Question During idle, one of my front case fans spikes up and then cools down every 10 seconds or so

[Noobpunk]

https://imgur.com/a/EHzRa4c

View: https://imgur.com/a/EHzRa4c


Hey guys,

So my fairly new PC recently has started behaving a little unusual. The front panel of the PC has 3 fans and one of them every 10 seconds or so speeds up, up to 3000rpm and then down to the levels that the other 2 fans are spinning at.

Does anyone with some experience in this know what the issue could be?

I must emphasise that the temperatures throughout my PC are all in the good and stable digits.

 

[JohnBonhamsGhost]

i would imagine either that fan, the header it is connected to, or any hub it may be connected to is malfunctioning.
i've never seen Corsair fans, especially stock installed case fans, reaching 3000RPM.

if it isn't already, try connecting this fan directly to a motherboard SYS_FAN or CHA_FAN header with a manual fan curve profile in place and see if it still fluctuates and shows incorrect stats.

 

[Noobpunk]

i would imagine either that fan, the header it is connected to, or any hub it may be connected to is malfunctioning.
i've never seen Corsair fans, especially stock installed case fans, reaching 3000RPM.

if it isn't already, try connecting this fan directly to a motherboard SYS_FAN or CHA_FAN header with a manual fan curve profile in place and see if it still fluctuates and shows incorrect stats.

Thing is, I think it is just the fan that does this because the other fans don’t behave like this one.

These fans are connected to the ICUE Commander Core, I will try tomorrow and see if I can adjust this at all. Because it does this when the fan profile is set to quiet.

Could this relate to changing my power mode to power saving at all?

 

[JohnBonhamsGhost]

Could this relate to changing my power mode to power saving at all?

shouldn't affect any fan RPMs at all.
and as long as they are connected to a controller that is in turn controlled by the iCUE software, it is up to iCUE what profile the fans use.
Windows wouldn't have anything to do with it at this point.

 

[Noobpunk]

shouldn't affect any fan RPMs at all.
and as long as they are connected to a controller that is in turn controlled by the iCUE software, it is up to iCUE what profile the fans use.
Windows wouldn't have anything to do with it at this point.

As the fans are connected to the ICUE, would you think it is more software than an hardware issue?
Mind you there was an ICUE update a few days ago..

I also have the noctua fans plugged into the CHA_FAN headers and they’re not behaving this way.

 

[Paperdoc]

Here's a possibility. When this is happening, see if you can watch the fans carefully. I suspect that the one that is behaving oddly actually is STOPPING, then re-starting.

All fan headers monitor their fan's speed signal for fan failure - that is, no speed signal. MANY will respond to that by sending out to the fan a signal to go to full speed so it CAN re-start. If that does not work, you get a fan failure message. But if it does work, the fan then is told to return to its current slower speed setting.

You are running all the fans on a Quiet setting - that is, slow all the time. It may be that that particular fan is a little stiffer, so it stalls more easily that the others, and you happen to be running it right at the stall limit. So frequently that fan stalls, gets re-started successfully then slowed down, and repeat. IF you verify that it is stopping and re-starting, use iCUE to re-specify a slightly higher minimum speed it will use for your quiet setting.

 

[Noobpunk]

Here's a possibility. When this is happening, see if you can watch the fans carefully. I suspect that the one that is behaving oddly actually is STOPPING, then re-starting.

All fan headers monitor their fan's speed signal for fan failure - that is, no speed signal. MANY will respond to that by sending out to the fan a signal to go to full speed so it CAN re-start. If that does not work, you get a fan failure message. But if it does work, the fan then is told to return to its current slower speed setting.

You are running all the fans on a Quiet setting - that is, slow all the time. It may be that that particular fan is a little stiffer, so it stalls more easily that the others, and you happen to be running it right at the stall limit. So frequently that fan stalls, gets re-started successfully then slowed down, and repeat. IF you verify that it is stopping and re-starting, use iCUE to re-specify a slightly higher minimum speed it will use for your quiet setting.

One thing I had noticed was when I changed the fans profile to balanced, it wouldn’t spike up to 3000 for any of the fans and it was stable; this being done in ICUE.

However if I close ICUE from the task manager, I also did notice that I would see occasionally the fan speed going down to 0 and then spiking back up again.

This is all being seen through hwinfo by the way. I have read that the monitoring softwares can interfere with ICUE so I pray it’s software rather than hardware..

 

[JohnBonhamsGhost]

what fan make & model are they?

 

[Noobpunk]

what fan make & model are they?

They are three LL120 RGB fans.

I think it may be a software collision. Because now that I have ICUE open, the fans are no longer behaving abnormally in quiet mode.

One thing I did notice though when I had ICUE closed but HWinfo open was, Fan 1 would go up to 3000rpm for a split second and then jump back into the regular speed whereas another fan would simply go to 0 for a split second and then return to its normal speed.

Only when I open HWinfo the fans behave abnormally. However the moment I open ICUE and open the cooling tab for the fans, the fans return to the quiet mode. Now closing both ICUE and HWinfo, the fans are so far behaving as they should.

 

[JohnBonhamsGhost]

LL120 RGB fans

as far as i've ever seen these only go up to 1500RPM max so there's definitely an issue with any software reporting up to 3000.

 

[Paperdoc]

I agree with JohnBonhamsGhost above. And that leads to a related question. I had assumed when you said the one fan in question jumps to 3000 rpm and then back down, you confirm that the fan really DOES jump to full speed and noise (whatever that is). That is, you are not merely reporting speed displayed without reference to reality. So, DOES that fan jump up and down in speed on some occasions?

Even if it does, WHAT that high speed is, is in question, as JBG said. It is not 3000 RPM. Many do not realize that several third-party utilities that tell you stuff like fan speed and mobo temperatures are NOT immediately correct, and must be calibrated before you can rely on their info. It appears this is one such case when you use hwinfo.

 

[Noobpunk]

I agree with JohnBonhamsGhost above. And that leads to a related question. I had assumed when you said the one fan in question jumps to 3000 rpm and then back down, you confirm that the fan really DOES jump to full speed and noise (whatever that is). That is, you are not merely reporting speed displayed without reference to reality. So, DOES that fan jump up and down in speed on some occasions?

Even if it does, WHAT that high speed is, is in question, as JBG said. It is not 3000 RPM. Many do not realize that several third-party utilities that tell you stuff like fan speed and mobo temperatures are NOT immediately correct, and must be calibrated before you can rely on their info. It appears this is one such case when you use hwinfo.

Just to confirm with you, that fan actually does ramp up. I hear the difference in the fan physically.

What I did last night was potentially a fix.

After removing the Corsair link from HWInfo, then fans no longer behaved abnormally. Tested this with ICUE closed/opened + HWinfo closed/opened.

Here is a thread similar to my case but on HWInfos forum; https://www.hwinfo.com/forum/threads/hwinfo-and-corsair-commander-pro.5746/

I will need to do more testing tonight however so far so good.

 

[Paperdoc]

I see. I would guess the problem is that both iCUE and HWInfo are trying to take some control of mobo resources like fan headers, and conflicting with each other. The thread you linked to suggests there is a setting in HWInfo for Safety that solves this issue, perhaps by halting certain HWInfo actions that conflict.