[Amps] oil vs distilled water cooling.

[John Lyles]

> ##  so exactly what is the issue with oil vs distilled water ???   Distilled water
> can be found in 4 litre jugs, cheap, at any grocery store.  To the best of my knowledge,
> distilled water will extract heat a lot faster / better than oil.   I only saw 1-2 small tubes in the
> entire Eimac catalog that used oil cooling.  All their big commercial broadcast tubes are either
> air, or water, or vapor phase cooled tubes.  Or multi-phase cooled.... which is a combo of
> water cooled and vapour phase cooling.

All of the amplifiers I design and use at work are water cooled, some 
flowing water and some multiphase, or hypervapotron mode, which is the 
same thing but a trademark. Its not a tough problem to crack, and I'm 
dealing with considerable HV. I studied the RCA, Eimac and Thomson 
recommendations some years ago, and use a common set of specifications. 
Keep the DC leakage current in the hoses < 500 uA per hose. So for 
whatever resistivity is maintained, the hose diameter and length is 
designed for this current (resistance) at the maximum plate voltage to 
be used. DI water is closed loop cooled, as its New Mexico and we cannot 
waste water. This complicates things with heat exchangers and secondary 
loops with cooling towers. Since the water is recycled, we have to watch 
the ph and the dissolved oxygen level as well. For < 5 kV DC I think the 
requirements are much less difficult to meet.

Now, adding an ionic additive like water wetter is risky for a pure DI 
or distilled water system. One reply said the typical content is:
 > DiIsoPropyl Alcohol Ether 1-4.9%
 > TriIsopropyl Alcohol Diether 1-4.9%
 > Sodium Molybdate 1-4.9%
 > Tolyltriazole 1-3%
 > PolySiloxane Polymer

That third one is a good corrosion inhibitor. I use the same molybdate 
as an ionic additive, that when put in solution with pure water, makes 
an excellent dummy load. The amount is about 0.5% by weight, and I use 
what are called water column loads. These are essentially 
multiwavelength long coaxial cylinders with a shorted center conductor 
at the far end. They contain a way for this solution to flow through, 
and if I use pure water the mismatch is terrible, like 1 dB RL. When I 
add the 0.5%, the match becomes 20 dB RL at a specific frequency. So a 
tiny bit of molybdate makes a huge change in RF loss. I have also 
noticed that it makes a large difference in the DC conductivity of the 
solution. My best soup is 530 microsiemens/cm^2. With just a few hundred 
cc dumped into a system with several hundred gallons flowing will change 
this value by a factor or 2 or more.

I would be wary of doping the distilled water with these chemicals 
without doing an experiment first.


> ##  last year, I tested distilled water  vs  tap water  with just a fluke 87 dvm.  The
> dc resistance of distilled water is sky high compared to tap water. Distilled water is
> a superb insulator.   Dunno what happens to the dc resistance once  water wetter is
> added to the distilled water.  I should re-run the test.   For a tube cooling application
> it?s a moot point, since it?s a closed loop anyway.
>
>