Water Cooling

[Jonathan Bright]

I've enjoyed this thread and have been thinking along similar lines for my JLH Class-A. But another thought is to look at something other than water for the transfer medium. I had not realised that liquids differed in their thermal characteristics until I read a book by Phil Irvine the great bike/car engine designer. He makes the point that oil absorbs thermal energy more quickly than water but also dissipates it more slowly. (This is probably why people poured boiling oil on their opponents when their castle was being attacted!) For our purposes this makes oil quite attractive. The thermal properties mean that it is better than water at drawing heat away from the relatively small confines of the area around the semi-conductor. On the other hand the slower dissipation rate will not be a great handicap. Most of us will be able to make the liquid to air heat exchanger as big as we like. There will not usually be space restrictions on a set up where the radiator is mounted outside the house and old car radiators are cheap etc.
Now I haven't done any further research on this but I have a couple of questions I wanted to chase up. One: the viscosity of oil may mean that a motor can't pump as much fluid through the pipes when compared to water. This reduced flow rate may negate some of the thermal advantage...... although the flow should improve as the system warms up the oil. I am guessing that it won't have corrosive problems and a gallon of olive oil could look attractive too.
Secondly, what does it sound like? What I'm really waiting for is for someone to do an A/B test and announce that the oil based design had smoother highs than the water cooled system!

[kilowattski]

Don't think for one minute they won't either.

[Jonathan Bright]

With, perhaps, both types having a"fluid" quality in the midrange.

[imix500]

I'm sure, as long as you have a frikin' $300 power cable on the pump.

[Brett]

Originally posted by Jonathan Bright
For our purposes this makes oil quite attractive. The thermal properties mean that it is better than water at drawing heat away from the relatively small confines of the area around the semi-conductor. On the other hand the slower dissipation rate will not be a great handicap.

From my sitting on the fence, pragmatist engineer POV, I fail to see why it's better, for the simple reason that the pump, fittings and heat exchanger (car radiator) are cheap, easily obtained and more than adequate for the task. Even with several Alephs a decent size radiator, sited well will dissipate more than enough heat. There is always a breeze up the side passageway of my house, which is where I'd place the radiator if I were going to do this (nearly did myself, but tubes sound better).

If you want to pick up more heat in the assembly with the semi's mounted on it, simply use a angled entry fitting, use a flange plate internally with a number of holes drilled in it, zigzag the tubing, or place a spring in the straight sections to make the water flow in there more turbulent to improve the heat transfer. Cheap, simple, not fancy, last forever. And as the absolute temperature in any well thought out system should be quite low, there are a lot of plastic fittings and adaptors that will allow you to use a 50/50 water/ethyl glycol mix and prevent oxydoreduction.

Grey, alaskanaudio, akira and Brian Donaldson have given a lot of good suggestions on how to do it. Another tip for bending the tube, plug the ends with a bit of scrap pine, fill the tube with fine sand and bend it around an object of approximately the correct radius. I've built car exhausts this way (with a bit of applied heat from an oxy) and everyone who saw them thought they'd been made with a mandrel.

[GRollins]

I certainly have no problem with trying to improve something that needs it, but until you try water cooling and discover for yourself just how effective it is, I don't see the point to all these embellishments.
I'm sure that some of them would look cool, though.
You want to beat your head against the wall? Bend your will against the portability issue. Other than that, if it ain't broke, don't fix it.

Grey

[imix500]

It certainly works well in my large frame lasers. The head and the power supply generate about 15 kW of heat and a few gallons of water a minute through 2 large "water blocks" keeps it nice and cool.

[lumanauw]

Are the transistors need tobe cold? I remember Mr.Pass said that mosfets likes it "HOT". So if it is a mosfet amplifier, class A (runs hot), and we use very effective cooling device, like water cooling, fans and radiator, will it change the sound?

I have encountered a commercial power amp that uses very thick aluminum block (not heatsink)+heatsink. I wonder if it is for adding weight or what.
Then I realize that the massive metal block is there to maintain "Thermal Inertia", so the temperature will be very stable.

There is a thread somewhere about "Thermal Memory Effect/Peufeu website". I know that this "thermal inertia" is opposite than that "thermal memory".

But which is more make sense and which is hoax?

[JRKO]

Its not very expensive nd looks good AND does the job!

http://www.quietpc.com/uk/watercooling.php

[SGregory]

I suppose that if there is a diffin sound quality between the type of cooling fluid, the direction of flow would also be important.?
Scott