Exploring Below Ambient Water Cooling

Page 13 - Seeking answers? Join the Tom's Hardware community: where nearly two million members share solutions and discuss the latest tech.
what kind of temp difference do you see with sub zero cooling?
i have to fight with a intel biased father on why i want to do sub zero cooling, and i have nothing to back me up lol
 
@ cbrunnem, I'm not sure I'd trust a spray. I've seen the steps needed to insulate a motherboard and I'm not sure a spray would do the trick. Short term maybe but no way long term.

@nna, what kind of temp difference? I'm not sure I see the point to your question. Temp difference is whatever the cooler can max provide.
 


i meant in comparision to traditional cooling, how much benefit does it have?
 
Only way to get sub ambient temps 24/7. Water can get you close to ambient and things like dry ice or LN2 will go sub. But they don't last. TECs are the best at keeping you sub ambient and allowing 24/7 running at those temps.
 


ok, coz im thinking of making a freon cooler (with an eco friendly substitution of course 😛)... and wouldn't want to see little to no difference if im gonna be spending money on it
 


This is a below ambient cooling solution thread not sub zero, temperature comparisons with what you can expect with this type of cooling are covered up in the beginning of the thread.

Sub zero cooling is mostly for high clock records and extreme bragging rights, it's not a viably convenient 24/7 cooling solution, because first of all it's just not necessary and extreme voltages can kill your hardware even under sub zero temperatures.

Why even go sub zero if you're not flat going for it, hardware expendable!

To me it's extremely funny when the extreme high clock bragging is proceeded by honesty.

Meaning the CPU that supplied that high clock is no longer operational, it was expendable from the very beginning!

My hardware is not expendable and all my CPUs are still fully functional with zero degradation.

Insulating the motherboard to prevent condensation damage, is not necessary for example to run a SB i5 2500K at 5.0ghz with this cooling solution, as long as you discover your condensation limits, and at what temperatures condensation begins to appear.

You don't even have to go as low as the condensation producing point to run 5.0ghz 24/7 with the 2500K.

This solution would allow higher IB clocks I'm sure if you compare what some are getting on air, at the mercy of ambient room temperature.

My setup never reaches ambient room temperature it's always below ambient, so 24/7 stable clocks are higher than air cooling, or standard water cooling closed loop cooling, because the radiators still get their cooling from ambient room temperature air.

The real purpose of below ambient cooling is to allow higher 24/7 stable clocks with the lower load temperatures for longevity of the hardware, and to allow a certain level of jolly time high clocking for sheer curiosity.

This cooling solution is very quiet, I do have fans running but none are running at their full speed rating that are high CFM and high noise levels, they're only cranked when necessary.

I can crank my fan speeds up to control the condensation and use ice to drop my water temperature down as far as 5c without motherboard insulation.

Which actually allows higher clocks than I'm even interested in running, but it's just not necessary to run those high clocks 24/7 daily, the Sandy Bridge CPUs I own do everything I need them to do and more running @ 4500mhz, and you can do that air cooling.

Ask yourself this; What would be my overclocking goals if I had sub zero cooling and where would I be content stopping?

Is your hardware expendable?

Ask your father if your hardware is expendable!

Ryan






 


No expert here but I would also think that a spray would limit the heat exchange from the actual board itself. Ryan is intentionally remaining over dewpoint, so it should not be an issue, but would be an interesting experiment in a sub-zero thread...
 


😛 thanks for the in depth awnser, and i got my terms mixed up... i have no plans for extreme overclocking for i do want to get some time out of my CPU

 
Update: I've been experimenting with various water levels using the peltier, I'm presently running exactly 4 gallons of water, which left a large air space inside the cooler.

The peltier is not cooling the air so that air space becomes a negative cooling loss to the water, thanks to my friend Art, I used some impact insulating foam slid inside gallon freezer bags, laid them in the cooler floating on the surface of the water and the negative cooling loss is gone.

Dropping down to 4 gallons of water, with the insulation floating on the top of the water, plus increasing the peltier clamping pressure, I can now drop the water temperature under full load!

Today was a rare day, I got rained out this morning my water temperature in the cooler was 16c, normally I would start the day with a jug of ice, if not, I would have been bumping 20c in 2 hours with no ice. (even with 10 full gallons of water)

Been running my DJ program all day with graphics, and surfing the net at the same time on and off since 9.00am @ 16c water temperature, it is now 3.00pm, 6 hrs of run time with no ice and my water temperature is 11.8c for a total temperature drop of 4.2c, since power up.

I would say at this point this peltier experiment is 100% successful, it is doing exactly what I was wanting it to do.

Eliminate the daily use of ice.

Now the test of time and how long this cooling solution runs problem free.



 
I don't really know, but since I'm not freezing ice in two freezers anymore, it's probably less than before, I don't run the machine 24/7, just when I use it, but the freezers were running 24/7 freezing the ice
 


I additionally still use the AC Pan Tabs listed at the beginning of this thread and a silver killcoil inside the output line to the pump.
 
Apparently I'm getting consistency, usually Saturday mornings are times when my wife has no problem with me spending time on my computer, so I get to catch up on all that's going on out there.

Started up the computer this morning at 5:41AM water temperature in the cooler was 16.7c, 6 hrs later surfing the net while my DJ program was running in the background using no ice, my water temperature is 11.8c at 11:41AM, dropped 4.9c in 6 hrs.

That's consistent with another test I ran a few posts up.

It's to the point I don't run the peltier at all until my water temperature gets up to the 16c range, 16c is 6c below my ambient room temperature.

At the time of this posting 12:12PM the water temperature is 11.6c, once I shut down today the cooler will hold the water temperature for quite some time.

:) Ry
 
Additional Note; Using the Peltier the water temperature drops much faster from 18c down to the 14c range, once the water temperature in the cooler reaches a level dropping below 14c the rate of temperature drop starts to slow.

Apparently somewhere around 10c seems to be a leveling off point, it will drop below 10c but takes a long time to do so, I believe this is a direct result of only supplying it with 12v when it is a 15v peltier.

However this is actually working great for my needs!
 
Well with the new system capabilities and your successes with the pelletizer setup you have it has convinced me to reapply the TEC’s to my next system thank you for this info your TEC works better than my old ones you can get a 60 F drop on the TEC itself mine only have a 40 F drop so I am going to get a couple of new ones for that extra 20 F drop it will not be the same as yours but I will implement the same kind of idea to drop the water temps I will have one directly attached to the CPU and the other 3 will have a set up like yours designed to cool the water and a second water loop for cooling the TEC’s instead of an air cooler but my intention is to get it as cold as possible to get the most clocks out of it.

I will have to do a lot of testing on this setup to get it just right but your tests and setup has shown me the need for them again the last time I used them was on the K7 series and they got really hot after a little clocking today they get hot after a lot of clocking but seem to hold stable if you can keep them cool.