Jump to content
Corsair Community

Setting up a proper silent fan curve


Recommended Posts

I have the following equipment;

  •  H150i Elite Cappelix (at the top of my case, and in the position of blowing out hot air)
  •  ML120 Pro 3 Fans (One of them is exhaust fan and blows hot air outside, two of them are located at the bottom of my case and blow cold air inside) 
  •  AF 120 Elite 3 Fans (All of them are on the side surface of my case and blow cold air into my case) 

I want to create a quiet "cooling profile" without compromising cooling.

Can you give me a suggestion?

This is the profile I am using now.

Also, what should I choose as a sensor source?

WhatsApp Görsel 2024-06-14 saat 11.51.53_3be4d36b.jpg

Ekran görüntüsü 2024-06-14 115035.png

Link to comment
Share on other sites

The fan curve for your radiator fans should be configured to use the coolant temperature, not the CPU.

Why? The fans on the radiator do not cool the CPU!! They cool the coolant! It's the coolant that then cools the CPU. As the coolant has a much higher specific heat capacity than the CPU, the cold plate and ... well pretty much everything else in the system. Because of that, it will warm up (and cool down) much slower than  your CPU. Also, it won't have the sudden temperature spikes that lead (inevitably) to fans constantly ramping up and down.

With that said, you should configure your fan curve with a max temp of around 40-45C (fans spin up aggressively) with a more gentle ramp up to, say, 38C. Your baseline/minimum temperature for the curve will be room ambient. Physics won't allow it to be cooler than that.

You're also exhaust warm air (from the inside of the case) through the radiator. This will have the effect of warming your coolant a bit - again, physics. To counter this, you should set your airflow fans (side, bottom, rear) to be based on the internal case temperature. You can get a good reading on this by placing the single temp sensor that comes with the Commander Core in the exhaust flow of your rear fan. You'll need a little experimentation but, in general, this curve should be similar to your radiator curve as it determines the minimum possible running temperature of your coolant. If the internal air flowing through the radiator is warmer than the coolant, then it'll just warm the coolant to the air temp (eventually). Again ... thermodynamics at work here.

I'd also suggest mounting the exhaust fan a bit lower if you can. That'll help pull out the GPU heat before it can impact the cooler. Ramping up those airflow fans will help to counter the GPU heat and the exhaust fan will pull that air past the radiator fans. You may want your exhaust fan a little faster than your airflow fans, too.

  • Like 1
Link to comment
Share on other sites

Sensor choice should be "H150i Elite Capellix Temp" also sometimes shown as H150i Coolant Temp.  That is the measure of heat added into the AIO from the CPU and including changes in ambient case temp.  It's a slow changing variable that also is a good representation of case temperature changes, so it works well for all fans including those not on the radiator.  

 

When you cold boot up the PC the AIO liquid temp will be equal to the room temp for the first minute and then slowly creep up 4-7C and level off.  This is the effective baseline system temperature and it's influenced by your room temp, the case design (closed vs open), and the case's location in the room (open and airy vs under the desk).  Most people are in that +4-7C above room temp range, but it is unique to your environment.  Whatever the value, this is where you set the fan speed you want for normal, quiet working conditions.

 

Let's say this is 30C.  Set you first and second points at 30C = Quiet fan speed (whatever that is to you).  Keep the second point at the same speed and move 3C or so up (33C).  This keeps the fans from fluttering around because the room temp is +-1C different at that moment.  From 33C start slowly stepping up fan speed.  Your target is going to be the normal maximum liquid temp you see, likely when gaming and not a CPU stress test.  Most users see around +10C above the baseline and most of this is from GPU heat.  Using the 30C baseline, you set 40C to the highest fan speed you are willing to tolerate when gaming minus 100 rpm.  For most people this will be in the 1100-1500 rpm range, but there is no magic number and a relatively small difference in cooling effectiveness in this range.  If it seems to loud, go down.  Be irritated is not worth 1C of any component temperature.  Save the last control point for an emergency blast -- something like 50C=2000.  You should not reach 50C.  If you do, you want to know.

 

There is another alternative to this that some may prefer.  You don't really need your fans to adjust their speed as temps change in 1C increments.  Component cooling isn't that sensitive where you need to modulate like this.  So you create a very flat quiet work curve like above, but keep the speed very even.  30C=500 rpm, 35C=600 rpm, 40C=750 rpm (or similar) and then again put a fan blast at 50 or 55C to let you know the temps are out of bounds.  This is your work curve.  The fan speeds always will be low and you can't overheat the CPU except by deliberately loading it with a synthetic stress test.  Then when you are going to game, render, or do anything else with high continuous load, you change the fan preset to Fixed RPM.  Again pick a speed that is just under your noise limit.  You leave it there until done, then toggle back to work fan curve.  Continuous static fan speed will become part of the background and you won't hear it.  This tends to be far less intrusive than constantly changing fans speeds that we do notice.  

Edited by c-attack
  • Like 1
Link to comment
Share on other sites

Thank you for your wonderful answers.

So what should I choose for the pump? Extreme, quiet, balance?

By the way, I have only one Commander Pro, 3 fans are connected to my motherboard, so I left the control of these 3 fans to Asus.

There is also a pump setting in the bios. Should I touch these places? Or is it better to leave it as default?

Ekran görüntüsü 2024-06-14 191653.png

Link to comment
Share on other sites

BIOS Pump setting won't matter.  It control the specific W_Pump or AIO_Pump fan headers on your motherboard.  Those headers are really CHAssis fan headers renamed and set to 100% for very basic, small AIOs that take their power from a fan header.  Nearly all current AIOs regardless of size use a SATA or Molex connector for power and a fixed pump speed.  

 

As for that, you'll control your pump speed in CUE only.  The BIOS is not capable of changing it, even when trying to adjust the CPU fan header where the tachometer wire is connected.  That is just a tach wire and nothing else, so no power or control along that path.  It does make up part of the boot protection sequence.  If your pump doesn't talk back to the motherboard on boot, it will throw up a CPU_BOOT_ERROR at the BIOS load screen and you'll know something is off.  

 

As for preferred pump speed setting, you probably can put it on quiet and leave it there forever.  AIOs have a short loop length and not too much flow restriction.  It doesn't take a ton of pump speed and there isn't a lot of flow rate difference between 2000 rpm and 2800 rpm.  In a custom loop with a lot longer length and possibly more vertical height to travel, pump speed is more important.  Also those models have a range of 800 rpm to 4800 rpm, so there is a lot more to play with on the top end.

 

You can do a quick assessment to see if pump speed matters for your CPU.  Put it on quiet and any moderate fixed fan speed like 1000 rpm.  Start a fixed load stress test like the one in the bench tab of CPU-Z.  This is a moderate fixed load under your watt limit, so it won't bang off the top end and is good for showing how the cooling works.  Start on quiet pump speed.  The CPU temps will hold pretty level and let it go for 5-10 seconds.  Now up it to medium.  Do the temps drop 1-2C as you change?  Wait another 5-10 seconds and go to Extreme.  Check for temp drop again.  You also can do this from Quiet to Extreme directly to try get a clearer picture from min to max.  Typically only people with large die, high core count CPUs will see differences between the pump speeds.  1-3C when at max load is probably about right.  However, for normal use you likely don't come anywhere near this level of CPU wattage and thus the Quiet speed works just as well with no difference in performance at all.  

Edited by c-attack
Link to comment
Share on other sites

1 hour ago, Ous said:

By the way, I have only one Commander Pro, 3 fans are connected to my motherboard, so I left the control of these 3 fans to Asus.

You don't have a Commander Pro. You have a Commander Core with that AIO. They are quite different.

While there are more fans than headers, you can use (multiple 2x) splitters or a powered fan hub (ideally). You can then control multiple fans as a single fan from iCUE using your temp sensors. This would be good for the intake fans (front/bottom).

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...