Vic:
I didn't attend any of the engineering meetings for the design of the
30L-1. I told them that I was simply too busy to bother with such a trivial
project.
Seriously, I can state several good engineering reasons, which I would
have offered to support the circuit strategy which was agreed upon (in my
absence.)
There are separate issues for DC and AC 'float' of the grid circuit.
>From a DC standpoint, it is desirable to apply a very high DC cutoff bias
during standby. The tubes generate a very predictable noise current all the
way down to 0-current. I am actually curious about amps with zener diodes in
the cathode circuit to accomplish this - what is the actual I(quiescent) they
achieve?
With the DC current isolated, it was much easier to read the grid
current using the multi-meter that was designed in.Bias adjustment/setting
becomes somewhat easier if the bias circuitry doesn't have to handle the full
supply current, as in cathode circuit biasing.
Also, the negative bias on the grid (for cutoff) worked out quite well with
their relay switching. They had to have a power supply for the relay no
matter what, which possibly could have been powered from a cathode return,
but the regulation problem would be severe.
Finally, they paid serious regard to linearity specs. Although I have
never attempted to document its operation myself, their use of RF feedback
for this purpose was a minor breakthrough. I assume that their final
measurments justified this, since it took some effort to implement. There are
a few notable examples of this in the annals of RF design but I don't want to
dwell on it. It is just one more of the obvious reasons for not wanting to
directly ground the grid pins.
I have discussed the floating-vs-grounded grid issue many times, and
in all of my discussions it always reverted to the potential problems with
grid-circuit inductance. My belief is that with a little effort you can keep
it (grid inductance) well under control and enjoy all of the advantages of a
biasing circuit that doesn't have to operate at high current levels. I just
built a socket for a new GS35b amplifier with built in grid bypassing, using
a novel but very simple capacitor structure. Its inductance is so low that I
really can't properly read it on my best Z-meter.
Eric von Valtier K8LV
|