[Elecraft] Remote Antenna Tuner

[George, W5YR]

As a general observation, one should bear in mind that the best of all
worlds is not always a "balanced tuner" driving a "balanced line" connected
to a "balanced load."

In the real world, antenna loads, especially wire antennas, can readily be
substantially unbalanced even though they are center-fed and are of a
"balanced" design. Proximity to other objects,  unequal capacitance to the
ground, etc. are usually the culprits.

By the same token, a parallel line is not necessarily balanced just because
it looks like it should be. Coax is actually balanced line although our
application of it most of the time is in an unbalanced manner. Ladderline is
designed to be operated as a balanced line, but again depending upon the
proximity of other objects near the line, it may become readily unbalanced.

Finally, which does the better job of feeding a balanced load: a "balanced
tuner" or a 1:1 current balun?

There is no "always" answer, of course, but as a practical matter, the
conventional T-network tuner followed by a 1:1 current balun will usually do
the better job of sending equal currents down each side of the line -
however unbalanced it and/or the load may actually be - than the typical
balanced  tuner network, depending upon its circuitry, tuning, etc. And that
is the requirement when driving a "balanced line/load."

The aim with a balanced load system is to feed equal and opposite-phase
currents in each side of the line. The 1:1 current balun, properly
constructed for the frequency range and power levels involved, will
generally do the better job with no attention to tuning, etc. Some balanced
tuner designs, as with some balun designs, are aimed at providing equal
voltages to drive each side of the line to the load. Unless the line/load
combination is actually quite well balanced, the resulting line currents can
be far from equal and opposite in phase.

It seems highly intuitive that a fully balanced tuner with a balun on the
input would always be the superior solution to driving a balanced line/load.
 It can do the job, but not necessarily is it inherently superior.

A revealing test for a balun designed for driving an unbalanced load is to
connect its output through a short length of feedline to a load consisting
of a 100 ohm resistor from one line conductor to a ground plane and a 50 ohm
resistor from the other line to the ground plane - a sheet of metal
connected back to the "ground" side of the unbalanced feed to the balun
input (or tuner output if a balanced tuner is used).

With an r-f voltmeter, measure the voltages across the two resistors while
applying r-f to the balun input. If the line currents are truly balanced,
the voltages will be in a 2:1 ratio as are the resistances. If, on the other
hand, the balun (or tuner) appears as a voltage source, then unequal
currents will flow in the resistors and their voltage ratio will not be 2:1.
This test was first described by Walt Maxwell W2DU in QST some years ago.

It is noteworthy that few if any  "balanced" tuners are available on the
market today. The T-network with output side balun is overwhelmingly the
most common to be found. A few commercial tuners are L-section networks,
usually for high power applications on the lower bands. The Elecraft rigs
use L-section networks in their internal tuners.

It has been shown that the T-network, comprising a variable or switched
inductance and two variable capacitors, provides the capability for tuning
the widest range of load impedances with the fewest components at the lowest
cost. That accounts for its command of the marketplace.

Roy Lewallen W7EL and others have published analyses that demonstrate that
under all but the most stringent conditions, placing the balun at the output
or the input of a tuner does equally well and does not change the stresses
the balun must deal with in handling certain load impedances. Again, this is
highly counter-intuitive, but the math and experience supports it.

Those who have built balanced tuners such as the Measures' design have been
pleased with them and no doubt they work and work well. But, for most
applications, an unbalanced T-network and a 1:1 current balun on the output
will do as well, and with some loads even better, with fewer components and
complexity..

Intuition works well in many fields, but in engineering it can be quite
misleading.

73/72, George
Amateur Radio W5YR -  the Yellow Rose of Texas
Fairview, TX 30 mi NE of Dallas in Collin county EM13QE
"Starting the 58th year and it just keeps getting better!"
w5yr@att.net







----- Original Message -----
From: "David A. Belsley" <belsley@bc.edu>
To: "Randy Moore" <wrmoore47@comcast.net>; <la3za@qsl.net>;
<Elecraft@mailman.qth.net>
Sent: Sunday, September 07, 2003 6:31 PM
Subject: RE: [Elecraft] Remote Antenna Tuner


> Sverre and Randy:
>   If it is just a matter of doubling the components, and even the cost,
I'm
> all for it.  Still, I doubt you'd have to double the costs.  But the added
> components do not double the complexity of concept, so let's go for it.
>
> best wishes,
>
> dave belsley, w1euy
>
> --On Sunday, September 7, 2003 3:01 PM -0500 Randy Moore
> <wrmoore47@comcast.net> wrote:
>
> >>
> >> I wonder, is there anything fundamentally difficult in making
> >> an automatic balanced tuner, say an automatic version of the
> >> BLT-tuner or another Z-match tuner? - or is the problem that
> >> there aren't enough people interested so that noone has
> >> bothered to try?
> >>
> >> --
> >> 73,
> >> Sverre