• 05.07.2020, 17:37
  • Register
  • Login
  • You are not logged in.

 

Dear visitor, welcome to Aqua Computer Forum. If this is your first visit here, please read the Help. It explains how this page works. You must be registered before you can use all the page's features. Please use the registration form, to register here or read more information about the registration process. If you are already registered, please login here.

maximum fan output selection for controller?

Tuesday, January 29th 2013, 11:35pm

hi folks :) ,
I'm a rookie in water cooling and I'm just planning to assemble my first rig. I'm planning to buy aquaero controller since it is the most complete hardware/software combo out there. yet I'm not sure if multiple fan controllers can be defined for a SINGLE FAN and then select the maximum RPM out of those controllers to govern a fan.
the reason I think I need it is as follows:
I'm going to use a single loop and my aim is MINIMUM FAN RPM , thus minimum noise, while keeping the computer alive. no overclocking, no maximum cooling, just minimal noise. so I'm thinking that I need to assign three fan controllers for the radiator fans (that would be controlled with the same fan controller profile) with the following profiles:
1. 20% power if GPU temp is up to 40. then increase linearly and max out at GPU temp to 60.
2. 20% power if CPU temp is up to 40. then increase linearly and max out at GPU temp to 60.
3. 20% power if water temp is up to 30. then increase linearly and max out at GPU temp to 40.
averaging the tempeatures is dangerous, for example, assume that water and GPU are cool enough yet my CPU is under load and it can use more cooling yet controllers, using the average temp as the gague, thinks it is providing enough fan RPM. selecting the maximum temp would be dangerous, too. assume for example a situation which GPU has the highest temp ouf of the three and it is around 50, so fan is not pushed to max. yet who knows, maybe water is already passed over designated 40 yet controller is using the higher temperature, in this case, GPU, as its guague.
(I hope I could have described the sitation clear enogh. :rolleyes: )

so bottom line: is it possible to define more than one fan controller and use the max of those to drive fans?

any ideas is appreciated,
cherers :D

Jeremy

Junior Member

RE: maximum fan output selection for controller?

Wednesday, January 30th 2013, 2:01am

so bottom line: is it possible to define more than one fan controller and use the max of those to drive fans?

any ideas is appreciated,
cherers :D
I'm a bit of a rookie with the system myself, but I don't believe what you're describing is possible. Given your goals, if you have an accurate way to measure the temperature of your coolant (I believe Aqua sells a product that ties into the Aquaero 5), I would base fan speeds off of that directly as that determines the temps that the components will reach. Also, keep in mind that water changes temperature more slowly than air cooled solutions do, so you'll probably want to be moderately aggressive in setting the temperature at which fan speeds reach 100%. Given that you have no desire to overclock, I would suggest 100% fan speed at a water temp of 35-40 deg C.

It'll take some experimentation on your part.

RE: RE: maximum fan output selection for controller?

Wednesday, January 30th 2013, 8:00am

I would base fan speeds off of that (water) directly as that determines the temps that the components will reach.
Thanks a bunch for your reply! Yet I think that based on the fact that GPU or CPU blocks will have a certain thermal resistance, as they inject heat inside the water, they create a temperature gradient. in other words, CPU and/or GPU temperature is not necessarily the same as the water. I think it might be anything between 5-25 degrees higher. so basing your fan power on just the water temperature seems a little bit risky.
P.S: speedfan has this option to select the maximum RPM if I still recall how it used to work ....
:?:



cheers ,
8)

cc01

Full Member

Wednesday, January 30th 2013, 8:42am

Basing your fan speed on coolant temp is exactly how you want to do it. Your coolant is the medium by which you are effecting heat transfer, so you want to keep the coolant temperature as low as as possible. Heat is not the same thing as temperature. Heat is the amount of energy transferred from one thing to another because of a difference in temperature, (ie from a hot object to a cold object) so to maintain the efficiency of your loop to cool your components you want to maintain as great a difference in temperature between those components and the coolant. The second heat transfer occurs at your radiator, so you also want to keep the ambient air temperature as low as comfortably possible without feeling like you are sitting inside a refrigerator.

This post has been edited 2 times, last edit by "cc01" (Jan 30th 2013, 8:48am)

EnigmaG

Senior Member

Wednesday, January 30th 2013, 8:43am

No GPU/CPU temperature jumping, no load water +ca 2° max load water +ca 15-20°(it is a constant value, it does´t change) and back in a blink,RPM has no effect at this .
The water warmed-up if the heat input higher as the output, slowly. Delta 100 Watt 1 minute 1 liter 1,5°Kelvin (the same the other way around)
So u have plenty of time to rise the RPM if the water temp are raising

jpmboy

Full Member

Wednesday, January 30th 2013, 3:15pm

It would be a mistake to link fan speed to cpu or gpu temperature since the fluctuation is fast - although there are ways to "dampen" the effect on the controller. Use water temp to control rad fan speed.
AsRock E3Gen3, 2700k @4.6 with cuplex HF, 2 HD7970s with aquaC waterblocks, 16G GSkill 2133, TJ09, ST1500 ps, plextor 256 ssd, 2x1TB WD VRs raid 1, HP 30 inch. Aquacomputer 720XT Mk IV.

Saturday, February 2nd 2013, 8:55am

Wow that is a really complicated system you describe and I can't see any benefit to it.
The Aquaero controls the speed of fans ... which cool the water ... as long as the water is kept to a low temperature the cooling effect on the water blocks is the same. The CPU and GPU temps will jump up and down with varying load but water temperature will stay relatively stable and change slowly.
The best way to control your cooling is to base the speed of your fans on the difference between the ambient air temperature and the temperature of your water. You do this by creating a virtual sensor.
You then apply that virtual sensor to a curve controller or multiple controllers so that all fan speeds climb as the water to air delta ( temp difference) climbs.
That way the fans stay at low speed when your computer is idling and speed up slowly when you put it under load. They will also then stay at higher speeds til your coolant has dropped back down in temperature after you stop using your PC.

Basing your fan speeds just on water temperature would mean your fans will speed up if the air temperature rises even though it would give no benefit.

This post has been edited 2 times, last edit by "Jakusonfire" (Feb 2nd 2013, 9:05am)

thickskull

Junior Member

Sunday, February 3rd 2013, 3:22pm

I just got my Aquaero 5 LT up and running and use water temp to control fan speed. One rad the fan stays on and ramps up with temp and the two fans on my other rad don't start untill the water temp gets over 28C and then ramps up from their. I don't like using air to water delta because if the room warms up my cpu will be hotter and may not handle my overclock. I have three inline temp sensors and may use the delta of two of them to ramp up the fan that stays on.

jpmboy

Full Member

Sunday, February 3rd 2013, 11:44pm

Wow that is a really complicated system you describe and I can't see any benefit to it.
The Aquaero controls the speed of fans ... which cool the water ... as long as the water is kept to a low temperature the cooling effect on the water blocks is the same. The CPU and GPU temps will jump up and down with varying load but water temperature will stay relatively stable and change slowly.
The best way to control your cooling is to base the speed of your fans on the difference between the ambient air temperature and the temperature of your water. You do this by creating a virtual sensor.
You then apply that virtual sensor to a curve controller or multiple controllers so that all fan speeds climb as the water to air delta ( temp difference) climbs.
That way the fans stay at low speed when your computer is idling and speed up slowly when you put it under load. They will also then stay at higher speeds til your coolant has dropped back down in temperature after you stop using your PC.

Basing your fan speeds just on water temperature would mean your fans will speed up if the air temperature rises even though it would give no benefit.



Excellent point! I'm gonna try your water-air temp delta controller suggestion.

This post has been edited 1 times, last edit by "jpmboy" (Feb 3rd 2013, 11:47pm)

AsRock E3Gen3, 2700k @4.6 with cuplex HF, 2 HD7970s with aquaC waterblocks, 16G GSkill 2133, TJ09, ST1500 ps, plextor 256 ssd, 2x1TB WD VRs raid 1, HP 30 inch. Aquacomputer 720XT Mk IV.

EnigmaG

Senior Member

Monday, February 4th 2013, 8:20am

My advice. Keep it simple, use water temp.

jpmboy

Full Member

Tuesday, February 5th 2013, 12:03am

My advice. Keep it simple, use water temp.


Yea - i set up ased on water- air temp and the fans run for a. Bery long time trying to match the two temps even with a startup temp of 2oC difference. It works, maybe a bit of tweaking with a "shut off fans" delta T would be good. But.... I went back to water temp.

Jacksonfire's suggestion is a good one, with a start and stop delta T defined. AS2012 has start, but not stop parameters.
AsRock E3Gen3, 2700k @4.6 with cuplex HF, 2 HD7970s with aquaC waterblocks, 16G GSkill 2133, TJ09, ST1500 ps, plextor 256 ssd, 2x1TB WD VRs raid 1, HP 30 inch. Aquacomputer 720XT Mk IV.

Jeremy

Junior Member

Tuesday, February 5th 2013, 5:32am

AS2012 has start, but not stop parameters.
0% on the fan curve = 0 volts = stop.

The biggest issue I have with using dT to determine fan speed is that the system won't compensate at all for changes in ambient. If your ambient is pretty steady year round this isn't an issue, but if your ambient is 65 deg F in the winter and 85-90+ in the summer like it is for me you might want to consider using straight water temp instead of dT. Sure, it'll be noisier in the summer, but that's because it needs more cooling when ambient is high. To me, that's acceptable.

This post has been edited 1 times, last edit by "Jeremy" (Feb 5th 2013, 5:35am)

thickskull

Junior Member

Tuesday, February 5th 2013, 6:33am

AS2012 has start, but not stop parameters.
0% on the fan curve = 0 volts = stop.

The biggest issue I have with using dT to determine fan speed is that the system won't compensate at all for changes in ambient. If your ambient is pretty steady year round this isn't an issue, but if your ambient is 65 deg F in the winter and 85-90+ in the summer like it is for me you might want to consider using straight water temp instead of dT. Sure, it'll be noisier in the summer, but that's because it needs more cooling when ambient is high. To me, that's acceptable.
Their is a startup temp in the curve controller just don't check the box that says hold minimum power in the fan setup. I set the curve controller to 0% power lower than the startup temp I wanted just to make sure.

Tuesday, February 5th 2013, 9:58am

Why would you ever want your fans to stop?

The air / water delta will never be zero so depending on your system capabilities you set a curve controller ... mine is set to max the fans at a 10C delta and fans 0% at -1C delta

The system automatically balances itself to about a 4.5C delta at idle and 1000RPM and the load delta is about 7.5C at just over 1500RPM

If you just use water temperature then you have to keep adjusting everything depending of the temperature of the room. If you have to do that then what is the point of using an expensive system like the Aquaero. You can connect a water sensor to any cheap fan controller. The whole point of using air / water deltas is that it doesn't make any difference what the air temperature is ... You always get the best possible cooling, any more is a waste of noise.

I thought this was the reason people bought these systems, its not that complicated.




Air in temperature stays the same ... water temp goes up so Air/water delta goes up so fans speed up.

This post has been edited 4 times, last edit by "Jakusonfire" (Feb 5th 2013, 10:28am)

mandrix

Full Member

Tuesday, February 5th 2013, 1:35pm

It took me a while but I eventually came around to operating on air/water Deta t. However I can see situations where one might want to change fan speeds based on other than water temperature.
Some motherboard components such as pch can run pretty hot, then there might be consideration for other equipment such as PA2's being undervolted. Fortunately I have a pretty roomy case/pedestal system so airflow is decent.

Do some of you actually see 10c delta between air and water? I don't think I've ever seen over 3c difference, and I do quite a bit of overclocked benching. That is what finally brought me around to using air/water delta input for curve controllers as hearing the fans ramp up constantly was quite noisy, and unneeded.
I will admit I have overkill on radiators, though, but they are doing exactly as I planned. Less fan rpm and a quieter machine overall.

Jeremy

Junior Member

Tuesday, February 5th 2013, 3:09pm

Why would you ever want your fans to stop?
In my particular case, the top fan is weird. I get much better idle temps with it on, but much better gaming load temps with it off. So I have it set to turn off when GPU temp reaches a certain point.

You wouldn't need to change the curves based on season if you used water temp instead of dT, but you might need to settle for a higher base dT. Fans start at water temp of 30, ramp to max by water temp 40. If you're benching, that wouldn't work, but most people set fans to 100% for benching runs anyway. Everyone's different, and the Aquaero can support a fair number of different options. That's the beauty of it, I just want to make sure he knows his options.

Wednesday, February 6th 2013, 2:41am

It took me a while but I eventually came around to operating on air/water Deta t. However I can see situations where one might want to change fan speeds based on other than water temperature.
Some motherboard components such as pch can run pretty hot, then there might be consideration for other equipment such as PA2's being undervolted. Fortunately I have a pretty roomy case/pedestal system so airflow is decent.

Do some of you actually see 10c delta between air and water? I don't think I've ever seen over 3c difference, and I do quite a bit of overclocked benching. That is what finally brought me around to using air/water delta input for curve controllers as hearing the fans ramp up constantly was quite noisy, and unneeded.
I will admit I have overkill on radiators, though, but they are doing exactly as I planned. Less fan rpm and a quieter machine overall.

That's the beauty of having giant cases with pedestals and separate sections or the like. It allows uninterrupted and unpolluted airflow for great deltas.

In a regular PC case a delta of anything under 10 is considered good.

I have seen lots of people claim very low deltas at all times and I always have to wonder if their sensor setup is really giving accurate results because to achieve that requires an extraordinarily efficient system. Without enormous rads the water temp simply has to be higher according to the laws of thermodynamics ... more heat energy is dispersed at higher temp differences. It takes a much larger surface area to disperse a couple hundred watts at 2C compared to 10C

mandrix

Full Member

Wednesday, February 6th 2013, 2:06pm

It took me a while but I eventually came around to operating on air/water Deta t. However I can see situations where one might want to change fan speeds based on other than water temperature.
Some motherboard components such as pch can run pretty hot, then there might be consideration for other equipment such as PA2's being undervolted. Fortunately I have a pretty roomy case/pedestal system so airflow is decent.

Do some of you actually see 10c delta between air and water? I don't think I've ever seen over 3c difference, and I do quite a bit of overclocked benching. That is what finally brought me around to using air/water delta input for curve controllers as hearing the fans ramp up constantly was quite noisy, and unneeded.
I will admit I have overkill on radiators, though, but they are doing exactly as I planned. Less fan rpm and a quieter machine overall.

That's the beauty of having giant cases with pedestals and separate sections or the like. It allows uninterrupted and unpolluted airflow for great deltas.

In a regular PC case a delta of anything under 10 is considered good.

I have seen lots of people claim very low deltas at all times and I always have to wonder if their sensor setup is really giving accurate results because to achieve that requires an extraordinarily efficient system. Without enormous rads the water temp simply has to be higher according to the laws of thermodynamics ... more heat energy is dispersed at higher temp differences. It takes a much larger surface area to disperse a couple hundred watts at 2C compared to 10C
Yeah I suppose I see very low deltas compared to some. Here I've been fussing around with calibration, and controlling over a 3c increase in water temps under load, lol. My setup has been very efficient at cooling all along. I guess having those two 480mm radiators in the pedestal gives them enough isolation to be really good, then I have a 240mm in the case.
Since I check the calibration of the sensors frequently I'm confident I'm seeing consistent numbers. Add to that I now have delidded my 3770K and that dropped the core temps under load a good bit. I'm very happy with my rig.

Measuring temperature is always subject to the accuracy of the equipment used at a given temperature, though, and fussing over a few C difference probably isn't worthwhile.

cc01

Full Member

Thursday, February 7th 2013, 11:13am

Out of curiosity I have tried a fan controller using a water/air delta instead of water temp only and it works well, although I think I need to fine tune it to my system. Under idle and low loads my system is quieter due to the fans not winding up as much however under load I get regular and pronounced fluctuations of fan speed. Delta doesn't increase beyond 10C and at idle sits around 4 to 5C. I have noticed that using this contoller doesn't work quite as well on hotter days than it does on mild or cooler days, but this is probably due to the inefficiencies of my radiators. I don't have the luxury of numerous and large radiators due to my case being so small, however I've been wanting to upgrade them to something chunkier for some time now ...

Jeremy

Junior Member

Friday, February 8th 2013, 3:01am

however under load I get regular and pronounced fluctuations of fan speed.
I have a few suggestions/notes you might consider that could help. One, how steep is your ramp? If it's too steep, you'll get large/noticeable increases in fan speed with small increases in temp. Two, consider a stepped ramp instead of a pure linear one. 25% for dT from 0-3 deg C, 50% from 3-6, 75% from 6-10, 100% for 10+. Those numbers are just a suggestion, but broad plateaus like that might keep the fans from changing speed so frequently. Constant speeds are generally less noticeable/bothersome because the pitch/noise doesn't change. Also, note the speed/voltage where the fan noise becomes apparent and tune your fan curve around that and only go above that when you really need to. I have one fan that's noticeable at 12 volts but near silent at 10 volts, so I keep it at 10 volts or less most of the time.

Similar threads