Chilled Water Cooling Version 3.0

[Silver Surfer]

All I am trying to say is that it is more efficient to cool peltier with chilled water
than cool water with peltier connected to monstrous fan.

Efficiency, and peltier, really do not belong in the same sentence!

You pretty much get what you pay for.

And by the way, my monstrous fans are maximum 21dba.

They are the fans that come on the Thermalright True Spirit 140 Power, look them up yourself, I can hardly hear the thing running.

 

[Lochekey]

So this is a little of topic Silver but I'm hoping not to far.

I'm wondering how will you think the true spirits would chill the loop without the TECs involved. I'm just wondering how they would compare against a traditional radiator.

My supposition is that they work well in your application due to the temperature differential between the TEC and ambient, whereas they would not work well in a traditional setup due to the lower temperature differential between the water and ambient that is typically maintained in a water cooling loop.

I have nothing to back any of this up so it is pure conjecture on my point but was just wondering if you ever tried such a thing.

 

[Silver Surfer]

So this is a little of topic Silver but I'm hoping not to far.

I'm wondering how will you think the true spirits would chill the loop without the TECs involved. I'm just wondering how they would compare against a traditional radiator.

My supposition is that they work well in your application due to the temperature differential between the TEC and ambient, whereas they would not work well in a traditional setup due to the lower temperature differential between the water and ambient that is typically maintained in a water cooling loop.

I have nothing to back any of this up so it is pure conjecture on my point but was just wondering if you ever tried such a thing.

If you had one mounted to a water block it would be possibly 5% of a 120 radiator 10% at the most just because of the size of transfer contact, heat sink base to water block. That percentage is just a flat out guess based on a past occurrence. I am relating back to the effects of one of the TEC assemblies I was running in version 2.0. When it was cut off, meaning the peltier wasn't energized, that TEC assembly would ever so slightly begin warming the water flowing through the water block. Wasn't as bad as if you had a radiator in the loop fighting against the chilled water trying to raise it back to ambient, but it did manage to raise it slightly.

Does that make any sense?

 

[Lochekey]

I think so. Like I said it was just a random thought and I really did not want to put the money out there to test it. I figured with your testing you might have some insight into it.

On another note can I ask why you did not include your gpu into the cold loop?

 

[Silver Surfer]

I actually did that in version 2.0 as it had 600w of TDP capability, but when it came to gaming and the GPU was outputting a constant heat load it slowly caused the coolant temperature to rise until the high overclock of the CPU went out of the stability window. At the time of the testing I was running a 3770K overclocked to 5ghz and that was an off the shelf CPU that had never been delided, the only way that CPU was stable was if the coolant temperature stayed below 15c, which at the time I was running the coolant at 10c. All was well as long as the GPU was just idling but once the gaming came into the picture the temperature gradually began increasing and in a very short time the coolant temperature had risen to 18c and beyond which was past the CPUs temperature stability window. That was with the GPU not overclocked at all, just dumping the full gaming load of temperature into the loop.

I had summarized at one point that TEC wise it would have at least took twice what I was running to resolve the coolant issue and maybe more than that and it was already using too much electricity to do that, as my plans for version 3.0 was to cut the electricity use not increase it. Of course I could always have dropped the CPU overclock but then regular radiator water cooling would have sufficed. My original goal was to operate below ambient to achieve the highest possible 24/7 stable overclock of the CPU.

 

[MaddMutt]

Dear Sir,

I know that you are not a constant member here but just visiting from OCN. I'm looking at an easier way for the chilled water thing. I have done the AIO then Multiple Custom Water Loops then Dry Ice. I'm looking at the http://www.performance-pcs.com/hot-...hp-790watt-cooling-capacity-waterchiller.html. I'm looking to just cool a pair of video cards. They range from the AMD 2xxx and up series to the Nvidia 8800 and up series. I still have all of my old basic WC equipment that I used to cool my CPU & GPU a couple of years ago. I have purchased a http://www.performance-pcs.com/ek-thermosphere-universal-gpu-water-block-copper-plexi.html with the necessary adapters. I'm hoping to be able to Bench my GPU's at a lower temp than AIR but not have to go through the trouble of having to purchase\transport\store Dry Ice until I'm ready to bench. I understand that I will still have to prep the GPU's for condensation anyways. I'm not sure if this would take my setup Sub-Ambient.

Thank You For Your Time

 

[Silver Surfer]

Dear Sir,

I know that you are not a constant member here but just visiting from OCN. I'm looking at an easier way for the chilled water thing. I have done the AIO then Multiple Custom Water Loops then Dry Ice. I'm looking at the http://www.performance-pcs.com/hot-...hp-790watt-cooling-capacity-waterchiller.html. I'm looking to just cool a pair of video cards. They range from the AMD 2xxx and up series to the Nvidia 8800 and up series. I still have all of my old basic WC equipment that I used to cool my CPU & GPU a couple of years ago. I have purchased a http://www.performance-pcs.com/ek-thermosphere-universal-gpu-water-block-copper-plexi.html with the necessary adapters. I'm hoping to be able to Bench my GPU's at a lower temp than AIR but not have to go through the trouble of having to purchase\transport\store Dry Ice until I'm ready to bench. I understand that I will still have to prep the GPU's for condensation anyways. I'm not sure if this would take my setup Sub-Ambient.

Thank You For Your Time

I am a constant member here, I just don't post in every thread I look at unless I have something to add. I actually was looking at one of those earlier chillers and decided against it and went with experimenting with the peltier cooling.

The reasons I did not choose the chiller;

#1
Is plainly stated: If you open the box you cannot return it, and there's too many unanswered questions with that amount of money, to take that chance.

#2
Is even though they claim it can go down to 4c, that's with an aquarium water mass that is affected by room temperature and an aquarium light.
How does it perform from active heat load from a CPU or GPU especially overclocked?
Keep in mind this is actually an Aquarium chiller being adapted to use on a CPU.
IMO, you would be better served finding someone actually using one of these Chillers to cool their computer, and see what the results are from them.
As I have zero experience on actual use of one.

#3
Was the actual lump sum cost at the time, and what I was considering buying back then cost around $800 bucks.

#4
The noise of operating one added to the computer running.

Regarding those reasons I opted to begin experimenting with peltier cooling over 5 years ago.

The GPU water block still requires heat sinks on the Memory chips and VRs and supplying air cooling over them.

If you do go for it, please let me know what happens?



Regarding your sig it was 17 days and not lasting 17 hours.

 

[MaddMutt]

I am a constant member here, I just don't post in every thread I look at unless I have something to add. I actually was looking at one of those earlier chillers and decided against it and went with experimenting with the peltier cooling.

I was going to ask if what you had done was a variation on a Phase Change System. I did not ask because I do not know enough to be able to tell the difference.

You talked about the 10 Gal Cooler filled with water and ice. I totally enjoyed that cooling power but do not want to have to leave to get the Ice and setup the cooler, ECt, ECT.
To me it's easier to already have the capability in a small unit and all I have to do is turn it on.

Thank You For Your Time

 

[Silver Surfer]

I was going to ask if what you had done was a variation on a Phase Change System. I did not ask because I do not know enough to be able to tell the difference.

You talked about the 10 Gal Cooler filled with water and ice. I totally enjoyed that cooling power but do not want to have to leave to get the Ice and setup the cooler, ECt, ECT.
To me it's easier to already have the capability in a small unit and all I have to do is turn it on.

Thank You For Your Time

Exactly, the 10 gallon cooler worked great but having to constantly keep it iced was not only laborous, getting dust and dirt in the system requiring special filtering to keep it out of the CPU cooling water block, plus the additional expense of freezing the ice. I was actually using 2 freezers and plastic gallon jugs swapping out the frozen in the 10 gallon cooler.

I wanted some way to continue what I had discovered without having to deal with the ice, that's when I started researching alternative ways of maintaining below ambient coolant temperatures. Looked at similar aquarium chillers as to what you linked but it was too much money to spend at the time, then I stumbled across a link on the internet of a guy that had used a peltier to chill his aquarium. So I thought well if it would work for cooling an aquarium could it be made to work for cooling a computer.

From there my peltier cooling experimentation began and FYI, the link below is the link I stumbled across over 5 years ago.

http://www.shine7.com/aquarium/chiller.htm

Maybe you'll get your own ideas from this, back then everyone was telling me you had to cool the hot side of the peltier with water and that costs a lot more money! This guy was successfully cooling the hot side with an air cooler so why couldn't I do it? So my experimentation began and led to the computer cooling I am running today.

Once you experiment on a small affordable scale and discover what it can do you'll want to take it further, at least I did.

 

[MaddMutt]

I Thank You on taking the TIME for the 101 class

Exactly, the 10 gallon cooler worked great but having to constantly keep it iced was not only laborous, getting dust and dirt in the system requiring special filtering to keep it out of the CPU cooling water block, plus the additional expense of freezing the ice. I was actually using 2 freezers and plastic gallon jugs swapping out the frozen in the 10 gallon cooler.

^ I read about that and was thinking that it was a great idea but I never got around to doing it

I gave up on Water Cooling and went to Dice because I wanted that extra step but it was to much money. You have a very interesting MOD of a Phase Change

Q: Your cooling is only limited by the AIR Coolers ??? Correct

 

[Silver Surfer]

Q: Your cooling is only limited by the AIR Coolers ??? Correct

I'm not sure exactly what you're asking?

The air coolers are more than enough to provide the cooling to actually go down to sub zero they are 360w TDP coolers.

In my testing for version 3.0 before it was put into full operation, these new TEC assemblies I allowed it to go down to 1.7c coolant temperature.

It was still dropping the temperature, but I am running steam distilled water and biocide and did not want the water blocks to freeze up on me so I stopped it from going any lower.

My temperature controller keeps the coolant temperature about 13c below ambient and that is a safe range that there is zero condensation forming.

Which is cold enough to allow my 8700K 6 core CPU to be overclocked to 5ghz at 1.250v and even be stable at ambient start up.

So actually my cooling is limited by the temperature controller, which keeps the coolant temperature from dropping too low and producing condensation on the CPU water block.

I hope that answers the question?

FYI: If I wanted to run sub zero temperatures with this same setup, I would have to add some type of anti-freeze to the coolant, and fully insulate the motherboard against all condensation and ice problems, and drop the temperature on the temperature controller to a below zero point.

 

[Silver Surfer]

Coming Soon: Chilled Water Cooling version 4.0, I will probably post the entire operational build right here at overclockers.com this time, in a new thread of course. SS