H100 Fan Control vs. CPU Fan Control

[bobn]

Is there any advantage in useing the build in H 100 fan controller as opposed to directly connecting the fans to a mobo cpu fan header? The H 100 will control fans based on coolant temp while the cpu fan header uses cpu temp. My understanding is pump rpm is fixed at 2000 rpm and has no relation to fan control.

 

I do understand the build in fan profiles and upcoming corsair link . Seems like useing a cpu fan header connection could be reliable and provide advanced fan control. What are the tradeoffs in haveing a radiator fan respond to changes in coolant temp as opposed to cpu temp?

[beekermartin]

As long as the pump speed doesn't change according to which profile is selected then using either the H100 or the mobo to control the fans shouldn't matter. The advantage to using the mobo is you can customize the fans to any profile you like using the bios or a fan controller software. Using the H100 controller requires opening the case and selecting one of the 3 profiles unless you get the corsair link.

[parsec]

Is there any advantage in useing the build in H 100 fan controller as opposed to directly connecting the fans to a mobo cpu fan header? The H 100 will control fans based on coolant temp while the cpu fan header uses cpu temp. My understanding is pump rpm is fixed at 2000 rpm and has no relation to fan control.

 

I do understand the build in fan profiles and upcoming corsair link . Seems like useing a cpu fan header connection could be reliable and provide advanced fan control. What are the tradeoffs in haveing a radiator fan respond to changes in coolant temp as opposed to cpu temp?

 

Yes, the CPU fan header provides good fan speed control, and likely more flexible overall than the Corsair controller. But in this case, the H100's fans are three pin fans, and not four pin PWM controlled fans. So connecting them to the boards CPU fan header will provide less control than the Corsair controller.

 

IMO, the main tradeoff with coolant temp control vs CPU temp control might be that coolant temp control is slower to respond to temp changes. OTOH, with liquid cooling the change in CPU temp will not be as abrupt or as fast then with an air CPU cooler. I doubt that CPU vs liquid temperature monitoring makes any difference with these coolers, otherwise users would see a spike in CPU temp at first, and then a decrease in CPU temp, which has not been reported by users.

[RAM GUY]

Correct me if I am wrong;

3-pin fans - the fan is controlled by voltage, meaning fan speed is adjusted by some sort of physical means, like a knob to turn or a button to press. Not sure if it's possible to adjust the fan speed through the motherboard (i.e. Speedfan, bios, other software won't have any effect on the fan's rpm).

[parsec]

RG, some mother boards have BIOS and/or software capability to control the speed of three pin fans, but it is usually not very flexible and useful. It is never as flexible as PWM fans or the capabilities of the Corsair H-series fan controller. That capability in mother boards has gotten better recently with some manufactures, given what I have seen, but still does not match PWM fans or the Corsair controller. I'm referring to software included by mother board manufactures with their products, that only works with their products. It never provides the flexibility of a manual control like you mentioned, but those are not automatic. Of course, I have never used every mother board and it's fan speed control software, so if some exist that are very good, I'd like to hear about it.

 

Otherwise, SpeedFan is the only generic software that provides fan speed control for PWM and (I think) three pin mother board fans that I am aware of. SpeedFan works, but IMO is a bit complicated, not particularly intuitive in use, and asks more from the user regarding learning how to use it than most people are willing to put into it, including myself. Or, it's documentation is poor or does not exist. I noticed that may be improving recently.

 

IMO, the main problem with mother board three pin fan speed control is which device is being monitored that will trigger changes in the fan speeds. Usually it is mother board temperature, which never changes in the way that CPU temperature changes. Then the software usually supplies a few predefined fan speed "profiles" that provide very little change in the fans speed, relative to that provided by PWM fans. A newer board I have allows selecting among three devices (including the CPU) on the board whose temperature will trigger increased fan speed, and allows the user to define their own fan speed profile, or a speed relative to temperature "curve". IMO, this still does not match a good PWM fan based software control, and so is not (yet) as useful for CPU cooling.

 

All of this is what I believe inspired the fan controller in the H80 and H100. I think mother board fan speed control has been kept simple (dumb) due to the wide range of case fans and their differing performance characteristics and the issue of correctly controlling all of them. We can see the potential problems with more sophisticated three pin fan speed control, such as those that have been noted in some cases with the Corsair controller. Some fans do not function as expected when used with it. PWM control is more easy to implement than voltage control, but I have a board where the PWM fan must be "calibrated" with the software so it will work optimally. The differences in fan performance seems to be a difficult issue to overcome.

 

All mother board fan speed control software is not created equal, with one manufacture in particular that is known for how poor their software has been in the past and continues to be. This type of software and the settings in it's UI or the boards BIOS can be confusing to some users, as we have seen in this forum.