To build the crowbar for your HT supply, based on a Thyratron as the
crowbar element, you should choose a good sensing device, such as a
shunt or current transformer. Since the power supply negative end
will jump below ground when the positive side arcs to ground, the
sensor needs to be able to withstand and isolate this without
destroying the thyratron triggering circuit. Opto coupling or the
current transformer are the best ways. Grounding should be to a
single low impedance point in the system, for everything.
One typical way is to use conventional comparators to sense when the
current spikes above a certain threshold (and maybe an integrator to
capture time x current). The integrator or filter is needed to
prevent nuisance crowbars from noise in the power supply rails and
the amplifier. The sensor will be in the B+ lead going to the RF
amplifier, after everything else including series resistors. The
comparitor will then drive a 600 V SCR, which latches (giving you a
memory to light a lamp and indicate crowbar fired, and shut off the
power). The anode discharges a capacitor (< 1 uF) of several hundred
votes through a pulse transformer, the secondary of which is directly
connected between grid and cathode of the thyratron. Be sure to have
some series R between the thyratron plate and the capacitor storage,
to limit peak I through the device. Connect the thyratron shunted
across the lead going to the RF amplifier plate HT connection back to
the common ground. And don't forget to have warmup timers on the
thyratron that prevent turnon of HT until the tube is hot and ready
to protect.
Finally, include a very fast shutdown command, from the SCR back to
the AC mains switch, preferable a solid state relay as opposed to
main relay contactor. This will prevent blowing fuses when the supply
crowbars itself. The crowbar will short out in the microsecond
regime, while the AC line shuts off within a half cyle or so(or the
next phase zero crossing in three phase supplies), in the
milliseconds. The fuse or circuit breaker should stand this oK.
You can test the effectiveness with foil, 30 AWG wire, all the
techniques that the tube manufacturers recommend. The wire should
survive while the power supply crowbars and shuts down. Then you can
rest assured that the tube will not be hammered by arcs. By the way,
crowbars fail, so an occasional test with some sort of relay or
switch should be done.
We use big crowbars on all our high power amplifiers, so we have many
tens of crowbars that have to work perfectly to prevent costly
damage. We use ignitrons and spark gaps mostly, but the thyratrons
are used to drive the ignitrons in some systems. Gridded tubes and
klystrons have an occasional arc over, for one reason or another.
Sometimes we blame it on the cosmos. Some blame it on parasites...
EEV is one of the best sources of Thyratrons in my opinion. I got
some Russian made tubes from Richardson, that lasted less than a year
before they began falsing on intermittently.
73
John
K5PRO
--
FAQ on WWW: http://www.contesting.com/FAQ/amps
Submissions: amps@contesting.com
Administrative requests: amps-REQUEST@contesting.com
Problems: owner-amps@contesting.com
|