Setting fans to 0 RPM

[Blueyoshi03]

So I have a fan hub connected to my CPU header so all 6 case fans scale to the CPU temperature, however the lowest my BIOS allows is something like 30% fan speed.

Is there a way to configure all of my case fans to go between 0-100% usage depending on CPU temperature?

Because I could have passive cooling through 40c then actually have fans turn on, then scale fans from 0% to 100% speed once it gets up to 70c.

Hope this makes sense

 

[Jamin280672]

So I have a fan hub connected to my CPU header so all 6 case fans scale to the CPU temperature, however the lowest my BIOS allows is something like 30% fan speed.

Is there a way to configure all of my case fans to go between 0-100% usage depending on CPU temperature?

Because I could have passive cooling through 40c then actually have fans turn on, then scale fans from 0% to 100% speed once it gets up to 70c.

Hope this makes sense

It does make perfect sense, and ive tried this myself before, unfortunately my bios doesn't allow for 0% either, I think the reason behind this is most, if not all boards have a feature in the bios "shutdown computer on CPU Fan Fail, and 0% would in affect trigger this feature.

 

[PaulCa]

It also comes down to the fan itself. Some PWM fans when they receive 0% will spin at full rpm, some won't go below 30%. The PWM fan spec allows this. The spec is much clearer that all fans should spin at 30%.

Your motherboard wants to avoid "stalled" fans. When you send them 20% say, it might not provided enough power to turn a stationary fan, so it will sit stationary consuming power and potentially burning itself out, or even literally burning.

One solution is to move the fan header to a fan header that does allow 0% and use software to configure the CPU fans. Obviously the downside here, and why I don't recommend it, is if that software fails to run or you forget to run it, the CPU could overheat.

Personally.... my CPU fans (on an AIO radiator) will not go below 750rpm, but they are silent enough at that speed so I'm fine. When my GPU sends 0% to stop it's fans, the GPU radiator fans will not spin lower than 500rpm and the auxiliary fan on the GPU itself to cool the VRMs and VRAM goes to full 100% rpm. However, if I get MSI Afterburner to send 20% or 25% then the fans all spin minimum rpm of about 500 and make virtually no noise.

 

[EsaT]

That's certainly harder question in case of motherboard based control.
Aquaero certainly can control fans how ever you want.
(even using average from CPU/GPU temp and physical sensor etc)
But haven't tried dropping CPU fan speed to zero and I have also speed signal going to CPU fan header to avoid motherboard going nuts over lack of CPU fan speed signal.

Using Aquaero you can just forget buggy BIOSes/softwares/Windows or even if mobo/CPU is in some kind hard crash/lock situation
Once properly installed/set up and programmed controls will run always when ever it gets power from PSU.
https://www.overclockers.co.uk/search?sSearch=Aquaero+LT

It also comes down to the fan itself. Some PWM fans when they receive 0% will spin at full rpm, some won't go below 30%. The PWM fan spec allows this.

So much for the advantage of PWM fans when you can't even control them in guaranteed way...

 

[Blueyoshi03]

Just strange that it’s so seemingly difficult to get this configuration considering a lot of GPUs can keep stationary fans until needed. My MSI 970 runs idle up to 50c before fans turn on.

So at some point I also want a 360mm aio with push/pull config connected to my GPU. Is there a way to have the push/pull fans react to the GPU like my CPU fan hub?

 

[EsaT]

Just strange that it’s so seemingly difficult to get this configuration considering a lot of GPUs can keep stationary fans until needed. My MSI 970 runs idle up to 50c before fans turn on.

So at some point I also want a 360mm aio with push/pull config connected to my GPU. Is there a way to have the push/pull fans react to the GPU like my CPU fan hub?

For one thing in average GPU loads surely have less/slower variation over time than CPU load.
I mean you start the game and GPU load goes up until you end game, then it basically stays low pretty much until you again start playing some game.
Again CPU/motherboard is designed from perspective that CPU load can vary constantly from near zero to full load.

Aquaero can control any of its outputs any way you want.
Controlling fans and pump of custom built loop with data/monitoring of water temperature and flow is what it's been designed to do.

 

[PaulCa]

So much for the advantage of PWM fans when you can't even control them in guaranteed way...

This is not the fault of PWM. It's the individual fans as manufactured. Out of all my fans:

Artic F14s 140mms will stop.
Artic F9 90mm goes to 100% rpm with a commanded 0%.
NZXT 120mm radiator fans maintain 500rpm
Corsair 120mm radiator fans maintain 750rpm (but as they are on the CPU header this might be BIOS intervening).

So the PWM fan spec does not state what the fan MUST do, but you can always ask the manufacturer or search reviews to find out if the fan in question will stop or not.

DC control of brushless fans is just wrong on so many levels. I'm amazed that manufacturers still make these fan controllers. Granted it works, but it's just a hack.

 

[LuckyBenski]

The most common behaviour is for PWM fans to maintain their minimum speed if given a signal too low i.e. one that would/might cause the fan to stall.

I'd say the second most common behaviour is for fans to run at full speed if the PWM duty cycle is too low - used to see this on older motherboards (10+ years ago).

In fact I've never seen a PWM system fan that will stop, personally. The feature on GPUs only came about a few years ago with 900 series and it requires the fans themselves to support it. If you wire the card up to normal fans they won't stop.

Bear in mind there is circuitry inside a PWM fan that manages speed, it's not simply the supply voltage being pulsed like most other PWM scenarios.

DC control of brushless fans is just wrong on so many levels. I'm amazed that manufacturers still make these fan controllers. Granted it works, but it's just a hack.

I see this mentioned all the time, and I'm an electronic engineer but what exactly is wrong with doing this?

 

[PaulCa]

I see this mentioned all the time, and I'm an electronic engineer but what exactly is wrong with doing this?

My experience with brushless motors is in higher power motors. 200, 300, 400W motors. As I understand it, the motor is designed to take a constant peak voltage on multi phase phase AC. To change the speed you change the duration of the peaks to push/pull for less time as the rotor passes the stator magnets. The frequency remains constant (in a sensorless brushless anyway) the peak pulses change in duration.

Lowering the voltage, for which there is some tolerance, is not how they are designed to be slowed. Lowering it too far can lead to "cogging" or a rotor stall causing excesses wear and heat build up. Also the circuitry controlling the AC switching is designed to run at a fixed voltage, most likely 12V in our case. Running it at 6V or lower and it's a case of YMMV.

As I said I'm surprised it works so well at all.

Now maybe PC fans, being much lower power are simplier, the set up I am used to would entail circuitry equivalent to an ESC (electronic speed controller). I donno.

 

[LuckyBenski]

My experience with brushless motors is in higher power motors. 200, 300, 400W motors. As I understand it, the motor is designed to take a constant peak voltage on multi phase phase AC. To change the speed you change the duration of the peaks to push/pull for less time as the rotor passes the stator magnets. The frequency remains constant (in a sensorless brushless anyway) the peak pulses change in duration.

Lowering the voltage, for which there is some tolerance, is not how they are designed to be slowed. Lowering it too far can lead to "cogging" or a rotor stall causing excesses wear and heat build up. Also the circuitry controlling the AC switching is designed to run at a fixed voltage, most likely 12V in our case. Running it at 6V or lower and it's a case of YMMV.

As I said I'm surprised it works so well at all.

Now maybe PC fans, being much lower power are simplier, the set up I am used to would entail circuitry equivalent to an ESC (electronic speed controller). I donno.

That's pretty valid. Lowering the speed a fraction may be less risky/more predictable, but pushing the fan right outside its normal behaviour is another story. Thanks for elaborating

 

[EsaT]

My experience with brushless motors is in higher power motors. 200, 300, 400W motors.

Those are running hell of lot more power demanding load than motor in PC fans.
Which basically don't have any external load (only friction of bearing) until work in pushing air starts to eventually cause it when RPM starts rising.
And from strength of airflow I would say that load doesn't start to increase much until RPM starts moving toward 1000.
I mean some 500rpm fan really can't push air much.

 

[chroniclard]

IIRC mine go down to about 25% and are silent at that speed. Do you need them at zero? will be almost silent at low speeds anyway.

 

[PaulCa]

My fans that continue at 500 or even 750rpm as so quiet I can here the AIO water pumps but not the fans.