[SOLVED] How does the pump in an Arctic Liquid Freezer II get power ?

Colif

Win 11 Master
Moderator
Trying to work out how the AIO pump on my Arctic Liquid Freezer II gets power.
I know it runs off the PWM signal but I don't know how it gets power.
Apart from being physically installed on the CPU, its not electrically attached as far as I can tell.

The only two cables that were attached to the AIO were both for the ARGB fans attached to AIO. There is nothing else running to MB.
There is a cable attached to the fans that runs down length of the air pipes that linked the fans to PWM and ARGB headers on board.

But now I have replaced those fans with a set of Lian LI fans that run off their own controller which is attached to those two headers.
Controller has its own PCIe 6 pin running it. I guess it could be power source.

So I know the pump runs at 100% speed at 40% Fan speed, but where is it getting power from?
I have to assume its the motherboard but not sure how its getting pump to work? How does the pump get the PWM signal? Magic?

its just a mystery to me. I had to ask so I could get some sleep :)
 

Lutfij

Titan
Moderator
View: https://www.youtube.com/watch?v=ZQRwn-lecnk


So I know the pump runs at 100% speed at 40% Fan speed, but where is it getting power from?
I have to assume its the motherboard but not sure how its getting pump to work? How does the pump get the PWM signal? Magic?

The pump get's it's power from the PWM header/connector or so that's what it looks like after watching the video and reading through their installation manual.
 

Colif

Win 11 Master
Moderator
Yes, the original unit comes with a 4 pin PWM and a 3 pin ARGB cable that run down the entire length of the tube. and attached to the CPU_OPT & ARGB header on board.

https://www.arctic.de/media/2b/3f/96/1690272955/Spec_Sheet_Liquid_Freezer_II_360_A-RGB_EN_update.pdf

We have unplugged both of these from the motherboard.

The Lian Li fans run off their own power cable that attaches to the Lian LI controller unit which is attached to three headers on MB
1 x ARGB header
1 x PWM header
1 x USB2 header

Lian LI controller powered by its own 6 pin PCIe cable running from PSU.

None of these touch the cable that runs through water lines. its not attached to anything now.

Some how the PWM signal being sent to the fans also gets to the AIO pump via the CPU)Fan header even though its not attached to it?
kSkYor5.jpeg


clearly I am missing something here.
 
Are there ANY cables going from your pump body/main unit, to ANYTHING, like an ARGB controller? There MUST be some type of cable running from your unit to something, otherwise there is no way possible the pump could be getting power. It can't magically get it just by being in contact with the CPU or via air transfer. LOL. If it's connected to an ARGB controller and that is the only thing that's connected, then it must be getting power via that. I don't think I'd allow that. I'd recommend connecting an appropriate PWM cable from the pump to any configurable PWM header on the motherboard so you don't burn out your ARGB controller, if that's the case.

Looking at the manual, the only thing it lists is a PWM or 3 pin fan header cable, but there HAS to be something else. SATA cable, ARGB cable, something. But even if that ARGB cable IS powering the pump, it would not be capable of providing an RPM signal so there should be no way to see anything related to the pump RPM in the BIOS. I really don't think you want this to be this way. I mean, I know I wouldn't. I'd connect the original PWM cable to a header to ensure it's getting power in the way that it was intended.
 
Last edited:
Looking at the bottom of the unit, I can see that there ARE two cables coming from it, but I can't see if it's one PWM and one 3 pin cable, or something else. Or, is this different than your unit? Because if there is nothing at all plugged into your pump, then it is NOT running.

install4.jpg
Pump is not speed adjustable even as it is PWM so you can plug it in anywhere on the MB. Don't know ehich MB you have but most modern ones have AIO_PUMP /which is not speed adjustable or W_PUMP which is and gives more power or any Case/System_FAN.
As for radiator fans, whatever you have, original or other make. directly or thru hub, they should be connected to CPU_FAN connector and only them not mixed with any case fans,
 

Colif

Win 11 Master
Moderator
Mystery solved. I had a friend helping me install it last night and I asked him if he left any of the old cables attached to CPU_OPT header and as it turns out he left the pump one attached.

The current fans are plugged into CPU_FAN so it will be getting its flow rate from them.

Good thing he did as PC wouldn't be working right now.
 
Pump is not speed adjustable even as it is PWM so you can plug it in anywhere on the MB. Don't know ehich MB you have but most modern ones have AIO_PUMP /which is not speed adjustable or W_PUMP which is and gives more power or any Case/System_FAN.
As for radiator fans, whatever you have, original or other make. directly or thru hub, they should be connected to CPU_FAN connector and only them not mixed with any case fans,
Every modern board I've worked with in the last, oh IDK, maybe five years, that wasn't a bottom of the barrel budget board, had adjustability on EVERY header no matter what it was called. Output type (PWM or DC controlled), input type (CPU, motherboard 1, VRM, etc.), curve customization, etc. regardless of whether it was a FAN_1, CPU, AIO, etc.

On cheaper boards there are likely some that don't have this adjustability, but in the last few years I haven't worked with any. Even the really cheap B460m ASRock board I used for a budget build for my sister's desktop had this.
 
Mystery solved. I had a friend helping me install it last night and I asked him if he left any of the old cables attached to CPU_OPT header and as it turns out he left the pump one attached.

The current fans are plugged into CPU_FAN so it will be getting its flow rate from them.

Good thing he did as PC wouldn't be working right now.
Yeah, I figured something had to still be connected and whether it was to the CPU-OPT, AIO or some other header, I knew there had to be something else it couldn't be working. At least you figured out what though.