On Sun, 05 Sep 2004 18:06:07 GMT, Richard Clark
wrote:

|On Sun, 5 Sep 2004 12:47:45 -0500, (Richard
|Harrison) wrote:
|
|The tank circuit is mostly a harmonic filter providing a very high
|impedance to the fundamental frequency and shorting out the harmonics.
|
|Hi Richard,
|
|Even here, the Goatman offered in his notes that his finals tank
|(actually a series resonant Z match) offered a loaded Q of 2! (If I
|read his scribblings correctly.)

Yeah and he also calculates

(http://www.techatl.com/wrek/docs/gnm69_25.htm)

the required plate load resistance as:


Eb
Rl ~ -------
Idc

Which for class C is off by about a factor of 2, but with Eb = 2500
and Idc = .25, he does the division and comes up with Rl = 1000. Hey
what the heck, what's a factor of 10 among friends.

If the calculation is done more accurately:

Eb - Eg2
Rl = ---------
K * Idc

Where K = 2 for Class C and
Eg2 = 300 (screen voltage)

Then Rl ~ 4 Kohm

Since the minimum output capacitance (Cp) of a 4CX300 is 4.5 pF, the
parallel equivalent of Rl and Cp is Rp ~ 3920, Xp ~ -388.

Thus the minimum possible Q ~ 10, which to someone who has built a few
VHF amplifiers, sounds much more plausible.

For example here's one I designed and built not much later that the
WREK(ed) transmitter.

http://www.qsl.net/n7ws/K7CVT_Amp.html

But it gets worse. Try as I might with the component values he
specifies, I cannot develop a plate load Z anywhere close to what is
necessary. He has a lot more inductance that he thinks, so maybe that
helps and I suspect his output lowpass filter (seen in the photos but
not on the schematic) is part of the matching network.

I'm really surprised that with the construction and documentation
presented he could get FCC type acceptance.