[BC] Antenna matching tuners: if from scratch

[RichardBJohnson at comcast.net]

Can you beg, borrow, or steal a 500 to 1000 foot roll of 50-ohm coax? Skinny low-power stuff will do. Then you need to find an old fashioned sweep generator like those used to tune TV sets in the "olden" days, with some tunable markers to show your frequency. Heathkit used to make these, as well as HP and other expensive brands. It needs to put out at least 1 volt because it will take 300 mV or more to turn on a diode. Then you need to find a coaxial diode with a BNC connector output. Typically they have a type 'N' fitting on the RF and, and a BNC fitting out the output. They are often used to detect TV signals in transmission lines.

You connect the output of your sweep generator to a coaxial tee. The through-connection connects to your 500 ft cable delay-line, and the tap connects to your RF diode. The BNC from the diode goes to an ordinary scope. It's set to auto-trigger. The detected signal will be in the kilohertz range.

Start by leaving the antenna end of the delay open. Set your sweep generator so you can see the tone envelope on your scope. You can turn on the marker that shows your frequency and observe its glitch on the scope. This tone is generated because the reflected signal will be mixed with the incident signal from the generator and detected by the diode. Because the signal is being swept, its frequency will have changed by the time it travels the delay-line, the produces the beat signal you observe on the scope. This is also a great way to measure return-loss at microwave frequencies that can't be observed with conventional instruments.

Set the scope gain to make the resulting waveform full-scale. This calibrates your system.

Connect up a 50-ohm load attached end of the coaxial delay-line. The tone will disappear. Crank up the scope gain until you can measure the voltage again. The return-loss is 20 log(A/B) where A was the open-circuit voltage and B is the terminated voltage. This should be 30 dB or better if the coaxial cable and the terminating resistor is any good.

Now connect up the antenna. Adjust the scope gain so you can observe the dynamic return-loss. Move the marker around so you can see your assigned frequency. Adjust the marker a few hundred kilohertz on both sides of the carrier frequency and mark the scope face with tape or grease pencil.

Now adjust the tuning stubs for the lowest return-loss, consistent with a reasonable bandwidth.

I used to do this at the top of TV towers! I'd lug up all the equipment, build a shelf of boards set through tower members, connect the equipment into some tower-lighting circuit by removing the bulb and screwing in an adapter. After several trips up and down 600 foot towers, you learn to tape a box over the PE cell first, so you have power on your first trip! Sometimes, with FM antennas, I'd have to bend antenna elements to get the return-loss down. RCA FM antennas were flimsy so they would sometime get distorted in shipping.

Cheers,
Richard B. Johnson
Book: http://www.AbominableFirebug.com/

----- Original Message -----
From: "wpio" <wpio at gate.net>

Wow,  so there's no rule of thumb. (yes I mean input matching tuners)

In this case, ice would not be an issue so we could go for dead on 
frequency.

However, I was hoping a Bird Thruline and cut/try cut try so to speak, 
would work, just didn't know if there was a perfered sequence.

I doubt if I can come up with a network analyzer.

Randy
wpio

Thomas G. Osenkowsky wrote:
> The correct way is to use a network analyzer to
> examine the bandpass. There may be several
> combinations that produce a 1:1