Topband: High Z vs Low Z shunt feed tower

[Tom Rauch]

>  "That is what I have.  One 15kV 200pf door knob cap in series with the
>  shunt."
> > ----------
> > Plus one other component -- the 9:1 un-un.

The 200 pF doorknob, assuming the R at that point where it is in 
series is 450 ohms (allowing you to cancel reactance with that 
capacitor), carries 1.8 amperes. The reactance of the capacitor is 
442 ohms, voltage across the capacitor is 1.8*442=760 VRMS, or 
1075 v peak.

Q of the shunt system at that point is essentially 1, so in your 
case it looks like a good result. It won't always be a good result, 
but in your case it is.

> > If you move the tap point down from the 450-ohm point to the 50-ohm
> > point, you'll be able to get rid of the un-un and the required gamma
> > capacity will be about 600 pF, three times what you're using now.  Tom
> > basically said to use the fewest number of components to reduce losses
> > and that the greatest amount of capacity is the most efficient.

He already has a good combo as far as how the system is right at 
the capacitor. Assuming your numbers are correct there would be 
147 ohms Xc and 5.5 amperes through the capacitor. Voltage 
across the cap would be 5.5*147= 808 volts and Q at that point in 
the feed system would be almost 3.

Remember the qualifications I included. I said for a given system, a 
change in gamma effective diameter to increase capacitance 
required, efficiency is improved and Q lowered. 

Let's say we keep the feed system at 50 ohms by moving the tap 
point (so we still only need one capacitor), and increasing shunt 
diameter until we need 2000 pF of capacitance, Xc is now 44 
ohms. Q is less than one, and since the resistance at that point is 
50 ohms (same current) voltage is 242v RMS across the capacitor.

So you see, for a constant value of R anything you do to reduce 
the reactance reduces Q and improves efficiency (because VAR 
power and distributed losses are less).

By far most of the losses (ignoring ground system losses) are in 
the conductors in the gamma and the tower. That system also has 
"Q" of it's own that figures into the system Q. The results above are 
only valid for what is happen right at the capacitor. The Q of the 
transmission line making up the gamma and the tower increases 
with higher gamma impedance   
 
> But, will it provide a better pattern or ERP?  That's the issue.  I am
> certain I could figure out 50 different ways to get a match.  Does it
> improve my SNR?  

SNR ratio, assuming you system does NOT have common-mode 
problems with noise, is unaffected by any of these changes. You 
can feed an antenna any way you like as long as it is done 
correctly without common-mode currents on the feedline, and S/N 
ratio is unchanged as long as pattern is unchanged.

Noise is an electromagnetic wave just like a signal, and you can 
not "dc ground it out" nor can you "shield it out". The only 
exception is where noise is generated right in or very near the 
antenna, like corona discharge in the antenna itself.  

If your system has common mode-noise problems, improving the 
ground system would reduce noise. So would adding a choke on 
the feedline (like a sting of beads).

My only point is, for a given system, anything you do to reduce the 
number of components and reduce the reactance required to 
cancel the feed system reactance will improve efficiency and 
increase bandwidth. In other words, if you have a 120 foot tower 
and you run a wire to the top for an omega match, the Q of that 
system will be higher and losses will be higher than running a 
gamma to the correct point for 50 ohms resistance with a simple 
single series capacitor cancelling reactance. The thicker you make 
that gamma (while still maintaining a tap point for 50 ohms), the 
more series C you will use and the lower Q and less loss you will 
have.

But all of this is splitting hairs. The bulk of loss will be in the 
ground system in most installations. Bandwidth is a good reason 
to mess with the feed system, efficiency or "noise" is not.

73, Tom W8JI
w8ji@contesting.com


--
FAQ on WWW:               http://www.contesting.com/FAQ/topband
Submissions:              topband@contesting.com
Administrative requests:  topband-REQUEST@contesting.com
Problems:                 owner-topband@contesting.com