[Collins] length of cable from KWM2 to 30L1

[Dr. Gerald N. Johnson]

The 20.5 feet is based on solid polyethylene coax like RG-58C/UPA. The 
coax in the relay is usually all air so while the coax velocity factor 
is 0.67, the velocity factor for all of the coax relay except the 
connectors is 1. So 3" of air coax in the relay will replace 2 inches of 
the coax.

Supposedly that "magic" coax reduced the intermod products of the S-line 
exciter and PA. I recall that recent tests have been unable to detect 
any such effect so the modern idea is to use whatever length of coax it 
takes to neatly connect the exciter to the PA.

Theoretically, if the input of the PA is properly tuned it makes a 50 
ohm load and no matter the length of the coax the impedance seen by the 
exciter isn't going to change. If the PA input tuning or the PA loading 
presents a mismatch, what the exciter sees at the other end of that 
"magic" length will be different on every band. The air coax length 
would be 30.6 feet or 9.3 meters. Close to a full wave at 10meters, and 
a half wave at 20 meters, a quarter wave at 40 meters, 3/4 wave at 15 
meters and 1/8 wave at 80 meters. Full wave and half wave makes very 
little impedance transformation even when wildly mismatched, quarter and 
3/4 wave makes the greatest impedance transformation when mismatched, 
and 1/8th wave doesn't do a lot.

I don't know the origin of that length, it would have been fun to have 
asked back about 1963 when nearly every body who worked on those rigs 
was still working at Collins.

12th edition 30S-1 manual claims to explain the "magic" coax length in 
section 3.3. It says there needs to be an even number of 180 degree 
phase shifts between the driver plate and the PA grid. Supposedly the 
"magic" length of coax and the Pi networks in the PA cathode input 
circuit accomplish that goal. The varying plate current in the PA it 
says reflects varying reactance to the driver plate and phase modulates 
the drive signal increasing intermodulation distortion. It says the 
KWM-1 needs another 2.5 feet on that coax run.

I'm sorry, 1/8 wave is 45 degree phase shift, 1/4 wave is 90 degrees 
phase shift, 1/2 wave is 180 degrees phase shift, 3/4 wave is 270 
degrees phase shift and a full wave is 360 degrees phase shift. There is 
an even number of 180 degree phase shifts only on 10 and 20 meters in 
the coax and the Pi networks at the PA and the driver plate should have 
the same phase shift on all bands. For me the explanation does not 
compute. And the length of coax has detectable loss at 10 and 15 meters 
making the driver work harder on the band where both the driver PA tubes 
and the 30S-1 PA tube have slightly lower gain.

73, Jerry, K0CQ, Technical Adviser to the Collins Radio Association.

On 12/13/2015 7:36 PM, Jim Isbell, W5JAI wrote:
> The book says 20.5 feet.   How would passing through a coaxial relay on the
> way affect that distance.?
> ______________________________________________________________
> Collins mailing list
> Home: http://mailman.qth.net/mailman/listinfo/collins
> Help: http://mailman.qth.net/mmfaq.htm
> Post: mailto:Collins at mailman.qth.net
>
> This list hosted by: http://www.qsl.net
> Please help support this email list: http://www.qsl.net/donate.html
>
>