[Amps] Solid state amplifier cooling

[John Lyles]

As Jim mentioned, multiphase cooling is the way it is done these days in high power RF tubes. The large Eimac and Thales tubes use it, as do the big MW klystrons for scientific work. We use this technique in our new amplifiers at work, but it isn't something that hams should desire, as it requires very pure low oxygen and low copper in water. It requires that pH and dissolved oxygen be monitored along with resistivity. The devices that use this technique must be designed so that the vapor phase re-condenses in the device, in tubes right in the jacket or 'boiler'. Out of this comes warm water, like a conventional water cooled tube. 

Example from work: a 7835 triode from Burle Industries (RCA) requires 150 GPM to dissipate 250 kW conventionally. A TH628 Tetrode from Thales requires about 65 GPM for same dissipation, or can operate 1 MW plate dissipation with about 120 GPM. HUGE difference in effectiveness. But not something I would recommend for kW level of RF and DC power. 

73
John
K5PRO


> Date: Thu, 6 Jun 2013 09:15:54 -0600
> From: MU 4CX250B <4cx250b at miamioh.edu>
> To: amps at contesting.com
> Subject: Re: [Amps] Message-ID: <963402869080783179 at unknownmsgid>
> CSolid state amplifier cooling
> 
> Interesting idea, Bill. I wonder if folks on the reflector have tried
> vapor phase cooling? In principle, that should be the most effective
> cooling scheme of all, given the high latent heat of vaporization of
> water. Of course, the devil is in the details, and the practicalities
> might be too burdensome. This is an interesting discussion.
> 
> Jim w8zr
> 
> ##  Vapour phase is not as good as you think it is. If u look at any old paper copy
> of eimacs care and feeding, re vapour phase cooling., you will see that the max effect
> occurs at 135 C..and not 100 C.   That entire section is not in the latest online version,
> too bad.   100 C is too high for SS devices..and 135 C is clearly way too much.
> 
> ##  Then you also have the problem with bubbles forming on the surface of
> what ever you are trying to cool.   Tubes made for vapour phase cooling
> have rough surfaces, to break up the bubbles.
> 
> ##  Eimac sez that forced air is good for 50 w per sq cm of anode area.   And  135 W
> for vapour phase cooling.  Water cooling is the best at 1000 w per sq cm.
> 
> ##  the ultimate setup   is multiphase cooling..a combo of water-vapour- then back to water
> cooling.   But that would be difficult to implement on a SS device.