[AMPS] voltage breakdown vs pressure

[John T. M. Lyles]

Hey Jon, Dick and all,

While I don't have a lot of "airtime" right now to elaborate, I would like
to add 2 cents worth to the discussion, as it is something that really has
a lot of practical meaning to us. Like W0ID, I work at 7000 feet above sea
level. High voltage is part of our business, with large RF
amplifiers/tubes/klystrons. We have to derate everything. When companies
near sea level sell us products, one of the topmost design aspects that we
are wary of is their breakdown ratings, clearances, sharp points. This may
also apply in ham amplifiers, and I would rate Alphapower's stuff as
excellent from a breakdown standpoint, due to the testing that Dick refers
to.

Our 14 inch diameter coaxial transmission lines operate at 201 MHz, with 3
MW peak power. We use about 3 PSI of dry air, to pressurize them. Not only
is the increased pressure better for breakdown ratings, but it also serves
to improve heat transfer from the center conductor out. We installed tiny
bleeder valves at the far end of the lines, where the gas barrier/vacuum
window at the coupling loop in the drift tube linear accelerator is
located. The pressure leaks there, and the air movement cools the window as
well as exchanging the atmosphere in there. Stagnant air with excess
ionization from inception of corona breaks down polymers and seems to be
worse overall.

Now, to summarize the results of theory. Without going into
multipactoring/electron avalanche (which is a real problem in our vacuum
systems), field emission, or into gettering in tubes (someone already made
good description of that), one can refer to most physics texts and look
this up. I am not a physicist, but I think the text HIGH VOLTAGE TECHNOLOGY
edited by L.L. Alston (British Railways Board) for Oxford University Press
1968, is the best. It is listed as a Harwell Post-graduate series text for
the UK Atomic Energy Authority Research Group.

Paschens Law relates voltage breakdown and pressure. It has a dip at an
intermediate pressure (which we would still consider evacuated). As you
move towards absolute vacuum, the breakdown properties are high, meaning
you can standoff a lot of voltage. Our klystrons and triodes/tetrodes are
pumped to 10^-6 or ^-7 Torr. At ^-2 or ^-3 we would have problems. Same
thing is true of the accelerator tanks. They have small drift spaces in
which the voltage might be a megavolt (Q is 60,000) across the capacitive
region. This accelerates protons, what we want to do in the first place.

In the dip of the curve, glow discharges are easily initiated. This is the
"neon" sign region. Also a lot of other devices operate there. Spark
arrestor tubes, gas regulator tubes (remember the OA1?). It would not make
a good vacuum capacitor or RF power tube. As you move up in air pressure,
towards atmospheric, the properties get better again. At > 1 Atmosphere,
things get better to a point. This is why pressurized systems work for high
altitude systems. Nitrogen and SF^6 improve things too. However, at high
pressure, field emission comes into play, and things begin to breakdown
again. Paschens law fails here.

As for the small spaces in a vacuum cap, the pressure had better be low to
be to the left of the minimum in Paschens curve. Otherwise they would
breakdown often. The breakdown at 1ATM is not as good. But some capacitors
made for Continental Electronics for transmitters are gas filled. This
gives a degree of holdoff, and the cooling that is needed for high duty
operation.

Somewhere recently I read an old paper on Vacuum Capacitors, but I cannot
recall now where. It talked about this very subject (1940s info). Oh well.

73
John
K5PRO

--
FAQ on WWW:               http://www.contesting.com/ampfaq.html
Submissions:              amps@contesting.com
Administrative requests:  amps-REQUEST@contesting.com
Problems:                 owner-amps@contesting.com
Search:                   http://www.contesting.com/km9p/search.htm