Topband: Efficiently Matching Low-Impedance Antennas to 50ohms

[Tom Rauch]

> I got curious (always dangerous) about how much power
> was actually making it through the tuner, so I used an
> RF ammeter to measure this.  On the input side, the
> Bird wattmeter and  the RF ammeter were within 5w of
> each other, so that gave me some confidence about the
> approach.

We probably know input impedance is 50 ohms or pretty close
to it on the input.
We don't accurately know load Z.

 I moved the RF ammeter to the antenna side
> of the tuner and got a real surprise.  The
> measurements indicated that there was about 3dB of
> loss in the tuner.  Wow!

I'm assuming the tuner was adjusted correctly. It should
always be adjusted with maximum possible capacitance and
lowest inductance that allows a match.

I pulled data from several 989C tuners I tested. Average 1.8
MHz efficiency (using good test methods and equipment)was:
12.5 ohms 75% (-1.2dB)
25 ohms 81% (-.92dB)
50 ohms 88% (-.56dB)

I really can't see yours being much different.

Without complex equipment or methods, the only good way to
measure or estimate loss is with temperature rise. Virtually
all loss in the 989 tuner is in the inductor. If it had 3dB
loss and you ran ~400 watts, the roller would seriously
overheat (to the point of discoloration) after about two or
three minutes of on time. The roller will only dissipate a
hundred watts or so before getting very hot.

You could put the tuners in a styrofoam box (like a cooler)
and measure temperature rise of the air with known applied
power over a set amount of time. Substituting a large light
bulb or resistor you could measure temperature rise with
known power.

I'm afraid that's the only way (outside of having lab
equipment) to measure loss with reasonable accuracy (within
10-20%).

> I borrowed a TenTec 238B tuner and replaced the MFJ.
> I found a match point to 50 ohms and then measured the
> current on the antenna side again.  The news was
> better -- only about 1.5dB of loss.

71 % eff. Sounds low at 20 ohms. I'd expect that tuner to be
in the 90% eff area.

> measurements.  That is, the field strength with the
> TenTec tuner was about 2dB stronger than with the MFJ
> tuner.

You have to be a little careful measuring fields near an
antenna. You may be reading a combination of electric and
magnetic coupling. The meter reading change log 20 might not
fully apply as it would when reading purely voltage or
purely current readings.

All that aside, a conventional L network with reasonable
components can easily get in the upper 90% efficiency range.
So can a T network with enough capacitance and a good
inductor (Q>300).

My bet is you are estimating efficiency lower than it really
is. That is very common when measuring this type of thing.
The L net on the common point of my phased array matches
about 25 ohms, efficiency is nearly 100% in that network
since it uses large components.

73 Tom