[Amps] water cooling

[John Lyles]

I would recommend to check into the datasheet for flow volume versus dissipation, if it is a well-spec'd tube.  As for size of condensor, you are talking vapor cooling I guess. One good source of info is to find an older Thomson tube catalog, when they described their vapotron cooling technique in the 1970s. Then they had Supervaporton. Once they went to Hypervapotron, it got a lot more complicated in some ways, simpler in others. For instance, no need for a condensor at all. But water purity and oxygen level, plus the type of plumbing used, became very important. This is similar to CPI's multiphase cooling, where the phase changes from steam to water in the water jacket of the anode. 

I have been using a pair of 4CW250,000B tetrodes all week as a pulsed 200 amp power load. No RF, just DC driven with a pulse generator. They are loading my plate power supply for testing, before I finish the RF amplifier that will be drawing that current next spring. The power supply and capacitor bank (250 uF) must function well together before i stick a very expensive cavity amplifier on the end of the cable. 

Each Eimac tetrode has 40 GPM flowing, water is about 1 Megohm-cm resistivity, to allow no more than a mA or so of current in each 9 foot long 1.5 inch ID water hose. The real amplifier will need 130 GPM in the anode, and 5 other water loops of < 1 GPM for filament, screen, and cavity. When water cooling shoot for keeping the temp rise across the anode at about 5 - 10 deg C or less, and keep the flow reasonable for the pipe size, using standard pipe calculations. Too much flow will erode fittings. Too high a resistivity will strip ions from fittings and cause metal to move around and get plated in places you'd rather not (like inside of anode or hose fittings). For hoses, design them as a cylindrical resistor of water, and figure out the ohms per cm or meter, and then just make them long enough to keep leakage small, under a mA if possible. Use rubber, or polymer hoses or even PVC pipe, nothing with a lot of dark carbon (black hose). RCA use to say something like 4 Meg ohm per kV of
plate voltage I believe. 

Happy Holidays
John 
K5PRO

> Id like to test a Class C  industrial tube (Not on a ham band!) Ive had for 
> almost 25 years when I stripped the self excited oscillator for parts. The 
> questions would be how to calculate the necessary flow volume based on plate 
> dissipation, size of condensor, suitable hosing, controls, sensors, and all 
> that good stuff.
> 
> Carl
> KM1H