Fans speed burst for a split second every 10min or so

[YanivF]

I have icue 5000x with 10 fans (6 connected to the h150i controller and 4 to XT). I noticed that every 10 min or so, the fans spin up for less than a second and then go back to normal quite mode. is that expected?

 

[c-attack]

Can you tell which controller’s fans are spiking?  “Quiet” on the XT does not mean the same thing as Quiet on the Commander Core that uses coolant temperature as the control variable. 

[YanivF]

Hi @c-attack, I'm not not sure as I can't tell visually. I just hear it spike but all the fans are set to 'quiet'. What temperature level of the coolant triggers the fans to speed up? the coolant temperature is <40c.

Is spike is expected? I would assume that once it reaches a trigger point there is some threshold before it goes back down if that is the reason

[YanivF]

re-writing the above to make it more clear - 

Hi @c-attack, I'm not not sure which fans spike as I can't tell visually. I hear it spike and but the time i look it's too late. All the fans are set to quiet mode. .

What temperature level of the coolant triggers the fans to speed up? I'm not running any large load yet and I didn't see the coolant temperature go to 40c yet.

Is spike expected behavior? I would assume that once it reaches a trigger point there is some threshold before it goes back down if that is the reason. Do you know if there is away to log when fans change speed?

[c-attack]

You might be able to catch it after the fact by looking at the graph in the cooling tab for the Commander Core and XT.  I suspect only the XT is spiking and the response below is what I generally suggest for most people regardless.

 

The Commander Core with pump will use the coolant temperature (H150i Elite Temp) as the control source for the fans.  This is good as it makes an appropriate assessment of environmental temperature.  Coolant temp is a slow to change variable and so fans should do the same --- all you need from case fans.  However, the other Commanders (Pro and XT) default to using CPU temp as the control variable and sometimes the water cooling curve as the control mechanism.  Those two don't go together well and CPU temp is not a good control variable for anything except a small CPU air cooler.  

 

In your Commander XT cooling tab, click + to create a new cooling curve.  A graph will appear below.  Go to the shape tools in the lower right corner.  The 2nd from the left is the "Quiet AIO" curve your Commander Core is using.  Select that.  Then go to the Sensor and change it to "H150i Elite Temp".  This makes it identical to the other controller.  Now change the fans above so they use the new 'Custom 1' control curve.  I also suggest people do the same on the Commander Core.  It doesn't change the curve, but it turns the preset into a visible and editable option.  This lets you see the control points and understand behavior vs guessing if it's doing what it should.  Same procedure as above.

 

If the Commander Core is the one spiking (or both controllers), then we need to look for any other RGB or hardware monitoring programs you might be running.  Common examples are HWiNFO or AIDA.  These can become entangled with CUE as they try to access the hardware at the same time.  

 

This is an older picture from something else, but it identifies the changes.  Note you will be selecting H150i Temp on the sensor.  The one below is showing the water temp sensor from my custom loop on a Commander Pro.  

Edited March 15, 2022 by c-attack

[YanivF]

Thank you very much @c-attack. I'll follow the steps you suggested. Question: I can also add the temp sensors to the XT, which I didn't would you recommend adding them and if yes, where would you place them

[c-attack]

You most certainly can add a temp sensor to the XT to do the same job.  Technically this is a superior choice since it is the native value for that controller and works at all times.  If you quit CUE, the Com XT cannot fetch the data from the Com Core.  The water temp solution is quick way to stop the fans from spiking with no hardware changes, but temp probe is better.

 

You can do a few things with it.  First is stick one somewhere near the rear exhaust, in or out of the case.  Case rear exhaust temp is heavily influenced by the GPU and the best choice for most.  Exact placement is not important, but try set and try not to move it.  Shifts can alter the end result temp and that is a relative value.  The trickier part to this is finding your "normal" range.  While it's probably between 25-45C, that's a wide field and heavily influenced by hardware and case layout.  You also can throw one up near the front intake somewhere, just so you can see a intake vs exhaust temp comparison.

 

Second option is to put the temp probe on the exhaust side of the H150i radiator somewhere.  Exhaust air temp from that will be approximately equal to coolant temp.  This lets you use the same curve across all fans and required minimal guessing and tinkering.  Any spot on the backside should do.  Option 1 may be better for those with heavy GPU use.  This option may be better for those with heavy CPU renders, encoding, or other professional workloads.  

[YanivF]

Thank again @c-attack

Got home and follow your instructions - XT fans were in quiet, while core fans were in balanced except for the pump that was set to extreme. I changed all the fans to the custom curve / quiet but wasn't sure if I should do the same with the pump.

 

 

[c-attack]

You can do the same with the Commander Core unit and I encourage people to do so. It’s the same curve, but now you can see the data points and know what’s happening. You can use balanced, quiet, or a combination. Neither should cause fan spikes when using H150i temp as the control variable so the curve aggressiveness is not the source.

 

One other thing I will mention is to be aware of PSUs that operate in passive mode (no fan) then kick on when certain temperature or power level thresholds are met. I was testing a new gpu last night and it alone can pull the 400W required to make my 1000W PSU spin up. The PSU would kick to life each time I ran the benchmark and it’s 12v start up whir can be difficult to distinguish between other fans unless you know what it is. The Commanders should leave a visible fan spike in the cooling graph if they are doing this.