Ian White's discussion on the sources of vacuum arcs and anode to
screen breakdowns in tetrodes is what I have witnessed in my years of
working with high power tetrodes. It is standard practice
(recommended by Burle, CPI/Eimac, and Thales Electron Tubes) to place
a spark gap across the screen bypass cap, typically rated 20% above
the highest DC voltage that will be used. The sockets available from
Thales sometimes come with a built-in gap, for instance the TH18555
socket for the TH555A tetrode has a small air gap with adjustable
threads, typically set for 2200 VDC breakdown at the altitude above
sea level where it will be used.
I have witness autopsied tetrodes which have blown screen wires, from
an arc. We rarely (!) remove load on an operating amplifier to
prevent sudden injuries to the engineer. However, the phenomina of
ion bombardment (gas) is a real problem. Tubes are conditioned to
allow the getter to absorb as much free gas before high power
operation commences. Another source of gas which Ian may have omitted
is when an element (typically screen, plate or grid) is
overdissipating, and outgasses. This usually causes a fault, where
the tube breaks down internally. There are also chemical getters
which are cold, and do not need the heater or anode heat to make them
function. RCA/Burle tubes made a lot of them inside their tubes. They
are typically inside a glass ampule which gets broken open once the
tube is being sealed up, where they attract residual gas molecules.
Now before someone says that these are only things that are
experienced in larger tubes with handles or eye hooks, the principles
of ceramic metal tubes are scalable, from small sizes on up. We have
Seimens or Claire gas tubes across some other tetrode screens, with a
small series RFC from the power supply. We use the RFC and a few ohms
wirewound as a fuse, when the tube arcs anode to screen, and the gas
tube fires. This prevents the anode B+ from backfeeding and
destroying the screen bypasses, the power supply, the wiring, etc.
Its a lot easier to replace the RFC than a tube or power supply.
Hams would be wise to use similar techniques if they use expensive
or rare tubes for high power amplifiers. The cost of protection is
small compared to the cost of the RF device.
As an aside, last summer I seriously overdissipated a 3-500Z in the
PDM switch tube socket of a Continental 314-R1 (Collins 828C1)
broadcast transmitter in my shack. It had a shorted damper diode so
the ST was running into a dead short and was being driven into
conduction a low DF by grid pulses that would represent carrier level
only. The 500Z anode glowed nearly yellow, and a deep blue glow was
seen around the edges. It had roughly 8500 volts across it! Luckily
the HV breaker tripped along with the plate overload, but this
happened fraction second later. The tube recovered OK, but it took a
beating.
73
john
K5PRO
_______________________________________________
Amps mailing list
Amps@contesting.com
http://lists.contesting.com/mailman/listinfo/amps
|