More vacuum, get it....(oxymoron)
To add another nickels worth, an internal arc would be quite quiet, but it
is between metallic electrodes usually, so some sound does come out. A
horrible sound I remember is a sort of metallic 'tink' that occurs when the
ceramic seal blows out in a 7835 super power triode. This happened several
times in the 1989-93 era here. I was watching the filament voltage
suddenly drop (it had a semi-'constant' current supply) as the air rushed
in and ruined one. As it turns out, a number of problems led to these
catastophies, including operating the tubes at about 80% of rated
dissipation, operating into significant VSWR at the turnon in a very high Q
load, and probably manufacturing defects in the alumina ceramic ring.
Sometimes an extrernal arcover would occur, and the ceramic would crack.
Someday I'll post a photo of one, if you want. Each failure cost us over
$100K USD.
We don't see these types of failures anymore, knocking on my wooden table
right now as I say it.
We have two 2 liter/sec Ion pumps mounted on each tube. When connected to a
4 kV DC supply, they will attract the gas molecules and clean up the tube
during operation. As higher power is reached, the grid, plate heat up more,
and outgassing continues, beyond what the manufacturer can deal with in
bakeout. Its interesting to watch the spikes of ion pump current, as gas
burps are released in the tube. If it gets above a certain amount, we trip
off the plate voltage with an interlock, to minimize arcing inside. We have
learned how to 'nurse' the RF duty factor, and process the tubes while
making RF. We slowly increase peak power and DF, while watching the vacuum
level. At some point, a stable vacuum is reached, typically in a 48 hour
run for a new tube. We always overprocess, beyond normal temperatures, so
that momentary excursions of RF power don't release a lot of gas and trip
off the RF system.
Even the filament being lit up produces gas molecules, which must be taken
care of with getter or ion pumps, before the plate power is energized. This
takes about 15 minutes with the ion pumps. The outgassing seems to be worse
at about half current in the filament, 4000 Amperes DC. At that point, if
we left the current there, we would really overwhelm the vacuum tube. So we
move past that point quickly during ramp up, to the nominal 6800 Ampere, 5
VDC filament power.
These tubes are so large, that as a condition of warranty, it is required
that they be put on the ion pump power supplies on the shelf, every 2
months or so. With the number of seals, and the 'virtual' leaks inside, the
pressure will slowly rise otherwise, and eventually ruin a tube when it is
first lit off. Once in service, the heat of the RF losses and filament
power will keep the getter happy, and the external ion pumps are just there
as extra precaution.
Thank goodness our amateur tubes aren't so finicky.
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
|