You have tempted me to jump into the battle, guys, despite the more peaceful
part of me saying "don't"!
Some facts:
- Oil, water, and most other liquid coolants, have very poor thermal
conductivity, compared to metals. For this reason their ability to transfer heat
to or from a metal surface depends critically on having a lot of turbulent
motion, to quickly move away and replace any molecule of liquid that has
equalized its temperature to that of the metal surface.
- The thermal contact between a liquid coolant and a metal surface depends
dramatically, and in a nonlinear way, on the flow speed. Faster flow rate is
better, always, within reasonable limits. Not only because it renews the liquid
faster, but mainly because it creates more turbulence. It shouldn't be fast
enough to cause cavitation, though.
- Some of you are mixing up different things. Yes, a liquid coolant will come
out cooler from the radiator, if it flows slower through it, and enters at a
fixed temperature. But that's mainly because the lower flow rate made it bring
far less heat into the radiator! Our purpose is not cooling the liquid as far
down as possible, but moving all the heat from the amplifier to the air, with
the lowest temperature difference possible. That needs the lowest possible
thermal resistance from the tube/transistors to the coolant, from the coolant to
the radiator, and from the radiator to the air. And to achieve this, a high flow
rate and speed is helpful. How much speed and flow rate is needed, depends on
many things, like the amount of heat, the permissible temperature difference,
the dimensions of all things, surface roughness, characteristics of the coolant,
air flow through the radiator, altitude above sea level, humidity, and on and on.
- In a combustion engine different flow rates are used in different areas, to
achieve optimal cooling efficiency. Very simply stated: If you allow a lot of
cooling fluid to flow through an area that produces little heat, that fluid will
NOT flow through those areas where more heat is produced, so those areas might
overheat while the overcooled areas will stay too cool, thickening the oil there
and causing more loss due to oil viscosity. So, carefully controlling how much
coolant flows in each place allows obtaining the optimal temperature at each
place of the engine, while needing the least amount of coolant pump power.
This might find application in an amplifier, too. Like directing most of the
flow around the plate of a tube, and a smaller amount to the base seals.
- In a well optimized system, coolant flow is so fast that there is very little
temperature difference between the fluid moving from the amplifier to the
radiator, and the fluid returning. Thanks to this, the whole system (all of the
tube, all of the radiator, etc) are almost at the same temperature. This allows
using the smallest possible radiator, among other advantages. In case of having
many transistors, a high flow rate makes all of them work nearly at the same
temperature, which is essential for even current sharing.
I followed this thread only superficially, but I think I saw questions about
immersing the tube and tank circuit and everything in oil. I fear that doing so
might cause trouble due to large stray capacitances. They are about 3 to 4 times
higher than in air, depending on the kind of oil used.
And it's a mess, of course!
Manfred
========================
Visit my hobby homepage!
http://ludens.cl
========================
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|