> Terry Gaiser - W6RU writes ..
>
> The amplifier was in STANDBY ..... NOT KEYED .... but with all
> voltages ON ! I had a reason to remove the top cover (high
> voltage interlock removed), I slowly raised the cover from the
> front of the amplifier leaving the rear of the cover sitting on
> the amplifier ... when I got the front of the cover raised about
> 3 inches ... all of a sudden the amp made this huge big grunt
> and the circuit breaker tripped off. After I removed the top
> cover and turned the amp back on I had idling plate current in
> the standby mode. Long story short ... I had one of the two
> 3-500Z's with a grid to filament short and that tube had metal
> particles rolling around inside of it.
Your anecdote missed two important facts:
1) The amplifier was stable with the top cover on ... you made a
change, probably in the plate circuit stray capacitance when
you moved the top cover. That change "tuned" the parasitic
circuit.
2) The amplifier has "all voltages on" ... as such, there was
undoubtedly some residual anode current (all voltages could
not have been "on" if the cathode were completely cut off).
This allowed gain (and oscillation) if the feedback conditions
were "tuned" properly.
The whole point of the "argument" is that ANY amplifier will
become an oscillator at some frequency if the gain and feedback
at that frequency are sufficient to overcome the circuit losses.
There are two ways to prevent oscillation (parasitic) ... either
ADD LOSS (the Measures approach) or design the circuit to
eliminate (minimize) the possibility that the feedback will be
of the proper phase at the same time the gain is sufficient to
sustain oscillation (the Rauch approach).
The rest of the argument is simply over the "trigger" to start
the oscillation. In fact, the triggering event does not matter -
it can be photon bombardment (Measures), unintended "mistuning"
(your cover removal), a gain shock (HV "spike" due to a mains
glitch), momentary overdrive (the ALC spike from many modern day
solid-state exciters) or simply gaussian noise ... some more
plausible than others.
In any case, not all unexplained, sudden, catastrophic high power
amplifier failures are due to parasites. With sufficient study
the failure mode in properly designed and properly operated
power amplifiers can almost always be found to be due to external
causes ... dirt/dust accumulation from lack of maintenance, a
transient failure in the antenna system, an operator error,
component drift with age/heat stress/overload, or even (horrors!)
a design that wasn't all that proper in the first place!
For those who prefer the supernatural preventive powers of loss
inducing devices, that is your right. However, do so without the
humiliation, ridicule, rancor and personal attacks. If you have a
scientific argument to make, do so with properly documented and
verifiable sources.
73,
... Joe Subich, W8IK
<W8IK@Subich.com>
<www.qsl.net/W8IK>
--
FAQ on WWW: http://www.contesting.com/ampsfaq.html
Submissions: amps@contesting.com
Administrative requests: amps-REQUEST@contesting.com
Problems: owner-amps@contesting.com
Search: http://www.contesting.com/km9p/search.htm
|