[Amps] low pass filter fail

[John Lyles]

Built a lambda/4 stub out of 9 inch diameter EIA hard line, for
the big amps at work. It was build primarily to pass a tuning shaft 
through the output line to the variable capacitor paddle for output 
coupling to the plate/screen resonator cavity. Originally I thought it 
might need water cooling, so that too was planned to pass through the 
quarter wave stub. It took some modeling and measurement to get the stub 
length to be high Z at the operating frequency as when the diameter of 
the outer conductor approaches fractional wavelength then the geometric 
length isn't exactly what works. This was all explained in Robert 
Pound's chapter in the MIT Rad Lab textbook on coaxial line center 
conductor supports. He figured out how to broadband the stubs so they 
wouldn't have VSWR when the operating frequency shifted slightly.

http://www.ieeeghn.org/wiki/index.php/Oral-History:Robert_Pound

Anyway, my big stub worked very well on suppressing second harmonic. 
Better than -50 dBc right out of the amplifier. Wish it still reflected 
a short to the 4th harmonic, 805 Mhz, but due to being far overmoded (9 
inch coax that is), it doesn't work.

Essentially it is forcing a short for the even harmonic at a particular 
distance from the tee to the plate region of the tube. It modifies the 
plate current or voltage waveform. This sort of modification is similar 
to what is done with class F amplification, using the technique of Tyler 
from Marconi, later used in RCA broadcast transmitters. In my amplifier, 
I probed inside the main cavity in a safe location, and discovered that 
there wasn't much second harmonic being generated, due to the quarter 
wave stub on the output.

73
John
K5PRO



> Date: Tue, 29 Apr 2014 02:17:38 -0700
> From: "Jim Thomson" <jim.thom at telus.net>
> To: <amps at contesting.com>
>
>
> ##  Do you really require a LP filter ?   You could also use a .25 wave long coax stub..with far end shorted.   You can also add additional stubs etc.
> Paralleling caps is the real answer to your exploding cap problem.  2-4 caps in parallel  would be the ideal ticket.   Your typ 2m ant will only be resonant on
> its 3 and 5th harmonic.   You can make stubs that will kill odd harmonics if I remember correctly. W2VJN has a good book on stubs.
>
> ## Parallel some caps..and you wont have any more issues.   I have used the teflon sheeting sandwiched between 2  aluminum plates,,and that works superb...
> provided it is constructed correctly.
>
> ##  I like the idea of a LP filter......... then it kills everything beyond the cutoff freq.   A bandpass filter would even be better.    You could also use a LP filter
> for TX....and a low power bandpass filter...for RX only.
>
> Jim  VE7RF
>
>