it happens at 1200-1300. (http://imgur.com/bxUM7bx)
You can actually hear it's changing and it drives me crazy.

On higher levels 1400 etc, system fan rpm is stable (http://imgur.com/q3TObW7)

Thank you for your advices
!