[TowerTalk] Stubs and Reactances in parallel

[Steve London]

This is a little complicated, so follow along if you are interested....

I have a 40 meter beam, and a rotatable, loaded, 80 meter dipole on the same 
mast, a few feet above the 40 meter beam. For years, I have positioned the 
dipole perpendicular to the beam to avoid interaction. Works well, but is 
inconvenient when working 80 and 40 simultaneously in a contest,

Now I am trying to position the dipole and beam parallel to each other. Big 
interaction to the 40 meter beam. No effect on the 80 dipole. Neither is surprising.

So, I thought, this could be easy to solve. Add a 40 meter 1/2 wavelength 
shorted stub at the 80 dipole feedpoint. Indeed, that solved the 40 meter 
problem, but at the expense of 80 meters. In theory, the perfect stub should 
have an infinite impedance on 80 meters. The reality, using real coax, is that 
the stub impedance is around 1000 ohms on 80 meters. When you put 1000 ohms in 
parallel with a purely resistive load, such as 30+j0 ohms, the result very close 
to 30+j0. Good. However, in my case the 80 meter dipole is resonant around 3800 
kHz, and at 3525 kHz, Z = 30-j140. When you put the 1000 ohm stub in parallel 
with 30-j140, the result is 47-j130. Bad !

I can't simply use a relay to switch the stub in/out - I want to work 80 and 40 
simultaneously.

And, in case you are still reading, my matching network for CW operation is 
series coils to cancel out some of the capacitive reactance, and a hairpin to 
bring the Z up to 50 ohms and cancel the rest of the capacitive reactance.

Is there another solution I am overlooking ?

73,
Steve, N2IC