There's an interesting method of broadbanding antennas mentioned in
various ARRL publications by Frank Witt, AI1H. It uses the transmission
line length you are going to need anyway to match at two frequencies in
the band (say 3.52 and 3.80 MHz). It's also mentioned in a text I use
teaching a graduate microwave course at Stanford, and I have been
wanting to try it.
Here is a summary of an 80m inverted vee I put up that matches 1:1 at
the cw and ssb frequency. The inverted vee is designed using AO to
resonate at 3.67 MHz (geometric mean of two frequencies of interest),
and to have enough height and included angel to have a resistance at
resonance of 70-75 ohms. The apex is at 120 ft, and the angle is about
120 degrees. I use the AO optimizer, then knock off 1% length from
experience.
Now for the interesting part. I measured a full wavelength of 50-ohm
coax using the MFJ-259 antenna scope. Using Belden 8214, it's about 205
feet long, connects to the antenna 1:1 balun and comes down to the
ground. Next is conected a 1/4 wavelength of 75-ohm coax.
At the transmitter end of the 75-ohm quarter wave, the match is perfect
(less than 1.1:1) at 3520 and 3800! This should work for a horizontal
dipole, or any antenna with impedance in the 75-ohm range.
To see if this was a fluke, I put up a 40m dipole at 50 feet for
Sweepstakes. It uses a half-wave of 50-ohm coax followed by the
quarter-wave section of 75-ohm coax, and it matched fine at 7.0 and 7.25
MHz.
To see how this works, the best tool is Wes Hayward's Microsmith Smith
chart program (ARRL), but you need a reasonable RLC model for the
antenna impedance. I do this by getting the impedance of the antenna
from AO, using a spreadsheet to figure out what the equivalent series L
and C should be, then using that in Microsmith. A better dipole
impedance model has the resistance across the inductor, so the variation
of R with frequency is better modeled.
Once you have the antenna impedance model, you can see why this scheme
works. The electrical line length (in wavelengths or degrees) is less
at the low end of the frequency range and more at the high end, and it
tends to wrap the impedance plot into a much smaller range. Once you
have gone enough half-waves, you can convert the impedance to 50 ohms
with a transformer or quarter-wave section. AO, Microsmith and the
MFJ-259 working together make it easy, and my curiosity is satisfied.
You can use this on any antenna, but if you want to do it with a Yagi,
you need to work harder on the model (try RLC plus a short negative line
length), and you will find it takes an awful lot of line length if the
resitive part is less than 70 ohms or so. But I plan to use a W2FMI
transformer to get the impedance up where I can use this on a 3-el 40.
Hope this is of interest, give it a try.
73 de Dave, W6NL
--
FAQ on WWW: http://www.contesting.com/towertalkfaq.html
Submissions: towertalk@contesting.com
Administrative requests: towertalk-REQUEST@contesting.com
Problems: owner-towertalk@contesting.com
Search: http://www.contesting.com/km9p/search
|