Jump to content
Corsair Community

QL120 Fan Speeds Lower Using Commander Core XT vs Commander Pro?

Recommended Posts

I've recently been doing some trial-and-error testing on my fan controllers, and I've noticed that my QL120 fans spin at a maximum in the 900rpm range when set on Extreme in iCUE using Commander Core XT as the PWM controller.  I changed configuration to using Commander Pro as the PWM controller and I noticed the same fans spin at around 1700rpm when using the same Extreme setting through iCUE.  This is almost twice the speed!  I was happy to see the improvement as my CPU temperature at rest went down from around 42C to 35C...I use twin front radiators with push-pull in a custom cooling loop.  The increased fan speed cooled my coolant down significantly as seen by monitoring my pump/reservoir temp, and the obvious drop in CPU temp at rest.  I've also been having some RGB issues with the Commander Core XT, which was a second reason I decided to go back to my old Commander Pro controller (and a hub for RGB).

Does the Commander Core XT limit fan speeds to levels less than what the Commander Pro allows in the pre-configured speed options?  I haven't tested yet, but does the Commander Core XT allow manual settings as high as the Commander Pro pre-configured settings?

I should note that I use a PWM hub for the fans connected to the Commander Pro, but I had my QL120 fans' PWM cables connected directly to the Commander Core XT ports.  I doubt this makes any difference, but just sayin'...

Link to comment
Share on other sites

Your observations are correct, but unfortunately the underlying reason is a bit unintuitive.  Corsair chooses the hide the control points (Temp vs RPM) on their presets.  Furthermore, Quiet, Balanced, and Extreme can have different values across different devices.  Your Commander XT uses CPU temp as the control variable with some small level of fan hysteresis to keep the fans from aggressively ramping up and down as your CPU naturally loads and unloads each fraction of a second.  The Commander Pro also uses CPU temp as it's control variable on the presets with no fan delays, except the temperature points are ranged for water cooling and run in the 20-40C range.  This effectively makes all Commander Pro presets run at 1400-1700 RPM at all times, unless you can get your room temp way down to keep your idle CPU temp around 20-25C.  This wasn't really intended and I am not sure why it didn't get corrected all these years later.  It's probably a hardware limitation.


However, the advice is the same as it was back then.  Everyone using a Commander Pro should make their own custom curve and the same for the Commander XT.  CPU temp is a terrible control variable and it has no connection to your internal case temperature.  You have a couple of other options.

1) Use the Commander Pro temp sensor.  In the box were a bunch of wires with small orange plastic tip.  These 10K thermistor wires can be bought from most anywhere and are universal 2 prong connectors on the Commander end.  This measures air temperature.  Best place to stick is somewhere near the rear exhaust.  This can be in or outside of the case if you prefer so you don't have to look at it.  All you need here is a relative value for control.  You'll need to learn the normal range, but once you do it holds with the usual season room temp changes.  Set quiet fan speed for your normal doing stuff on the desktop temps.  Set a tolerable medium speed for whatever temp is gets up to during extended GPU loads like gaming or renders.  That will the highpoint for most users. The Commander XT also has these sensor registers, but may not have the 10K wires in the box.  Besides providing a steady control variable with a direct link to case temperature, the temp sensor is the native control variable for the controller.  It works at all times whether CUE is running or not, including boot, BIOS, and shutdown.  No more fan bursts as your load up or quit and protection from unnoticed app crashes.  This also includes custom water cooling temp probes.


2) If you have a Corsair AIO, you can use coolant temp.  In most installations this also does a good job measuring case internal ambient temp and most of its values from internal heat rather than CPU load.  Either way, both those sources trigger fan speed changes.  The only catch is CUE must be running for the Commander Pro or XT to fetch the data from the other controller that provides the coolant temp.


3) If none of the above work, your best best is to use GPU temp or "motherboard temp".  The later is preferable because it is a steady variable, unlike GPU temp that can shift 20-30C as it loads and unloads naturally during a game.  You don't really want the case fans to react to that.  The trick with motherboard temp is identifying and finding it.  CUE does not do a great job with some motherboard controllers and often provides duplicates or impossible values.  You may want to load up another monitoring program to identify the motherboard temp, then match it to Temp #X in the CUE.  You then can assign Temp #X as the control variable in your custom curve.  Motherboard temp, AIO coolant temp, and any air temp measure typically all have the same range and usually run in the same 25-45C range.  

Link to comment
Share on other sites

This answer is exactly what I was hoping to see...  You have provided detail and perspective that I find very interesting and allows me to configure my cooling to a much better level.  Thank you!

My temperature control goals are two-phase.  First and foremost, CPU cooling via radiators/coolant.  But not as obvious is the need to cool my Gen. 5 SSD that creates a lot of heat rapidly.  Because the SSD and its heat sink are located below my graphics card in the lower part of my case where the fans' air circulation is not as great (blocked by the card), I want to maintain positive air pressure inside my case to dissipate the case's air/heat which is done via fan speed control.  Then, of course, I want to control the push/pull fans for my radiators to effectively cool the CPU.  So, my cooling objectives revolve around case ambient air temperature and coolant temperature.

I have two rear fans as intake.  I have eight top fans as intake.  And I have 16 front fans in push/pull configuration on either side of two 480mm radiators, configured as exhaust.  My goal is to maintain higher fan speeds at the rear and top intakes than the front exhaust to maintain positive case air pressure.

I have one fan controller for the rear and top, and one fan controller for the front fans.  Therefore, I can use a temp sensor, as you suggested, for a relative value for control of case air temp, and I can use my Corsair XD5 Pump/Reservoir Combo's coolant temp probe to use coolant temp as a relative value for control of the radiator fans that cool the CPU.  I'll probably put the air temp probe somewhere below my graphics card near or in front of the SSD heat sink.  The case air temperature is higher here.

The air temp probe will be connected to the controller for the rear and top fans.  The coolant temp probe (from the pump/reservoir combo) will be connected to the controller for the radiator fans.  If temp sensors are the native control values for each controller, then the two controllers should be able to work independently to control fan speeds in the respective areas of the case via the two respective probes.

It will be a bit tricky to get the values correct to maintain positive air pressure, but also effective CPU cooling.  But with the volume of coolant (1200-1300ml) used in my loop, and 16 fans to cool the radiators, I think this set up will be effective.  I can also set the pump rpm higher or lower to cause the coolant to circulate through the radiators faster if lower radiator fan speeds are used.  Very interesting!

Thanks so much for the feedback, and I hope any other readers may gain some perspective here regarding how they configure their system for both case air and coolant temperature control.

Link to comment
Share on other sites

This case seems suspiciously like a 1000D.  That makes the CPU cooling fairly easy.  If it's just the CPU alone on the 2x480, you really don't have to do anything.  Even at nominal speeds 2x480 can dissipate a 300W load with minimal impact.  If you run CPU renders, encoding, or other long runs like, you can set a comfortable fixed speed before you start and be done.  If it's CPU + GPU (either combined or individual 480mm radiators), you still can use moderate speed and not worry too much about the performance.  This is the advantage of having multiple long radiators -- fan speed become far less important and the surface area of the radiator is going most of the work.


The M2 drives are trickier.  I routinely tell people radiators are insensitive to fan speed, meaning it takes a substantial change in fan speed to cause a relatively small change in coolant/component temp.  A M2 drive is the opposite and it is sensitive to both fan speed and airflow focus.  If this is a 1000D, you've got a long distance from any of those fans down the the lower motherboard heat shield area.  This is where those dual chambered cases PSU in the back and open fan slots in the bottom are useful.  The fans at the back and top are likely too far away to have a direct cooling effect.  If those temps really are of concern or you get into some heavy data transfers, you might look at fitting a small fan in that area to blow air off the heat shield on the lower motherboard.  You often can knock 10C off the temps that way. 

Link to comment
Share on other sites

You're exactly right.  I have a 1000D with just the CPU on the two front radiators and it cools very quickly when CPU temps jump.  The radiators are XR7s, 54mm thick.  I agree that surface area, more than fan speed, does most of the cooling.  But this fits right into my goal of lower fan speeds at the radiators, where air exhausts, to help keep positive air pressure inside the case.

BTW, I chose to exhaust through the front because the rig sits to the right of my desk with the back facing me.  I didn't want the heat and air flow disturbing me exiting the rear of the case.

I recently built a rig for my son using Corsair's newer dual chamber 6500 case.  Nearly the same set up as mine, but his cooling seems better configured.  As you say, the PSU in back with the fans at the bottom create great flow and cooling.  I plan on building a second rig for myself and family use, and I think I'm going with the 6500 case again for that one.

As for the 1000D, thanks for the suggestion of a small fan blowing on the SSD heat shield.  I'll look into this.

Great conversation, thanks for your insights!

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
  • Create New...