Amps
[Top] [All Lists]

[Amps] Water cooling

To: amps@contesting.com
Subject: [Amps] Water cooling
From: Manfred Mornhinweg <manfred@ludens.cl>
Date: Sun, 21 Aug 2011 01:47:21 +0000
List-post: <amps@contesting.com">mailto:amps@contesting.com>
Dan,

for my new low cost solid state amp, which is slowly gaining shape in 
the design software and via real world tests of subcircuits, I was 
considering two heat sinks forming a duct, through which air is moved 
(on demand, controlled according to heatsink temperature) by two fans on 
the ends. I actually have the heat sinks already. But water cooling has 
always been in the back of my mind as a great way to make the amp 
smaller, lighter, and less noisy. Water cooled amps travel very well to 
DXpeditions. You can carry them around easily in the same bucket that is 
then filled with water to act as heat sink!

> You also want to do something to ensure turbulent flow behind the
> transistor as otherwise the actual thermal resistance can be MUCH higher
> then you would expect, stuffing a spring into the pipe is effective. 

That's a good idea, but I understand that given enough flow velocity 
(actually a pretty low one is enough), the flow will automatically be 
turbulent. I guess it's the degree of turbulence that matters! How well 
the water layers closest to the copper get interchanged with the upper 
layers.

Since you seem to know about watercooling, can you give me a hint as to 
how I can calculate the resulting thermal resistance? I know very well 
how to calculate the amount of water flow required for a specific heat 
rise at a given power level, but I just don't know how much copper/water 
contact surface I need to achieve a certain thermal resistance there!

Like a very simple practical case: If I need to get rid of 1000 watts of 
heat, and I guarantee that I will put enough cool water through the 
thing so that the water comes out at less than 40 degrees Celsius, and I 
  need the copper surface in contact with the water to never heat above 
60 degrees Celsius, then how much surface I need at the copper/water 
interface?

I understand that here the flow velocity and surface roughness play a 
role, but the sheer numbers evade me! Can you help me? If you do, you 
might change the FodPower 1500's design from somewhat noisy air cooling 
to nearly silent water cooling!

Manfred, XQ6FOD.

========================
Visit my hobby homepage!
http://ludens.cl
========================
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps

<Prev in Thread] Current Thread [Next in Thread>