How should I connect my PWM pump + fans? (Z87X-OC Motherboard)

Associate
Joined
17 Nov 2008
Posts
1,680
I have a watercooled setup with a PWM pump, and five Corsair PWM fans on the rads. I bought the Gigabyte Z87X-OC motherboard because it was an advanced board and should suffice for my fairly simple build. However, I've encountered a rather glaring issue that I'm pretty disappointed with.

The motherboard was advertised as having 6 PWM fan headers, however after looking at the manual it seems only one header (CPU_FAN) uses pulses to change the speed of the fan. The others regulate speed by changing the voltage supplied.

This means any fan/pump that takes power from the PSU and speed from the motherboard needs to be in the CPU_FAN header, as if they are connected to the other headers they will run at max speed. I have a 5-way fan splitter that I also want to use for the radiator fans, so they speed up based on the CPU temp. However, the splitter requires power from the PSU, which means it also needs to be in the CPU_FAN header...

So here's my predicament, either I regulate the pump speed through CPU_FAN and connect the fans to the SYS_FAN headers individually, which would mean they aren't regulated based on CPU temp (I also can't connected all of them without some 2-way splitters). OR I have the pump running at full speed (noisy) and regulate the radiator fans.

How do other people with watercooled setups deal with this? It seems like a pretty big oversight, is it the same on other motherboards? Cheers.
 
I wouldn't want to run a pump off a fan header on a motherboard at all tbh.

Have you thought about an aquaero?

http://www.overclockers.co.uk/showproduct.php?prodid=WC-011-AQ&groupid=701&catid=2331&subcat=189
Nice idea but it's hard to justify spending another £50 on something like this, I've spent enough on this build already. Why wouldn't you want to run a pump from a fan header? What other options do I have to stop it running at full speed?

I also have another problem with the fan splitter, the more fans I connect to it the faster each individual fan runs. For example with 1 or 2 fans it runs at ~1200 RPM as I'd expect, but with 3, 4, 5 it runs at 1500, 2200, 3100 RPM. Any idea why this is?
 
I also have another problem with the fan splitter, the more fans I connect to it the faster each individual fan runs. For example with 1 or 2 fans it runs at ~1200 RPM as I'd expect, but with 3, 4, 5 it runs at 1500, 2200, 3100 RPM. Any idea why this is?

I suspect you may be seeing an issue where the tach feedback from more than 1 fan is being received by the motherboard. There should be only one of the 5 fans giving tach feedback on pin 3.

Does the fan actually get faster or is it just reporting a faster speed. I would suspect that 3100 RPM is probably above the rated speed of your fans.

What splitter, pump and fans are you using?
 
I suspect you may be seeing an issue where the tach feedback from more than 1 fan is being received by the motherboard. There should be only one of the 5 fans giving tach feedback on pin 3.

Does the fan actually get faster or is it just reporting a faster speed. I would suspect that 3100 RPM is probably above the rated speed of your fans.

What splitter, pump and fans are you using?

These are the fans I'm using. It might be reporting the incorrect RPM but they're definitely getting faster. When you plug in the 4th and 5th fan to the splitter you can hear it.

This is the splitter.
 
I wouldn't want to run a pump off a fan header on a motherboard at all tbh.

Have you thought about an aquaero?

http://www.overclockers.co.uk/showproduct.php?prodid=WC-011-AQ&groupid=701&catid=2331&subcat=189

You realise it only takes the PWM signal from the fan header and draw the power from (I'm guessing) a molex connector on the PSU?
This means any fan/pump that takes power from the PSU and speed from the motherboard

I've got the same problem. I'm just running the pump at max, because I couldn't think of a decent way around it.
Since I'm only running 2 fans I guess I could just run a normal 4-pin splitter and power both fans off the CPU_OPT motherboard header and the pump off the CPU_FAN header.
 
I've got the same problem. I'm just running the pump at max, because I couldn't think of a decent way around it.
Since I'm only running 2 fans I guess I could just run a normal 4-pin splitter and power both fans off the CPU_OPT motherboard header and the pump off the CPU_FAN header.
Yeah, a 2-way splitter would work fine because it doesn't get power from the PSU. I was thinking of getting a 2-way splitter and putting it on the CPU_FAN header, and running both the 5-way splitter and pump from that. Presumably the PWM pulse would limit both the pump and the fans to a percentage of their max speed, which is basically what I want.
 
Yeah, a 2-way splitter would work fine because it doesn't get power from the PSU. I was thinking of getting a 2-way splitter and putting it on the CPU_FAN header, and running both the 5-way splitter and pump from that. Presumably the PWM pulse would limit both the pump and the fans to a percentage of their max speed, which is basically what I want.

Yeah, the only issue is they'll both use the same percentage, which might be fine...

Someone on the forum was making 3pin to pwm converters you could try a few of those?

I think that was for controlling 3-pin fans with a PWM signal though.
 
I think that was for controlling 3-pin fans with a PWM signal though.

Yes it is. It can run 4 pin fans as well though, it just doesn't make use of the fan's PWM function. I have made one or two with one PWM feed split off to dual PWM converter boards which worked quite nicely and one or two with dual converters on one board.

A pump would ideally have its own circuit as they tend to be higher power than fans.

On your splitter does the pump or a fan have the connector with the green wire attached?

3100 RPM is indeed above the rated speed of your fans but could be coming from the pump, which may, or may not have a higher RPM. The noise increase could be a consequence of running more fans. I've never used a retail PWM splitter, having preferred to make my own custom cables, but there is no reason why fans should be running any faster if you have 1 or 1000 fans attached in a daisy chain, unless the PWM signal is being deteriorated by the fans, or isn't strong enough to activate the mosfet on multiple fans.
 
Last edited:
Back
Top Bottom