Mobo not recognizing chasis fan

I have an ASRock X99 WS motherboard and I've been trying to set the cooling in a way that allows for quiet operation when using productivity software, i.e. when the CPU is idle most of the time, and keep components cool when running heavy workloads, such as GPU Compute and games.

I eventually decided to go for the following configuration, which I present here in case people find it useful:

1x ThermalTake Commander FP connected via PWM (4-pins) to the CPU1 fan header.
2x PWM CPU Fans connected to 4-pin ThermalTake Commander FP fan header.
1x PWM Front Upper Fan connected to 4-pin ThermalTake Commander FP fan header.

Speed of CPU1 fan header-attached fans is set to zero for temperatures below the CP's idle temperature. This ensures that the system is quiet when the CPU is cool. It is set to ramp up the speed as the CPU's temperature rises.

The reason I used the ThermalTake Commander FP is to ensure that the front upper fan's speed is in sync with the CPU fans' speed, so that they are fed with sufficient, cool fresh air.

For the cooling of the motherboard and the supply of fresh air to the GPUs (I use Founder's Edition cards, which extract air from the case, potentially causing negative air pressure), two more fans are configured as follows:

1x PWM Front Lower (high speed - up to 2000rpm) Fan connected to 4-pins CHA2 fan header, which uses voltage regulation (i.e. as if it was a 3-pins). I noticed that the fan is activated when voltage is at least 40%, hence I configured this fan to operate at 40% for up to 40C motherboard temperature and then quickly ramp up speed, if motherboard temperatures rise.

1x PWM Back (up to 1000rpm) Fan connected to 3-pin CHA1 fan header. I set this to operate 100% at all times. It runs quietly, at about 1000rpm.

When the CPU is idle and the fans connected to the ThermalTake Commander are off, the lower fan feeds more air into the case than the back fan pulls, ensuring positive air pressure.

When the CPU fans are ramping up, which typically happens alongside the GPU fans, the back fan can't pull as much air as the front upper fan pushes, which allows for some fresh air to go to the GPUs.




Edited by ntsarb - 22 May 2017 at 1:39am

overclockmonkey wrote: Hello everyone My Z97 Anniversary isn't recognizing the chasis fan at all.  Neither is the application SpeedFan. I want to change the fan speed but I'm tired of using the UEFI all the time.  FANtastic tuning is having trouble recognizing it as well as the hardware moniter in AsrockTuning.  However  Fantastic tunning does work with it even though it always states it's at 0 rpm.  The problem is that I can't keep using it since it lags for some reason even though I have 8gigs. Not to mention at times it gives me trouble by saying "out of memory". The app shop however doesn't even bother to run, it always says that. However I didn't have that problem when I had the 4690k, now that I have the G3258 (which was what the motherboard intended for), now I can't use it. Last note: The 2 fans in my case are connected to a fan hub that came with the case. Here's a picture of it. Since both case fans are 3 pin instead of 4, I attatch them to that. The left side of the hub goes to the chasis fan header and the right side goes to the psu. TL;DR z97 anniversary motherboard can't recognize the fans connected to chasis fan header. FanTastic works but it always says that the header is at 0rpm.  AsRockAtunning lags  a bit and is "out of memory"  from time to time

Starting from your first post, lets examine your statement about how the fans, etc, are connected to the fan hub of your case:

Since both case fans are 3 pin instead of 4, I attatch them to that. The left side of the hub goes to the chasis fan header and the right side goes to the psu.In your picture, the right side is the four pin PWM fan connection, and the left side is the 12V power to the hub, not that this matters much.

How do I know that? From a picture and text from the Phanteks Enthoo Luxe manual, this one:



This is a four pin PWM controlled fan hub that can work with three pin fans.

The first major detail we need to examine is the four pin PWM input on the right side of the hub, and the description in the text. Notice that the instructions specify connecting that input to a four pin PWM CPU Fan header.

Reading the next paragraph, we find a warning about most mother boards with four pin Chassis fan headers that seem to be PWM controlled, but they aren't. This is exactly what is happening to you.

Usually, ASRock boards have a four pin Chassis Fan1 fan header. This Chassis Fan1 header may or may not be PWM controlled. On a sub-$100 board like the Z97 Anniversary, I doubt it will be a true PWM fan header. The four pin Chassis Fan2 header is even less likely to be a true PWM controlled fan header.

Your own description of how the fans worked fine when you had the four pin input to the fan hub connected to the CPU Fan1 header, but not after moving it to one of the Chassis Fan headers, verifies what I said above.

Are you using the Chassis Fan1 or Fan2 header now for connection to the fan hub?

The next paragraph is important, since it describes what happens when the fan hub is not connected to a true PWM fan header, which is your major problem:

which is limited to a total device consuming 30W in total.

This is telling us that you have three fans being powered by the single Chassis Fan (1 or 2) header, since your current input connection will not use the 12V power from the PSU.

What bothers me in that statement is the limit of 30W total. That assumes all mother board CPU Fan headers are capable of providing 30W at 12V. That is 2.5Amps, which is way beyond the capacity of most if not all current mother boards. Your board will likely provide 12W max from the CPU FAN1 header, which is 1Amp, but the Chassis Fan headers may be even lower than that, about 6W or 0.5A.

Since we have no idea what the fans you have connected to the hub are rated at for power, you should check that and to be safe unplug at least one of them until you change the current configuration. That is very important! The strange problems you are having might be related to the probable overload on the Chassis Fan header.

Finally, we see in the instructions that only the fan that is connected to the hub's FAN1 output will have its speed/RPM signal sent back to the mother board. It is impossible to monitor the other fan's speeds when using the fan hub.

Of course, that happens when the fan hub is connected to a true PWM fan hub, and is working correctly. Who knows what happens when it isn't as in your current configuration.

TL;DR The four pin Chassis Fan headers are not true PWM controlled fan headers, so the fan hub does not work correctly when connected to those fan headers. Part of the incorrect operation of the fan hub is the three fans connected to the hub are drawing power from the Chassis Fan header instead of the PSU. The Chassis Fan header will be overloaded by the power needed by the three chassis fans, which may be causing some or all of your other issues.

I can think of two simple fixes for you that will have the fan hub working correctly, and allow your four pin PWM fan on your CPU cooler (IF that is what you have, we don't know) to change its speed with the CPU temperature.

1. Connect the fan hub to the CPU FAN1 header with the four pin/wire cable. Connect the CPU cooler's fan to either the CPU FAN2 header, or the fan hub's FAN1 header.
2. Get a PWM fan splitter cable, with one input and two outputs. One of the two outputs will have four pins, the other will have three pins. The four pin output will be connected to the fan hub. The three pin output is connected to the CPU cooler fan.
3. If you use a PWM fan splitter cable, connect it as I suggest in (2), which will guarantee the correct operation of the fan hub. But this method will not allow you to monitor the CPU fan speed. You can also try connecting the PWM splitter's three pin output to the fan hub, and the CPU Cooler's fan to the four pin output. This configuration will still send the true PWM control signal to the fan hub, so it will work correctly, but you will lose the RPM signal from the hub's FAN1 connection. But you will get the CPU cooler's fan RPM signal sent back to the board.

Speed Fan does not work with every mother board. It also can't work with fan hubs that have one input and output, and be expected to control each fan individually.

The "lag" you talk about in A-Tuning may simply be its sampling or polling rate. That is, A-Tuning only reads the fan speed data once every four or five seconds. All monitoring programs work like this, but some have a shorter interval between data reads, or can be adjusted.

When you see the "Out of memory" message, have you checked your actual memory usage? Checking Resource Monitor by clicking on the Memory button at the top will show how much memory is being used by what, and how much is available.

Yea, I see now you've edited it and included more info in doing so. re: UPDATE

I'm more confused than when I first posted now though :)

Your post immediately above says the three fans are controlled. Yet in your update it says the app couldn't read the tach wire/RPMs of the fans.

What I believe your seeing is the app/BIOS varying the voltage applied to the one fan that it _can_ read. And by default the other two are reacting as such.

Still, what with the out of memory and other abnormalities you have going on there I would separate off the chassis fans to the hub you have and leave the CPU fan as the only one on the CPUFAN header.

And TY for recognizing my efforts here helping, or at least my trying to.

Well, as I am a User here and in no way associated with ASRock you can do whatever suits you.

What I said is fact and there's no getting around it. One MB header might power two fans but it is impossible to control those same two fans

Opps. I just re-read your 1st post.

No!

You can not read two fans connected to one Mb header. It just isn't electrically possible. No mechanically.

It will only read one. Think. How would any sensor be able to determine each fan when both fans signals are coming at it on the same wire???