While digging to read one of AG6K's favorite references - Gossling (IEE,
1932) on Flash Arcs in High-Power Valves - I stumbled across another
classic that I just had to tell ya'll about.
THE BALANCING AND STABILIZING OF HIGH-FREQUENCY AMPLIFIERS, WITH SPECIAL
REFERENCE TO POWER AMPLIFIERS FOR RADIO TRANSMITTERS
by W. Ure, E. J. Grainger, H. R. Cantelo
Journal of IEE (UK), Vol 72, 1933.
It has a brief sketch of HF amplifier developments, with Miller and Nichols
(1919) understanding of the nature of oscillation of amplifiers to the
neutralization schemes of Hazeltine and Rice (1923-24).
"The development of the screen-grid valve has to some extent superseded the
use of neutralized circuits, notably in receiving amplifiers, but this type
of valve so far has only found a place in the earlier stages of
transmitting amplifiers. This is partly due to the difficulties introduced
by its dependence upon screen voltage, which may be affected by an
inductive voltage-drop along the screen lead causing the screen to act as a
control grid, and partly due to the residual capacitance between the grid
and anode being too large to ensure stability at high power, particularly
at the high frequencies."
Then it goes on to discuss balanced bridges and how the Rice and Hazeltine
neutralizing ciruits can be ebodied that way. For instance, the Rice
circuit consisted of tapped coil, with out of phase signal coupled back
one side from the plate (through a capacitor). Hazeltine used the tapped
coil at the plate, and derived the signal for the grid through a capacitor.
Some of these circuits were patented then.
In the discussion, the generation of spurious frequencies is elaborated on:
"It is not always an easy matter to damp out parasitic oscillations. In
cases where the desired oscillations do not suppress the parasitic
frequency, common practice is to insert non-inductive resistors of equal
values in each of the bridge arms."
"Methods of suppression of general application where the desired frequency
and parasitic frequency are sufficiently apart consist of the use of:
Extra capacitance connected between grid and filament of each valve to
ensure that feed-back effects are in the wrong phase relationship to
produce regeneration.
Series resistance in individual valve anode or grid leads to act as damping
resistance.
Combinations of paralleled inductance and resistance inserted in individual
valve anode or grid leads, so proportioned that the combination has
negligible impedance at the desired frequency, but considerable impedance
at the parasitic frequency. This last methode is due to the Western
Electric Co. "
Then they discuss layout problems, and get to the subject of instabilities
in push-pull amplifiers, such as from asymmetric drive. The placement of
the RF ground at the center tap of an output coil is discussed, and how it
can be problematic when the two grids are not driven equally or out of
phase.
Methods of neutralization while cold or hot are covered, much the same as
we use today.
Finally, they even discussed choke resonances! These guys would be right at
home on this mailing list, if they were still around. A way to get around
the problem when bandswitching from 4 to 18 MHz coverage, was to short out
turns in the choke to move resonances for various frequencies.
When I read classic work like this, I think back to the tattered 1948 ARRL
handbook that my ancient neighbor gave me -in the late 1960s. I remember
all those guidelines that you could not question such as the proper use of
butterfly capacitors, proper bypassing of the center tap of an RF coil in a
P-P amp, keeping leads short, etc. Well, now I see where all that
knowledge was first written down for the RF engineers. I think I'll look up
some references from this paper....
73
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
|