[Moon-net] VE7TIL EME website...

[Leif Asbrink]

Hi Scott,

There is a photo of your balun at your website. 
This is an interesting design that I have not 
seen discussed before.

The design seems to be a 35 ohm coaxial line
to transform from 50 ohms to 25 ohms. (made from
two parallel 70 ohm cables)

The 35 ohm cable serves two purposes.
1) To change impedance from 25 to 50 ohms.
2) To act as a balun that prevents current to
   be injected into the boom tube.

To satisfy 1) the length has to be adjusted by
the velocity factor inside the cable to make it an
electrical quarter wave.

To satisfy 2) with a cable that runs parallel to
the boom tube, the length of the cable from the
radiator to where it is connected to the boom
tube has to be one quarter wave at the wave-speed
of the transmission line formed by the outer conductor
with its thin dielectric layer on its outside and the
boom tube away to the point where the outer screen
and the boom tube meet (at the connector)

To make 1 and 2 happen for the same cable is not 
trivial. It may be a good idea to move the point of 
grounding a bit behind where the impedance change 
takes place since the outer wave-speed is likely to 
be higher than the inner.

In your design however the cable forms a coil, and 
there is a rather large capacitance between the hot 
and cold ends of the cable at the point where it 
leaves the plastic box. A capacitor will lower the 
resonance frequency for 2).

The cable is squeezed however and the area enclosed 
by the current on the outer side of the screen is 
rather small. A squeezed coil has a lower inductance 
than a round one and the amount of squeezing can be 
used to tune the inductance to match the capacitance 
at the point where the cable leaves the plastic box.

If the capacitive current through the "capacitor" 
does not match the inductive current on the outside 
of the cable you will send current in three directions 
at the feedpoint. The current that flows from the 
inner conductor to the right hand side equals the 
sum of the current sent into the left hand side 
and the current dumped into the boom tube.

Having say 10% more current on one side may not be 
so harmful as one might first think because even 
if you remove the left half of the radiator, the 
right side and the boom tube would form something 
that resembles a GP antenna which essentially has 
the same radiation pattern as a dipole. The GP is 
not ideal however and some energy may be lost in 
radiation from the boom tubes.

To measure is to know. If you have the equipment 
try to measure the impedance between the left side 
and the boom tube. Best is to remove the radiator 
and replace it with a resistor so you can see the 
very high impedance at the resonance. Fig. 1 at
this site gives an idea:
http://www.sm5bsz.com/antennas/cpudes02.htm

For a cable with PVC outer coating you might
see an attenuation in order of 35 dB at the 
notch when the loop looks as on your photo.
With a cable having an outer coating with good
RF properties, i.e. low dielectric losses,
 you might see 45 dB.

The following is approximate and for a feed 
impedance of 50 ohms for the measurement at
the above site: 
If you see 40 dB attenuation, 1% of the current 
that flows through the 50 ohm resistor (that 
replaces the radiator during measurement) is 
flowing through the 50 ohm impedance of the 
level meter. (The approximation is to neglect 
power lost in the balun. At large isolations 
this approximation is good.) 

The impedance from the balun to the level meter 
is then 100 times bigger than the impedance to 
ground at the balun. 25 ohms to the virtual 
groundpoint at the load resistor and 2500 ohms 
to the level meter (with 2500 ohms as a resistive 
impedance internally in the balun to ground.)

The power budget becomes R*I*I or 50*I*I useful 
watts in the load and 2500*0.01*0.01*I*I lost 
power. The ratio loss to useful is 0.25/50 or 
0.5% which is perfectly ok for influence on S/N 
and power budget. At high power levels the 
associated heat could be a problem but not likely
in an array of several antennas that share the power.
With a good RF material on the outside resistive 
losses should be negligible, particularly at a 
low feed impedance like 25 ohms. The finite 
attenuation at the notch is then caused by 
radiation losses.

Squeezing the loop might improve the isolation 
from 20 to 40 dB (by shifting the notch frequency)
and reduce the power sent into the boom tube from 
5% to 0.5%. The gain improvement could be something 
like 0.2 dB so it is not a big thing, but the 
impact on G/T could be significant.

It might be interesting to point the antenna at 
90 degrees elevation and listen for a vertically 
polarized signal coming in from the horizon. It 
should be extremely weak, maybe 30 dB weaker than
a horizontally polarized signal. In case boom tube 
currents matter you might hear it pretty well....

I am sending a copy to MOON-NET. I do think
baluns and the way they work is of general 
interest. It is also a safety precaution. 
If I made a mistake I am sure someone will 
correct me.....

73

Leif / SM5BSZ