Return to

Watercooled chiller?


Not really sure. The oil might/probably would get thicker around the water block with it being below ambient temp there. Might help slowly cool the oil over all too? Though not sure if that would get over all below ambient at which point the case would condense water on it.


I suppose oil based fluids wouldn’t have a problem in pumps. I may consider doing a build like this sometime in the near future.


Your all covering everything that has already been done.

If you want a new idea. Here is mine.

  1. Make a box fully fluid cooled on all heat generating components.
  2. Seal and hard vacuum that box. So no air and what little water vapor is in there will evaporate not condensate.
  3. Externally chill the cooling fluid to your tastes.

No one is really going to do this practically because it’s not really practical and vessels under hard vacuum can be very dangerous. IE like old CRT display tubes when you broke them. They were a vacuum inside.


Debaurs phase change 3M fluid seem interesting as well not not really serviceable. Works on the phase change being at a temp I guess above ambient so no condensation.


Peltier external loop mineral oil build? Please show me :slight_smile:


It’s on the dark web with naked geisha girls with fans cooling you and your gaming rig :slight_smile:


I was just thinking a regular water look through the block, just surrounded by oil externally.

I would not want to pump even light oil through a PC water cooling setup, they are not made for it and likely would have some problem sooner or later.


This would certainly require some testing. Orings, such as those use on G1/4 fittings, in pumps, and the rubber used for some tubing is not compatible with hydrocarbons. You will need to test all of these components to ensure that the rubber bits do not swell or deteriorate in mineral oil.

Another concern is pump power draw. The thicker mineral oil will require the pump to work harder that it would pumping water. Test pump power draw and monitor pump temperature.


You’re saying that the pumps couldn’t handle it?


They are designed around moving water, oil is significantly more viscous. Yes.


Seems like this could be solved by mounting the motherboard horizontally with the CPU facing down. Seeing as how, AFAIK, nobody makes a case like that, I’d probably use an upside-down Bullet BH8. It looks like the outer shell, which is where I’d guess the feet are mounted, might be removable with enough coaxing.


that’s not how condensation works, by bringing the cpu or the board it’s attached to below the dew point moisture from the air will naturally collect on those surfaces regardless of orientation.


Saying I don’t know and it needs testing before committing hardware to it.


If that’s the case that kind of rules out what I was thinking of… I was thinking of pumps that would rotate the fluids… however if it were possible to have pelletier’s inside the case… then the heat could be dissipated possibly externally. Hmm, I wonder how effective… hmm. One way could be to lierally have an element that is both inside and out side the case at the same time. The cool side would be inside the case and the hot side would be outside of the case, having a seal.


If you want to circulate oil in a tank literally just put a fan in it, right in the oil.

A “normal” mineral oil rig does use a pump to circulate oil through radiators to cool the oil like a standard water loop but it needs to be a bit beefier to compensate for the added viscosity

Idk what kind seal you plan on using but mineral oil tends to disagree with most rubbers and some plastics


Not necessarily. Oil has a lower friction than that water and can in many cases off set the higher viscosity.

But with something like this a fluorofluid would be the best route.


Sure, but wouldn’t the CPU itself stay mostly above ambient since it’s the source of all the heat we’re trying to get rid of? I thought it was mostly the cold-side hoses that caused problems.


You would be riding a fine line


Well, yeah, probably.

Next time I’m not po’, I might try it with some really cheap parts to test the concept, then build something not as cheap if it works.


you also need to take into account the lower heat parts of the chip those will also conduct heat and possibly go below dew point, this is especially worrying around the pins and socket.