Overclock.net banner

1 - 20 of 20 Posts

·
Registered
Joined
·
113 Posts
Discussion Starter #1
I finished my rebuild a few weeks ago and installed Aquacomputer inline temperature sensors (2 per loop). One is located at the exit side of the block (hot reading) and the other is the fluid out of the radiator (cool reading). The Delta between both readings is usually less than1C, why is there no noticeable difference? For reference here's what I'm running:

GPU Loop:
EKWB Nickel plated 1080FTW water block with nickel plated copper backplate

CPU Loop:
EKWB ASUS X99 monoblock Nickel plated copper

Cooling is the same for both loops
EKWB D5 PWM pump
Alphacool 250mm D5 mount w/res
16mm PTEG tubing
Alphacool Nexxos 480x60mm with Corsair ML 120fans in a push/pull

Case: Corsair 1000D

Exhaust fans:
3xML120 up top and 2xML120 at the rear

Ambient room temp is 21.1C

everything is controlled by an Aquero 6XT


Screenshot 1 shows normal operation while running a game for a few hours


Screenshot 2 shows my fans at max speed, notice how both the hot/cold coolant temps stay within the same delta. The arrows in blue point to where I increased the fan speed


What seems to be the issue with my coolant not getting 'cool'?
 

Attachments

·
Registered
Joined
·
1,558 Posts
You usually don't see a Delta in water cooling loops across the same media(the water inside the loop) unless you are dumping a ton of heat into it like cooling 6 1080s that are mining. You just won't see anything past 1-2c difference between any 2 points inside the loop

Water to air Delta can see bigger gaps and that's usually what we mean when we talk about Delta temperatures, all loops reach an equilibrium at some wattage input heat to wattage dissipated and the water won't get any hotter or cooler until one of the 3 things that effect temps change, input heat(the system no longer generating as much heat or is generating more heat), ambient temps(the air going into the rad is cooler or hotter), or heat dissipation(this one can change 2 ways but really only one without changing the loop, increasing or decreasing surface area or increasing or decreasing the fan speeds)
 

·
Registered
Joined
·
113 Posts
Discussion Starter #3
You usually don't see a Delta in water cooling loops across the same media(the water inside the loop) unless you are dumping a ton of heat into it like cooling 6 1080s that are mining. You just won't see anything past 1-2c difference between any 2 points inside the loop

Water to air Delta can see bigger gaps and that's usually what we mean when we talk about Delta temperatures, all loops reach an equilibrium at some wattage input heat to wattage dissipated and the water won't get any hotter or cooler until one of the 3 things that effect temps change, input heat(the system no longer generating as much heat or is generating more heat), ambient temps(the air going into the rad is cooler or hotter), or heat dissipation(this one can change 2 ways but really only one without changing the loop, increasing or decreasing surface area or increasing or decreasing the fan speeds)
Shouldn't the radiators make some difference in the coolant temperature regardless? The heat dissipated into my coolant should be going somewhere and not remaining static through the radiators.
 

·
Registered
Joined
·
164 Posts
Shouldn't the radiators make some difference in the coolant temperature regardless? The heat dissipated into my coolant should be going somewhere and not remaining static through the radiators.
They do, but effectively the temp of the whole loop increases or decreases. Relatively, the water flows so quickly, the loop is so short, and there is so much cooling capacity that the entire loop/loops reaches a temperature that is more or less uniform.

This behavior is totally normal and your temps look great. Don't sweat it.

As somebadlemonade said, monitor your intake air to water temp delta T. That is the more meaningful data.
 

·
Robotic Chemist
Joined
·
3,010 Posts
You have to understand how fast the water is moving through the loop. The water flow rate is fast enough, and its heat capacity high enough, that cooling the water by even <1°C in the radiator moves a LOT of heat (watts) to the air per minute. You should see something around a 1°C rise in the GPU/CPU block and a 1°C drop in the radiator.
 

·
Registered
Joined
·
113 Posts
Discussion Starter #6
Thanks everyone, I appreciate the help. I should know this since I'm an engineer... It's been a long time since I last took a thermo class but I figured the radiator could dissipate heat more effectively despite the high flow rates.

Normally a 1C increase in the air temp correlates to about the same increase in the CPU/GPU temps. There are a lot of restrictions in my loop as well which means even at full speed my pump is fighting back pressure, therefore, the true flow rate is much lower. I'll have to add in a flow sensor in each loop to see what rates I'm moving fluid at for reference. I may have broken something with the PWM on both pumps because I can no longer control their speeds. They both only run at 100% (~4830rpm) which isn't necessarily a bad thing.

Anyway, I probably made myself look dumb because I tend to over think things just like I overdesigned my system for what I need. It's been one of those weeks so excuse my rather questionable thinking here.
 

·
Tetrapyloctomist
Joined
·
1,708 Posts
I probably made myself look dumb
Am dumber than you :D

On a general level, what you could do:
- Make your life simpler. You shouldn't complicate things by adding curves and tables and multiple conditions and variable RPMs.. Fans set, C/G-PU too hot? Fans set higher. The end. The fewer intermediates, the fewer 'steps' any device-related process entails, the slimmer the chances of failure, the better you'll be. Less is more here.
- It's hard to have a flow rate issue unless one uses, i forget the name now, those plug-n-unplug fittings* for each and every block. And even then, still hard to have real flow rate issues. Have you started with the simplest explanation? Something, somewhere, being partially clogged?
- I understand why you've installed so many flow sensors (am guessing so that you can pinpoint the issue) but once set, you should probably remove them. There are two kinds of sensors for this, both have a limited lifetime, both can cause issues. See less is more.
- Are your pumps PWM models? Asking due to the RPM issue you're describing. How are they being controlled (you mention AQ.. curve stuff again, something simpler)? And how are they powered?

* Quick disconnects! That's the name.
 

·
Registered
Joined
·
759 Posts
It depends on the flow rate. With my pumps are full speed, my coolant raises only 0.8C. Lower the pump speed and the coolant wil get warmer after the blocks and cooler in the radiator. With my pump at the slowest speed, my coolant delta is 9-10C. I’m using an EK D5 PWM as well.
 

·
Shooting down fallacies
Joined
·
8,201 Posts
It depends on the flow rate. With my pumps are full speed, my coolant raises only 0.8C. Lower the pump speed and the coolant wil get warmer after the blocks and cooler in the radiator. With my pump at the slowest speed, my coolant delta is 9-10C. I’m using an EK D5 PWM as well.
And even this will vary from build to build, and is typically in no way linear. There is a point in the flow rate where going under it produces a steep loss in cooling performance, while going above it has a very slight incline in gains. In some cases with a D5, running above a certain speed (usually up around 3500-4000 rpm) will actually increase water temps as the heat from the pump becomes more than the gains from the extra flow. That's why it is always best to experiment and see how your build responds to different scenarios.
 

·
Registered
Joined
·
113 Posts
Discussion Starter #10
Am dumber than you :D

On a general level, what you could do:
- Make your life simpler. You shouldn't complicate things by adding curves and tables and multiple conditions and variable RPMs.. Fans set, C/G-PU too hot? Fans set higher. The end. The fewer intermediates, the fewer 'steps' any device-related process entails, the slimmer the chances of failure, the better you'll be. Less is more here.
- It's hard to have a flow rate issue unless one uses, i forget the name now, those plug-n-unplug fittings* for each and every block. And even then, still hard to have real flow rate issues. Have you started with the simplest explanation? Something, somewhere, being partially clogged?
- I understand why you've installed so many flow sensors (am guessing so that you can pinpoint the issue) but once set, you should probably remove them. There are two kinds of sensors for this, both have a limited lifetime, both can cause issues. See less is more.
- Are your pumps PWM models? Asking due to the RPM issue you're describing. How are they being controlled (you mention AQ.. curve stuff again, something simpler)? And how are they powered?

* Quick disconnects! That's the name.



It depends on the flow rate. With my pumps are full speed, my coolant raises only 0.8C. Lower the pump speed and the coolant wil get warmer after the blocks and cooler in the radiator. With my pump at the slowest speed, my coolant delta is 9-10C. I’m using an EK D5 PWM as well.
And even this will vary from build to build, and is typically in no way linear. There is a point in the flow rate where going under it produces a steep loss in cooling performance, while going above it has a very slight incline in gains. In some cases with a D5, running above a certain speed (usually up around 3500-4000 rpm) will actually increase water temps as the heat from the pump becomes more than the gains from the extra flow. That's why it is always best to experiment and see how your build responds to different scenarios.
I like making things complicated, it's more enjoyable to build them/look at :)

Yes, they are both EK D5 PWM pumps. I'm not sure why I can't control them anymore, maybe it's something I did at one point during my various builds. I like having the temps sensor installed so I can see if there's an issue at any given point. Removing both sensors would be a big headache and involve draining 4L of coolant between the two loops and some reworking on a piece or two of PTEG which I don't feel like doing right now.

When I try adjusting the power to each pump in Aquasuite it shows a change in voltage but no change in RPM. The same was true when trying it with the pumps plugged into my mobo. Maybe the RPM readings are simply not showing up accurately? I'm including a picture below.

Here's my build for reference, some areas have changed slightly since I took these pictures. The temperature sensors were not isntalled when these were taken. (refined the look in some areas)
 

Attachments

·
Shooting down fallacies
Joined
·
8,201 Posts
D5 pumps are not voltage controlled. They are designed to see a constant 12V. PWM versions are speed controlled by PWM, Vario by a knob. You had something go wrong somewhere along the line.
 

·
Registered
Joined
·
759 Posts
I like making things complicated, it's more enjoyable to build them/look at :)



Yes, they are both EK D5 PWM pumps. I'm not sure why I can't control them anymore, maybe it's something I did at one point during my various builds. I like having the temps sensor installed so I can see if there's an issue at any given point. Removing both sensors would be a big headache and involve draining 4L of coolant between the two loops and some reworking on a piece or two of PTEG which I don't feel like doing right now.



When I try adjusting the power to each pump in Aquasuite it shows a change in voltage but no change in RPM. The same was true when trying it with the pumps plugged into my mobo. Maybe the RPM readings are simply not showing up accurately? I'm including a picture below.



Here's my build for reference, some areas have changed slightly since I took these pictures. The temperature sensors were not isntalled when these were taken. (refined the look in some areas)


You need to set your pumps as PWM controlled instead of voltage controlled in aquasuite.
 

·
Registered
Joined
·
2,111 Posts
Hi there

Less than 1°C difference between IN and OUT is pretty normal there,in my case this I seen when I run 360mm on top and on bottom 240mm 60mm thick radiator and I placed same Aquacomputer inline sensor after GPUs and OUT has been close to inlet of Reservoir, under or in idle I could see similar thing around 0.5-1°C difference between IN and OUT,under liad load this difference would raise by 1.5°C or more

In current loop with 4*360mm radiators plus MO-ra3 360mm difference between IN and OUT is around 1-1.5°C(loop order look like this Reservoir with D5->1st bottom 360mm Radiator->2nd bottom 360mm radiator - >D5->GPUs in series->MO-ra3 360mm->1st water temperature sensor->3rd 360mm radiator - >4th 360mm radiator->CPU block->water temperature sensor->inlet of Reservoir), flow in my case is around 100LPH

Under load difference between the IN and OUT is around 1.4-2°C as max, maybe bit more but water delta T under load is 2.2-3°C in gaming or rendering

In Aquasuite you need to assign pumps if they're same to one of the controllers, if they're PWM then you should control them easily, assign controller to pumps(not sure which fan channel you are using on Aquaero for pumps), create curve for pumps etc and you can control them


Hope this helps

Thanks, Jura
 

·
Registered
Joined
·
113 Posts
Discussion Starter #14
Any idea why I can't control the pump speed with my Aquaero 6 XT? I read somewhere about their PWM for pumps only working with Aquacomputer ones which doesn't seem right.
 

·
Registered
Joined
·
113 Posts
Discussion Starter #15 (Edited)
Hi there

In Aquasuite you need to assign pumps if they're same to one of the controllers, if they're PWM then you should control them easily, assign controller to pumps(not sure which fan channel you are using on Aquaero for pumps), create curve for pumps etc and you can control them

That was the first thing I did, I posted a screenshot showing them running at max speed despite only seeing ~7V from the XT 6. Here's another full picture below. I'm also attaching an old picture of one of my 2 EK D5 PWM pumps to show it's in fact a PWM unit.

So from what I'm gathering, I can't control my EK D5 PWM pumps with PWM? There's no dial on the bottom like on my previous Alphacool D5 i.e. these are PWM pumps. I have a molex (2 of the 4 pins, GND and 12V) then a 2 of the 4 pin fan header type plug with blue/green wires i.e. RPM and PWM but neither pump will work with the XT 6 via PWM, so am I SOL?
 

Attachments

·
PC Enthusiast
Joined
·
1,075 Posts
IIRC for 1GPM of flow it takes somewhere around 200-250W of heat to raise water temp by 1C, so what you were seeing is normal. If you know the flow rate and the hot and cold temps you can calculate the amount of heat being added to the loop, I've even seen people having this displayed on the front of an aquaero before.

Edit: If you are seeing the voltage going to the pumps drop then you aren't controlling them using PWM, when in PWM mode the voltage should always be 12V. There must be a setting somwhere in there you need to change so it controls by PWM instead of by voltage.
 

·
dontworkSundysormedaysoff
Joined
·
229 Posts
That was the first thing I did, I posted a screenshot showing them running at max speed despite only seeing ~7V from the XT 6. Here's another full picture below. I'm also attaching an old picture of one of my 2 EK D5 PWM pumps to show it's in fact a PWM unit.

So from what I'm gathering, I can't control my EK D5 PWM pumps with PWM? There's no dial on the bottom like on my previous Alphacool D5 i.e. these are PWM pumps. I have a molex (2 of the 4 pins, GND and 12V) then a 2 of the 4 pin fan header type plug with blue/green wires i.e. RPM and PWM but neither pump will work with the XT 6 via PWM, so am I SOL?
Try connecting a pump to motherboard pwm to rule out the xt6
 

·
Shooting down fallacies
Joined
·
8,201 Posts
That was the first thing I did, I posted a screenshot showing them running at max speed despite only seeing ~7V from the XT 6. Here's another full picture below. I'm also attaching an old picture of one of my 2 EK D5 PWM pumps to show it's in fact a PWM unit.
If your XT6 is changing the voltage, then it is not configured for PWM control. A PWM controller will NOT change the voltage. It will always be a constant 12V. Either you have the pumps on headers that are not capable of PWM control, or you have them set to control by variable voltage and not PWM.
 

·
Registered
Joined
·
1,165 Posts
Water is very good at absorbing heat. It takes over 4 watts to heat 1 gram of it by 1c. A normal loop flow rate is 3000 to 4000 grams per minute which is why it takes a lot of heat to increase the temp by 1c (over 250W at 4Lpm)

The same is true at the other end, the radiator/s. The rad can only dissipate as much energy as is put into the system. If 300W is put in for about 1c increase in temp then 300W comes out for the same temp decrease.

The temperature that the rad needs to be over ambient to dissipate the wattage being put in is the system delta.
 

·
Registered
Joined
·
759 Posts
That was the first thing I did, I posted a screenshot showing them running at max speed despite only seeing ~7V from the XT 6. Here's another full picture below. I'm also attaching an old picture of one of my 2 EK D5 PWM pumps to show it's in fact a PWM unit.

So from what I'm gathering, I can't control my EK D5 PWM pumps with PWM? There's no dial on the bottom like on my previous Alphacool D5 i.e. these are PWM pumps. I have a molex (2 of the 4 pins, GND and 12V) then a 2 of the 4 pin fan header type plug with blue/green wires i.e. RPM and PWM but neither pump will work with the XT 6 via PWM, so am I SOL?
You can control EK D5 pumps with the Aquaero. I'm doing this as well.
I can see from the controller screenshots that you're pumps are setup as voltage controlled instead of PWM controlled in Aquasuite. This doesn't work with the EK D5. Only PWM control works.

To adjust this, go to the FANS tab in Aquasuite, select the header of your pump and check PWM control instead of voltage control. Now you will able to adjust your pump speed from the controller screen and you should see always 12V then.
 
1 - 20 of 20 Posts
Top