Experimental Post Rad chiller Concept

[aigomorla]

Hey guys...

i wanted to give my friend some credit whose building 2 blocks for me.

Its a post rad chiller, something that goes after the radiator to boost your water performance.

This unit should either do 1 of 2 things:

1. Bring the overall system lower in heat by removing it, hence stress the radiator less.
2. Bring the coolant slightly below ambient so i get better cooling performance.

http://www.xtremesystems.org/f...howthread.php?t=172946


This is whats been taking up all my time. And martin is just awesome. :]


Anyhow those who like water will be very interested in this.

 

[ChefJoe]

I think you should stick a thermometer into the reservoir of any of your current water cooling setups and see how hot the coolant gets under load. That'll help you to figure out how much of an effect water temperature has compared heatsink/TIM/spreader layers.

 

[aigomorla]

Originally posted by: ChefJoe
I think you should stick a thermometer into the reservoir of any of your current water cooling setups and see how hot the coolant gets under load. That'll help you to figure out how much of an effect water temperature has compared heatsink/TIM/spreader layers.

Martin is doing all the testing for me while he builds. My idea actually interested him so he offered the time to do a complete detailed test.

He has awesome eq for a good test as well.

And i already have temp probes in my reservoirs. LOL... you should expect that from me.

http://i125.photobucket.com/al...aigomorla/IMG_0716.jpg

Those little black wires are the line feeds for the probes. :]

 

[DerwenArtos12]

So martin is going to use that 1300W heater coil to test the capacity of your chiller? Nice.

 

[aigomorla]

Originally posted by: DerwenArtos12
So martin is going to use that 1300W heater coil to test the capacity of your chiller? Nice.

LOL... martin is very resourceful.

If you look at the liquid cooling section, you'll see a lot of tests conducted by him, with very good data and results.

:T

 

[DerwenArtos12]

Originally posted by: aigomorla

Originally posted by: DerwenArtos12
So martin is going to use that 1300W heater coil to test the capacity of your chiller? Nice.

LOL... martin is very resourceful.

If you look at the liquid cooling section, you'll see a lot of tests conducted by him, with very good data and results.

:T

I have. Wasn't it martin that had the mini-tutorial on using a water heater coil and some PVC tubing to make a load tester? I'll have to find it again. It was a great idea.

 

[Rubycon]

Why even use a radiator in this case? Connect everything to a Koldwave unit or Neslab Chiller and you're good to go!

 

[DerwenArtos12]

Originally posted by: Rubycon
Why even use a radiator in this case? Connect everything to a Koldwave unit or Neslab Chiller and you're good to go!

because those are both really extremely expensive and this is a DIY project that he's trading with the milling guy for the labor and he's using the radiator to get things down to a few degrees above ambient and then this with three 89W TEC's to get at or just below ambient. It's specifically being designed so that it can use higher watt TEC's but, for this project he's not trying to get sub-ambient, though he may at idle. He's actually using a pair of radiators, one to get the main loop down and one to cool the hotside of the TEC's. He's basically going to sandwich the TEC's between two of the exchangers and run his main loop through the cold side and a secondary loop through the other. The guys over at XS have figured with this set-up he should be able to keep at ambient at load; which, as I understand it, is the exact point.

 

[aigomorla]

Originally posted by: Rubycon
Why even use a radiator in this case? Connect everything to a Koldwave unit or Neslab Chiller and you're good to go!

i estimate this.

my system is pulling near 400-500W on everything. That means my TEC would need to remove that much more heat.

So 400-500W tec draw... only an estimate would require around roughly 700-800W of energy to remove.

Ummm.... thats a lot of power.


By having the rad up front do most of the cooling for me, i only need to drop a small amount of tecs after it. Roughly around 91Wx3 at a lower downvolt setting to around 60W. 180W removed at the cost of ~ 230-250W??

this number is something i can handle.


Also the radiator is a safety net. Meaning cuz of it i can never take the fuild below condensation level. Also, if my TEC's go out, theres no worry about my hostide or coldside getting out of control.

 

[imported_dinGLeBeRRy]

Originally posted by: aigomorla

Originally posted by: Rubycon
Why even use a radiator in this case? Connect everything to a Koldwave unit or Neslab Chiller and you're good to go!

i estimate this.

my system is pulling near 400-500W on everything. That means my TEC would need to remove that much more heat.

So 400-500W tec draw... only an estimate would require around roughly 700-800W of energy to remove.

Ummm.... thats a lot of power.


By having the rad up front do most of the cooling for me, i only need to drop a small amount of tecs after it. Roughly around 91Wx3 at a lower downvolt setting to around 60W. 180W removed at the cost of ~ 230-250W??

this number is something i can handle.


Also the radiator is a safety net. Meaning cuz of it i can never take the fuild below condensation level. Also, if my TEC's go out, theres no worry about my hostide or coldside getting out of control.

yea! rubycon sounds like he cant afford a rad thats why. dont go diggin in your car for parts now.... hahha

 

[imported_dinGLeBeRRy]

Could you sub H2o for anti freeze? what would happen there??

 

[aigomorla]

Originally posted by: dinGLeBeRRy

yea! rubycon sounds like he cant afford a rad thats why. dont go diggin in your car for parts now.... hahha

Sorry this is my Admin acting gay. :X


And im not kidding about him being my admin.

 

[Rubycon]

TE coolers have the advantage of not being position sensitive and rugged. They also make no noise. Efficiency on the other hand, is poor when you get into that power range. That is why I mentioned the chiller.

You could always use a window shaker to produce cooler air and channel that through your radiators.

 

[aigomorla]

Originally posted by: Rubycon
TE coolers have the advantage of not being position sensitive and rugged. They also make no noise. Efficiency on the other hand, is poor when you get into that power range. That is why I mentioned the chiller.

You could always use a window shaker to produce cooler air and channel that through your radiators.

ahhhh if you mean phase chillers.. theres tons of issues that come with that.

i need to insulate the board, the lines, and anything else that can condensate.

Too much work. There isnt much gain from phase to high end water. About another 10% or so. I dont want the full 10% maybe somewhere around 5% would be nice.

 

[Rubycon]

Yes a chiller working on the refrigeration cycle. Does not mean you're cooling glycol loops to -20°C. (although you could for crazy cocks!)

Use an enthalpy controller and adjust a safety margin and you will never see condensate yet the lowest possible temps without the hassle of insulating everything on the cold side.

 

[imported_dinGLeBeRRy]

Originally posted by: Rubycon
Yes a chiller working on the refrigeration cycle. Does not mean you're cooling glycol loops to -20°C. (although you could for crazy cocks!)

Use an enthalpy controller and adjust a safety margin and you will never see condensate yet the lowest possible temps without the hassle of insulating everything on the cold side.

that would suck if you got condensation in your gpu.... ughhh.......

 

[Rubycon]

Originally posted by: dinGLeBeRRy

that would suck if you got condensation in your gpu.... ughhh.......

Of course it would but with proper safeguards in place it would never happen.

 

[aigomorla]

Originally posted by: Rubycon
Yes a chiller working on the refrigeration cycle. Does not mean you're cooling glycol loops to -20°C. (although you could for crazy cocks!)

Use an enthalpy controller and adjust a safety margin and you will never see condensate yet the lowest possible temps without the hassle of insulating everything on the cold side.

LOL rubycon....

Try to say Enthalpy controller 20 times fast without stopping.


The controler would be very nice if i knew how to build one.

 

[DerwenArtos12]

Did dingle come to AT just to bug you?

 

[Rubycon]

Originally posted by: aigomorla

LOL rubycon....

Try to say Enthalpy controller 20 times fast without stopping.


The controler would be very nice if i knew how to build one.

10 speak "enthalpy controller"
20 goto 10



You can purchase one and tinker with it or build your own. Choosing the humidity sensor is the trickiest part. Gold leaf capacitance transducers (say that 10x ) are responsive, easy to wire and reasonably durable. Most mid grade outdoor hygrometers are based on such transducers.

It's also possible to control condensation with the approach method based on conductance of surfaces. The design and build is easier however preventive maintenance, monitoring, and upkeep is high with such a system. Heck you're watercooling so that fits right in!

 

[aigomorla]

Product is done.

Martin has both blocks made. Pictures are up at XS.

If by some reason a guy name NaeKuh POSTS on this post, he's not me/creator of this project.

Anyhow he should be loading some heat on it, he already calculated the pressure drop.

Well see if this idea works. If it does, i have a feeling he's gonna get shoved with orders for it on xs. :T

 

[tylerdustin2008]

Originally posted by: aigomorla

Originally posted by: Rubycon
Why even use a radiator in this case? Connect everything to a Koldwave unit or Neslab Chiller and you're good to go!

i estimate this.

my system is pulling near 400-500W on everything. That means my TEC would need to remove that much more heat.

So 400-500W tec draw... only an estimate would require around roughly 700-800W of energy to remove.

Ummm.... thats a lot of power.


By having the rad up front do most of the cooling for me, i only need to drop a small amount of tecs after it. Roughly around 91Wx3 at a lower downvolt setting to around 60W. 180W removed at the cost of ~ 230-250W??

this number is something i can handle.


Also the radiator is a safety net. Meaning cuz of it i can never take the fuild below condensation level. Also, if my TEC's go out, theres no worry about my hostide or coldside getting out of control.

If you try to cool down all the heat in the system, every water block is going to start to condensate. And you don't want that, because that means everything will need to be insulated.

So you want to put to much of a TEC on there.

 

[aigomorla]

Originally posted by: tylerdustin2008

Originally posted by: aigomorla

Originally posted by: Rubycon
Why even use a radiator in this case? Connect everything to a Koldwave unit or Neslab Chiller and you're good to go!

i estimate this.

my system is pulling near 400-500W on everything. That means my TEC would need to remove that much more heat.

So 400-500W tec draw... only an estimate would require around roughly 700-800W of energy to remove.

Ummm.... thats a lot of power.


By having the rad up front do most of the cooling for me, i only need to drop a small amount of tecs after it. Roughly around 91Wx3 at a lower downvolt setting to around 60W. 180W removed at the cost of ~ 230-250W??

this number is something i can handle.


Also the radiator is a safety net. Meaning cuz of it i can never take the fuild below condensation level. Also, if my TEC's go out, theres no worry about my hostide or coldside getting out of control.

If you try to cool down all the heat in the system, every water block is going to start to condensate. And you don't want that, because that means everything will need to be insulated.

So you want to put to much of a TEC on there.

Radiator is safety net taylor. The rad will pull the sub ambient water back to ambient if the tec's pull out too much.

I dont think i have enough TEC power to overpower my PA also. So i think im safe.

 

[Nemesis 1]

Originally posted by: Rubycon
Why even use a radiator in this case? Connect everything to a Koldwave unit or Neslab Chiller and you're good to go!

Exactly. Using a rad on the same loop as chilled water makes NO-sense. You would be heating the chilled wayer going threw the rads. Not good.


Sorry didn't read far enough thout you were talking TEC chiller. If your going to use a TEC as a chiller . You really need to look at Temp controllers. Say you want the cold side loop at 10C that should be safe in a Air conditioned space from condensation. Plus once your loop is @ 10C it requires less power to maintain it. Controllers are the only way to go. and their expensive. TO cool 400 watts of heat energy requires a 800watt heat side of TEC to be cooled threw rad. I just can't see any way that using a rad in the cooling loop can be as good as just plain ambient air. Without the chillers.

 

[aigomorla]

Originally posted by: Nemesis 1

Originally posted by: Rubycon
Why even use a radiator in this case? Connect everything to a Koldwave unit or Neslab Chiller and you're good to go!

Exactly. Using a rad on the same loop as chilled water makes NO-sense. You would be heating the chilled wayer going threw the rads. Not good.


Sorry didn't read far enough thout you were talking TEC chiller. If your going to use a TEC as a chiller . You really need to look at Temp controllers. Say you want the cold side loop at 10C that should be safe in a Air conditioned space from condensation. Plus once your loop is @ 10C it requires less power to maintain it. Controllers are the only way to go. and their expensive. TO cool 400 watts of heat energy requires a 800watt heat side of TEC to be cooled threw rad. I just can't see any way that using a rad in the cooling loop can be as good as just plain ambient air. Without the chillers.

because instead of drawing that much power off the wall, the radiator will bring coolant levels down to near ambient. After that the TEC works and brings it down even further. Im not pushing a lot of voltage on these. 91x3W is most likely enough to bring me slightly into sub ambient cooling territory.

The radiator helps me avoid overshooting my TEC's. And also helps me avoid condensation all together.