How do you control your fans/ pump. Advice plz.

Joined
Feb 6, 2013
Messages
786
I am about to start my second hardline build, I bought a pwm d5 from singularity. I will have 6 fans total. I understand that aquewro has the best software, so I keep reading. However, I have a Lianli pc011 white case(the old one because I diont like the look of the dynamic) and I dont want to use a bay controller. I would prefer to control through windows. I would like the system to start at full speed obviously through bios and then settle to be temp controlled. In the past I have just bought vario pumps and set it to 4. I appreciate any advice.
 
I am about to start my second hardline build, I bought a pwm d5 from singularity. I will have 6 fans total. I understand that aquewro has the best software, so I keep reading. However, I have a Lianli pc011 white case(the old one because I diont like the look of the dynamic) and I dont want to use a bay controller. I would prefer to control through windows. I would like the system to start at full speed obviously through bios and then settle to be temp controlled. In the past I have just bought vario pumps and set it to 4. I appreciate any advice.


The Aquacomputer Aquaero CAN be a bay controller, but there are cheaper models that sell you just the controller without the screen and you can mount it anywhere in your case. I have the screen model, and it really isn't that useful. I barely ever look at it. All of my setup, control and most of my monitoring, I do via the software.

The primary control for Aquaero units comes via their Aquasuite software. The software communicates with the device, but once set up, the device runs independently (as long as you don't use motherboard/chip sensors for temp control, which you don't want to do with a water loop anyway. It's always best to control fan speed based on loop water temp)
 
As an Amazon Associate, HardForum may earn from qualifying purchases.
I apreciate the information. So that is a 100 dollar add on to my loop. What accessories would I also have to buy to go with it?
 
Thanks again for the info. I have decided to use the motherboard to control it all, with the pump set to max.
 
Thanks again for the info. I have decided to use the motherboard to control it all, with the pump set to max.

Does your motherboard have a temp sensor input?

Some do, and if so that is a good alternative way to control fan speed.

Then, just buy an in-line water temp sensor (these are fairly cheap), place it in your loop, connect it to the motherboard, and use that temp to control fan speed.

It is not very good to use CPU core temp to control fan speed in any water loop (not even an AIO)

When you do this, if the system is idle for a while, since the CPU's are relatively cool, the fans slow down. This allows the water in the loop to heat up. Then, if the load suddenly increases on the CPU, the CPU starts heating up, the fans spin faster, but since there is already so much hot water in the loop, your CPU (or GPU) spikes higher than intended until the fans can remove all that excess heat from the water and reach a steady state.

The way to do it is control the whole loop based on target water temp. When you do this, the problem above doesn't occur. You can figure out your target loop temp by maxing your fans, and loading up the CPU with some tool like Prime 95, then let it reach a steady state where everything is heat soaked. Now note your CPU temp, and your water temp. Subtract the water temp from the CPU temp to find your delta over your water temp.

This is mostly constant give then same amount of CPU load.

Now determine the target temp for the CPU at max load, subtract out that delta T, and set that number as your water loop target temp.
 
  • Like
Reactions: Dan_D
like this
I'd just get a PWM hub if you have too many fans to connect to your board. Silverstone has a cheap one which I've used a few times for 2-3 fans and worked fine.
 
I'd just get a PWM hub if you have too many fans to connect to your board. Silverstone has a cheap one which I've used a few times for 2-3 fans and worked fine.

I do what this guy does. Get a pwm hub, plug the fans into it and the hub to your mobo. Be sure your board can support the amps if you add more than 5 to the hub.
 
I vote for the Aquaero as well. However, if $100 is a little much to justify, Aquacomputer makes a "baby Aquaero" called the Quadro that's a bit more reasonable at about $45. I don't own one but from what I've read and seen it uses the same software and is capable of most the things needed for a typical simple loop.
 
I do what this guy does. Get a pwm hub, plug the fans into it and the hub to your mobo. Be sure your board can support the amps if you add more than 5 to the hub.

I use the Silverstone 8 port PWM hubs. They have a molex connector, so you don't actually pull any power off of the 4-pin header you plug it into. It just shares the pin4 pwm signal to all the fans.

Only downside with this approach is that you only get one readable fan RPM signal. You have to trust that the other fans on the splitter are doing their thing. This is kind of unavoidable with modern large systems with many fans though.
 
I run a D5 PWM , three 120 mm PWM fans and one 140 mm DC fan all via my motherboard software (ASUS AI Suite III) and have had pretty good results. I've daisy-chained the three radiator fans (Fractal Design Venturi HP-12s) and connected them to a single mobo header and run the D5 off of the PSU with the PWM cable connected to a dedicated pump header on the motherboard. I currently control my radiator fan speed based on the CPU temp and have been satisfied with the results.

I can see where Zarathustra[H] is coming from with his recommendation, but controlling your fan speeds with CPU temp or water temp honestly isn't going to make a difference in the end. If you are effectively pre-heating your loop with low fan speeds set by low CPU temps, all that means is that the CPU temp will ramp up faster and you'll reach steady state more quickly. This could possibly negatively impact the boost behavior of some CPUs, but I don't see any other downsides. The main benefit I could see to going with this method would be to smooth out temperature transients from the CPU, stopping your fans and pump from ramping up and down all the time (which could be a pretty big plus for some people.) I may spring for an inline temp sensor and give it a try just because I have an unused T.Probe connection on my motherboard right now and it gives me an excuse to tinker ;).

To hijack the topic a bit, what are people's thoughts on Aquaero and Aquasuite? For someone like me or the OP, what are the benefits to using a system like this as opposed to using the motherboard? Separating the power from the motherboard is nice, but it's awfully expensive if that's the best thing about it.
 
I can see where Zarathustra[H] is coming from with his recommendation, but controlling your fan speeds with CPU temp or water temp honestly isn't going to make a difference in the end. If you are effectively pre-heating your loop with low fan speeds set by low CPU temps, all that means is that the CPU temp will ramp up faster and you'll reach steady state more quickly. This could possibly negatively impact the boost behavior of some CPUs, but I don't see any other downsides. The main benefit I could see to going with this method would be to smooth out temperature transients from the CPU, stopping your fans and pump from ramping up and down all the time (which could be a pretty big plus for some people.) I may spring for an inline temp sensor and give it a try just because I have an unused T.Probe connection on my motherboard right now and it gives me an excuse to tinker ;).

To hijack the topic a bit, what are people's thoughts on Aquaero and Aquasuite? For someone like me or the OP, what are the benefits to using a system like this as opposed to using the motherboard? Separating the power from the motherboard is nice, but it's awfully expensive if that's the best thing about it.
Using water temp to control your fans (and specifically the water/ambient ∆t) is the better way to do it because your fans aren't cooling the CPU anymore - they're cooling the water. And they can't cool the water any cooler than the ambient air they're forcing through the rads. So, if the question is "how fast should I run my fans?", the answer is a water-cooled computer is always "How much hotter is my coolant than the ambient air?"

To segue into your second question, the ability to do what I just described above is what sets the Aquaero apart from typical mobo fan control. You simply cannot do in most BIOSes what the Aquaero can do. Even on high end motherboards with temp in headers, my experience is that the manufacturer assumes you're going to be controlling a fan or fans on temperature readings from something like a CPU or GPU heatsink, where 30°c is idle and 80°c is hot. I've never been able to adjust the curves to a watercooling scenario, where 20°c is idle and 30°c is hot - thus those headers (in my limited experience with them) are useless.

There are lots of other reasons the Aquaero is better: it's completely standalone meaning your OS doesn't need to be running for it to work like other software-based solutions. It's so capable that you're unlikely to need to use motherboard fan headers ever again; you can bring the Aquaero with you from system to system. It sends a fake tach signal to your motherboard's CPU_FAN header, and you can configure what makes that tach signal stop. What if your pump is still running but there's no flow? Add a flow meter to the Aquaero and you can actually protect your PC from this condition. It's got audible alarms you can configure. You can do a variety of different styles of fan and pump control: "curve" control like we're all familiar with, on/off control, even PID control like is used in industrial loop controllers. It's expandable too - you can slave other Aquabus devices to it to get more I/O options.

There's just so, SO much more there than you'll find on motherboards or in most software. I'll explain how I've got mine configured if you're interested.
 
  • Like
Reactions: drklu
like this
Using water temp to control your fans (and specifically the water/ambient ∆t) is the better way to do it because your fans aren't cooling the CPU anymore - they're cooling the water. And they can't cool the water any cooler than the ambient air they're forcing through the rads. So, if the question is "how fast should I run my fans?", the answer is a water-cooled computer is always "How much hotter is my coolant than the ambient air?"

To segue into your second question, the ability to do what I just described above is what sets the Aquaero apart from typical mobo fan control. You simply cannot do in most BIOSes what the Aquaero can do. Even on high end motherboards with temp in headers, my experience is that the manufacturer assumes you're going to be controlling a fan or fans on temperature readings from something like a CPU or GPU heatsink, where 30°c is idle and 80°c is hot. I've never been able to adjust the curves to a watercooling scenario, where 20°c is idle and 30°c is hot - thus those headers (in my limited experience with them) are useless.

There are lots of other reasons the Aquaero is better: it's completely standalone meaning your OS doesn't need to be running for it to work like other software-based solutions. It's so capable that you're unlikely to need to use motherboard fan headers ever again; you can bring the Aquaero with you from system to system. It sends a fake tach signal to your motherboard's CPU_FAN header, and you can configure what makes that tach signal stop. What if your pump is still running but there's no flow? Add a flow meter to the Aquaero and you can actually protect your PC from this condition. It's got audible alarms you can configure. You can do a variety of different styles of fan and pump control: "curve" control like we're all familiar with, on/off control, even PID control like is used in industrial loop controllers. It's expandable too - you can slave other Aquabus devices to it to get more I/O options.

There's just so, SO much more there than you'll find on motherboards or in most software. I'll explain how I've got mine configured if you're interested.

I don't personally like using the Water/Ambient delta for fan control. If you do, on hotter days your system will run hotter than on cooler days.

I like to have my CPU and GPU be cooled to the same target temp regardless of ambient.

I use 33C as that seems to be the loop temp that reliably results in below 40C load temps on my overclocked Pascal Titan X.

On hotter days the fans spin faster, and on cooler days they spin slower. It's never going to be close enough to 33C ambient in my office when I am using it that my fans can't maintain that temp, silently at idle and with varying degrees of fan noise at load.
 
  • Like
Reactions: drklu
like this
I don't personally like using the Water/Ambient delta for fan control. If you do, on hotter days your system will run hotter than on cooler days.

I like to have my CPU and GPU be cooled to the same target temp regardless of ambient.

I use 33C as that seems to be the loop temp that reliably results in below 40C load temps on my overclocked Pascal Titan X.

On hotter days the fans spin faster, and on cooler days they spin slower. It's never going to be close enough to 33C ambient in my office when I am using it that my fans can't maintain that temp, silently at idle and with varying degrees of fan noise at load.
I understand that reasoning. My computer's space is very well climate controlled, so my results are pretty consistent.
 
I use a pair of splitty9s for my fans in a thermaltake x71(big sumbitch). 2 corsair ml 120s on rad 1( hwl gtr 240 bottom compartment ) and 1 ml 120 in the open slot blowing into the psu (there is a 360 mount down there). 4 noctua 140mm 2000s. 3 on rad 2(hwl sr2 420-front) 1 on exhaust and 3 noctua 120mm 2000s on rad 3 (xspc r360 OG up top). The s9 controls both pwm 4pin and 3pin fans and does it very well. i dont have any 3pin fans but i do have a pair of ddc 3.1s in the backup rig that are 3 pin. i can control them with good ole speedfan! I dont use fan curves at all. All of the noctua run at 800rpm (silent) except the rear exhaust 140mm i keep that at 1200rpm (audible but not loud) its connected to the mb. The ml120s are loud mfers so i keep them almost completely off until i game or stress the rig. Ive got both of my d5s running off a swiftech hub but generally run them at 100%. All but the exhaust noctua are connected to 1 s9 and the ml120s on the other. I use speedfan for any and all adjustments. Cuz ima miser when it comes to fan controllers/hubs etc...ive got a digital display temp gauge on the bottom front of my res which is really cool(its cool knowing the fluid temp). I could go by water temp but my temps all around are excellent.
The aquaro is badass but its tooo friggin expensive for me. Maybe some day...

Edit- sorry for the wall o text.
 
Using water temp to control your fans (and specifically the water/ambient ∆t) is the better way to do it because your fans aren't cooling the CPU anymore - they're cooling the water. And they can't cool the water any cooler than the ambient air they're forcing through the rads. So, if the question is "how fast should I run my fans?", the answer is a water-cooled computer is always "How much hotter is my coolant than the ambient air?"

We can abstract it out as far as we want (air -> radiator -> water -> waterblock -> IHS -> silicon), but in the end you're still bound by the same requirement of removing the heat generated by the CPU. Heat is removed from the CPU more efficiently at lower coolant temps but is less efficiently transferred to the air. As the coolant temperature rises, this balance tips the other way and eventually your loop reaches a balance no matter what. As long as this is accomplished at reasonable temperatures for the CPU (or GPU), then nothing else is really important. No matter where you choose to measure your temps, the goal is always the same.

To segue into your second question, the ability to do what I just described above is what sets the Aquaero apart from typical mobo fan control. You simply cannot do in most BIOSes what the Aquaero can do. Even on high end motherboards with temp in headers, my experience is that the manufacturer assumes you're going to be controlling a fan or fans on temperature readings from something like a CPU or GPU heatsink, where 30°c is idle and 80°c is hot. I've never been able to adjust the curves to a watercooling scenario, where 20°c is idle and 30°c is hot - thus those headers (in my limited experience with them) are useless.

There are lots of other reasons the Aquaero is better: it's completely standalone meaning your OS doesn't need to be running for it to work like other software-based solutions. It's so capable that you're unlikely to need to use motherboard fan headers ever again; you can bring the Aquaero with you from system to system. It sends a fake tach signal to your motherboard's CPU_FAN header, and you can configure what makes that tach signal stop. What if your pump is still running but there's no flow? Add a flow meter to the Aquaero and you can actually protect your PC from this condition. It's got audible alarms you can configure. You can do a variety of different styles of fan and pump control: "curve" control like we're all familiar with, on/off control, even PID control like is used in industrial loop controllers. It's expandable too - you can slave other Aquabus devices to it to get more I/O options.

There's just so, SO much more there than you'll find on motherboards or in most software. I'll explain how I've got mine configured if you're interested.

Yeah, seems like it's probably huge overkill for me. I can already base my fan speed on whichever temperature sensor I choose on my motherboard and overtemperature protection will save me from any real damage, so I think I will just save the hundred bucks! Well, $90 since I'm going to go ahead play around with an inline temp sensor anyhow.
 
We can abstract it out as far as we want (air -> radiator -> water -> waterblock -> IHS -> silicon), but in the end you're still bound by the same requirement of removing the heat generated by the CPU. Heat is removed from the CPU more efficiently at lower coolant temps but is less efficiently transferred to the air. As the coolant temperature rises, this balance tips the other way and eventually your loop reaches a balance no matter what. As long as this is accomplished at reasonable temperatures for the CPU (or GPU), then nothing else is really important. No matter where you choose to measure your temps, the goal is always the same.
That argument goes out the window as soon as you start cooling more than one thing, though. Basing fan speed on component temperature in a water-cooled system isn't ideal. It certainly works - watercooling is so overkill for modern hardware that you could (and many do) just set their fans at a single speed and forget it - but it's not the logical way to approach it.

I do understand $100 being a tough pill to swallow tho. If you're interested in trying it for less, Aquacomputer also makes a smaller unit called the Quadro: about $45, uses the same software, the board just has a few fewer bells and whistles.[/QUOTE][/QUOTE]
 
That argument goes out the window as soon as you start cooling more than one thing, though. Basing fan speed on component temperature in a water-cooled system isn't ideal. It certainly works - watercooling is so overkill for modern hardware that you could (and many do) just set their fans at a single speed and forget it - but it's not the logical way to approach it.

That's a fair point. Water temp does become the common factor if you're cooling multiple chips on the same loop.

I do understand $100 being a tough pill to swallow tho. If you're interested in trying it for less, Aquacomputer also makes a smaller unit called the Quadro: about $45, uses the same software, the board just has a few fewer bells and whistles.

That might be something worth looking into.
 
Many great responses here. I think the bottom line for me results from the reason i water cool. I like how it looks and like to tinker in between new hardware builds. Basically my loop is always going to be overkill for what I am cooling so I cant justify the expense on a new build. In a year when I get the urge to tinker and between full rebuilds I may find the purchase justifyable.
 
if 100 bucks on a cooler is tough to swallow, i dont think water cooling is going to be a solution for you. Even if you find something cheaper, you're never going to be maintenance free. Even fully sealed setups have a lifespan that's measured in months, not really years between needing some kind of cleaning and refilling otherwise you start seeing a significant reduction in cooling performance. So there's the added cost of that. Plus cheaper systems tend to break more often.

I try and stay with the aio cheap solutions but it seems like they're never good as-is for long. And then I'm dropping a hundred bucks here and a hundred there to make it right. One cannot overstate the attractiveness of just having a normal air cooler when money is a factor.
 
if 100 bucks on a cooler is tough to swallow, i dont think water cooling is going to be a solution for you. Even if you find something cheaper, you're never going to be maintenance free. Even fully sealed setups have a lifespan that's measured in months, not really years between needing some kind of cleaning and refilling otherwise you start seeing a significant reduction in cooling performance. So there's the added cost of that. Plus cheaper systems tend to break more often.

I try and stay with the aio cheap solutions but it seems like they're never good as-is for long. And then I'm dropping a hundred bucks here and a hundred there to make it right. One cannot overstate the attractiveness of just having a normal air cooler when money is a factor.
The Aquaero isn't a cooler...
 
Hah, my bad. Yeah, if you are spending money on a glorified fan controller... Money is not a factor... So disregard.
 
Shit load of radiator, 11 fans running at 800 rpm, why finesse when you can brute force.
That's a reasonable economical effective way to do things!

Full disclosure; I'm an instrumentation and controls engineer, so setting up complicated systems kinda rustles my jimmies! XD
 
I am about to start my second hardline build, I bought a pwm d5 from singularity. I will have 6 fans total. I understand that aquewro has the best software, so I keep reading. However, I have a Lianli pc011 white case(the old one because I diont like the look of the dynamic) and I dont want to use a bay controller. I would prefer to control through windows. I would like the system to start at full speed obviously through bios and then settle to be temp controlled. In the past I have just bought vario pumps and set it to 4. I appreciate any advice.

You have a PWM pump so use the PWM controller on yer mb. I've never needed a 3rd controller for my pumps, always used PWM off the mb. My fans are controlled by a simple fan controller, set at 1 speed, silent. I've been doing it this way forever.
 
if 100 bucks on a cooler is tough to swallow, i dont think water cooling is going to be a solution for you. Even if you find something cheaper, you're never going to be maintenance free. Even fully sealed setups have a lifespan that's measured in months, not really years between needing some kind of cleaning and refilling otherwise you start seeing a significant reduction in cooling performance. So there's the added cost of that. Plus cheaper systems tend to break more often.

I try and stay with the aio cheap solutions but it seems like they're never good as-is for long. And then I'm dropping a hundred bucks here and a hundred there to make it right. One cannot overstate the attractiveness of just having a normal air cooler when money is a factor.

100 dollars alone is not an issue, its another hundred plus accessories on top of the 2000 plus I have just spent on this current build that makes it tough to swallow.
 
I've been out of the custom diy water cooling game for a long time so I didn't realize the hundred dollar aquaero wasn't an aio cooler. Personally, the idea of a hundred dollar fan/pump controller makes me throw up a bit in my mouth. Especially since it functionally overlaps heavily with things you already have in pretty much any motherboard you'd be putting a diy system in. But then again, diy water cooling is often not tied heavily to practicality or necessity, so to each their own. peacock the crap out of what you enjoy.
 
if it changes fan/pump speed based on temps then it does.
Show me a motherboard that can control fans with a PID controller based on the ambient/coolant ∆t as calculated from the average of three air temperature probes and the average of two coolant temperature probes, and then I might concede your point.

I'm not saying motherboard fan control doesn't work, and I'm not saying that level of complexity is even necessary.

However, even the most advanced, watercooling-friendly mobo fan control is absolutely primitive when compared to what the Aquaero and Aquasuite can do. Only people who have never used an Aquaero would say that it "functionally overlaps" motherboard control.
 
functional overlap doesn't require that the two things do things the same convoluted way.

Modern cooling controllers (like those found in the motherboard and other third party devices like the aquaero) adjusts speeds of fans and pumps to minimize power usage / noise / wear and tear so they're not running at some fixed speed all the time. That's the function. How it does it doesn't matter. What sensors it uses to determine how to adjust those speeds doesn't matter.

The result is the same.. Cpu gets cooled to a defined level via a variable cooling system. That's what they both do and that overlap is significant. What lies outside of that overlap is all the additional sensors and eye candy monitors and indpendent controllable outputs and perhaps even resolution of control in the system. But functionally they serve the same purpose and most people would see it as a redundant solution.
 
Stop pushing the Aquaero nonsense then. No one NEEDs it, they want it.
The only thing I'm "pushing" against is this notion that it's basically equal to motherboard fan control. That's a little bit like saying a car and a horse have functional overlap because they'll both get you there eventually, or that a campfire and a modern furnace are the same thing because they both provide heat.

I know no one needs it. It's $100! There's lots of other places to spend that money in your system. OP asked about fan control in a watercooled system, then BrotherMichigan asked what makes the Aquaero special. That's all I was trying to reply to.
 
Last edited:
Stop pushing the Aquaero nonsense then. No one NEEDs it, they want it.

No one NEEDS a water cooling loop either :p

For my particular application my Aquaero was a godsend. I wish I had bought it two years earlier when I first built my loop. It has made everything so much better.

I can now have it fine tuned so I have completely silent fans at idle regardless of room temperatures and still make sure my GPU never hits 40C without constantly messing with it, or having temp overshoots potentially crashing the system.

All of my hacks I tried before to get things to work were just dissatisfying pains in the ass. It wasn't until I went with an Aquaero I was finally happy with my loop.
 
Seems what this place is about. The enthusiast.

No one needs a car, let alone a corvette, or Ferrari. But damn if you can and wNt it. Go for it.

I am buying the aquaero because I want the best and Zara and others have messed around enough to know it’s very nice and makes life easier. So why wAste money with hassle factor and not having the right tool?
 
Back
Top