[Amps] 1/4 wave shorted stubs on amp output

[John Lyles]

It is for a cavity amplifier, with capacitive-coupled output paddle. Coax is 9 3/16 inch EIA, so I despise having to modify it. The stub is designed on paper and in machine shop, having an acme threaded screw inside that will change the penetration of the paddle, all while running at ~500 kW average power.
I was just curious for the collective experiences and wisdom here. The Amps forum contains a world resource of people who have done many things with many RF amplifier variations. 

Yes, you can expect some plate current (or drain, or collector) variation if you put a LPF on the output, and an out of band harmonic is not terminated but is reflected back. I had a lot of trouble with a particular FM broadcast amplifier cavity years ago, when it played wonderfully into a flat Bird resistor load, but when 
LPF was added (to meed FCC requirements) the thing when ape and created spurious oscillations. You can bet that the plate current waveform was modified. Now the second harmonic might be 20-30 dB down naturally in a cavity circuit due to Q. Still there is a slight effect, although small. The harmonic termination could change things a percent or two. But it can affect the propensity for a power tube to behave as expected, also. 

The plate current waveform may be distorted, which really isn't a fundamental frequency problem in the end result due to the Q of the cavity. However, in a more conventional Pi network or link coupled output with moderate Q and some bandpass or lowpass response, one might see something. It could mainly effect the DC current in a small way, which might bump or drop efficiency a small degree. I haven't tried it. Gerald, it can work as a way to feed B+ but only over a narrow band, like 10% or less BW. Some industrial RF heaters shunt feed their B+ this way, through inside of a tubing that is the output inductor. While it isn't a 1/4 wave stub, it is essentially the same thing, feeding at a high impedance for RF. Just what you'd like a choke to be.  

In the 1980s I built quarter wave stubs for putting on the front of a 2m repeater site near Philadelphia to protect the receiver from both nearby FM stations (overload) and from harmonics of transmitters. I used semi-rigid 0.25inch coax. 
I put a 1/4 wave DC short on a GMRS base station radio last summer, 162 MHz, that prevented static build up from the monopole on the roof top. Before that, they were seeing static buildup problems. These little stubs are wonderful devices if you have a network analyzer to make sure they are on the correct side of the Smith Chart, accounting for the connector insertion phase and stray L at the soldered shorted end. 

I am concerned with this big one as I don't want fireworks in my amplifier. 

73
John
K5PRO




>
> ## I bought george's [ w2VJN]  book from inrad a few yrs back
> It's  superb.  There is also excellent material  in the latest ON4UN
> low band dx book.   George's book show's  other designs too.. all using
> coax. 
> 
> ##  apparently, whether you put the  say shorted  1/4 wave stub  right at the
> output of the amp.. vs  a 1/4 wave away from the amp  depends on whether you 
> have a simple PI  net... or  a PI-L.  They work great on the output of the 2 x GS35B
> 6m amp.   Attenutaion  improves by a whole bunch, when bigger coax is used.  Huge
> diff between 213  and  heliax.... for the same  1/4 wave stub.   Somebody pointed out 
> to W8JI  that  they could hear his 2nd harmonic... when JI  was TX on 160m... and folks 
> could hear him easily.. when operating on 80m.   JI  installed the stub  right at the 160m ant 
> feed point... end of problem.
> 
> ##  I didn't  think of the stub interacting with plate current waveform modification.  Will you
> get the same waveform modification, when using a LP Filter... say with a cut off of 33 mhz? 
> What if the LP filter has a cut off just a bit higher than the band in use... like  monoband LP
> filters ?? 
> 
> Jim   VE7RF 

> ------------------------------
> Message: 5
> Date: Tue, 30 Mar 2010 08:26:21 EDT
> From: TexasRF at aol.com
> Subject: Re: [Amps] coaxial stub on amplifier output
> To: jtml at vla.com, amps at contesting.com

> I will display some ignorance here:
>  
> Does the plate current waveform distortion cause any problem at the  
> fundamental frequency?
>  
> Your idea for using a quarter wave stub like that is interesting. For a  
> single band amplifier, it could replace the rf choke for B+ and avoid all 
> those  pesky resonance problems. Of coarse the plate blocking C would need to be 
>  removed and used to connect the plate load C (in a pinet) to the output to 
> keep  the B+ off the load C.
>  
> 73,
> Gerald K5GW