What's new
  • Please do not post any links until you have 3 posts as they will automatically be rejected to prevent SPAM. Many words are also blocked due to being used in SPAM Messages. Thanks!

Triggering fans on coolant temperature instead of CPU temps?

bjcsoln

Well-known member
Joined
Apr 19, 2013
Messages
324
Location
Vancouver
So, now that I have my AIDI64 panel (sort've) setup, I noticed that the default ASUS qfan controller profiles have the graphs triggering the fans based on CPU/GPU temps.

Question - does it make more sense to trigger the fans based on the coolant temperature? Its the coolant that takes away the heat, so, for example, as long as its staying at 32c (as an example), even if the CPU goes up to 50c increasing the fans is only going to help keep the coolant, well, cooler?

What do other people do for their profiles? do you trigger off of hardware temps (like CPU or GPU) or off of the coolant temps?

Thanks,
 

bjcsoln

Well-known member
Joined
Apr 19, 2013
Messages
324
Location
Vancouver
Okay quick update - looking in my BIOS to be honest I don't know what temperature device they are being triggered off of..Based on what I could observe, I think its the CPU temp (when it hits about 50c I can see the RPMs going higher but I am not sure.
 

Izerous

Well-known member
Folding Team
Joined
Feb 7, 2019
Messages
1,174
Location
Edmonton
I started water cooling back around ~2003 or so. I stopped using external sensor taped to the blocks and coolant sensors around 2006 and never tried to use them again. Yes more data is never bad in theory but useless data is useless. Short version coolant temps don't really show what is going on that well, loop equalization, lag of temperature, where the sensor is (pump ads heat too) etc I found that the CPU/GPU temps were all that really mattered and pulled all of the external sensors.
 

sswilson

Moderator
Staff member
Joined
Dec 9, 2006
Messages
19,710
Location
Moncton NB
Yeah, I'd think that the "lag" mentioned above would be the problem when it comes to using coolant temp as your trigger. CPU/GPU temps would have spiked to potentially unsafe levels for quite a long period of time before those temps would affect a large enough change in coolant temp to drive fans to bring down the temps. The same would be true the other way.... CPU/GPU temps would drop almost immediately but coolant temp would keep the fans running for 10 - 15 minutes before those temps would drop.
 

lowfat

Moderator
Staff member
Joined
Feb 12, 2007
Messages
10,305
Location
Grande Prairie, AB
I wouldn't tie the pump speed to coolant temp but I see no issue w/ fan speed being tied to it. The temp of the GPU and CPU depend wholly to the water temperature anyways as long as flow rate is sufficient.
 

Izerous

Well-known member
Folding Team
Joined
Feb 7, 2019
Messages
1,174
Location
Edmonton
I wouldn't tie the pump speed to coolant temp
Agree with this part 100%. Pumps should essentially be cranked to highest comfortable speed and left alone. dual D5S in series and isolated and they are nearly inaudible
 

Entz

Well-known member
Joined
Jul 17, 2011
Messages
1,536
Location
Kelowna
Yeah der8auer did a video on that recently. Makes a world of difference.

I have tried it both way ways and honestly I prefer CPU. Between coolant "stabilizing" time (lag above, no idea what to call it lol) and ambient temperature differences it always felt like I was trying to hit a moving target.

That being said I actually prefer a constant fan profile that is within my desired threshold for noise. Constantly changing fan noise bothers me more than just a static one. Everyone is different however.
 

clshades

Well-known member
Joined
May 18, 2011
Messages
3,413
Location
Calgary
Agree with this part 100%. Pumps should essentially be cranked to highest comfortable speed and left alone. dual D5S in series and isolated and they are nearly inaudible
That's not how to run a pump at all. Whoever came up with that little gem is waaaay wrong.
 

Izerous

Well-known member
Folding Team
Joined
Feb 7, 2019
Messages
1,174
Location
Edmonton
That's not how to run a pump at all. Whoever came up with that little gem is waaaay wrong.

You mess with dynamic pumps speeds or a different approach? By comfortable speed/temp/noise maybe not detailed enough. The D5S's I let go wild since I can't even hear them. The DDC I used to use I had dialed back a little due to noise but I don't try at all to control beyond a static speed.

Benefits do decress off after certain flow rates but they don't completely flatten out.
 

bjcsoln

Well-known member
Joined
Apr 19, 2013
Messages
324
Location
Vancouver
Weighing in on ClShades point (I think this is where he is going on this), in a small closed loop system, I also don't think the water flow speed (and therefore the pump) makes that much difference. Its really about the process for how much energy the water can remove from the block as it passes over.

Question (and maybe an assumption here) - but if the coolant is cooler, then it should be able to transfer more heat away from the block?

Vs (as an example) if the water if warmer, but moving faster across the block.
 
Top