[Antennas] Frank Witt's Tuner Loss Test

[George, W5YR]

After reading Frank's excellent presentation of his "indirect method" for
estimating tuner loss in the Sept/Oct 2003 issue of QEX, I was encouraged to
evaluate my several years old MFJ 989C tuner.

Briefly, Frank's method involves adjusting the tuner with a 50 ohm load for
an SWR of 1:1 or a maximum return loss value. Then, SWR and return loss  are
measured for loads of 25 ohms and 100 ohms. Frank has derived equations for
tuner loss in dB, using return loss or SWR or calculated values of
reflection coefficient rho. He also translates tuner loss in dB into a
percentage of input power that is lost. His method has passed rigorous
scrutiny and is theoretically correct. It is also very straightforward and
easy to use.

The results are presented here with no claim for accuracy or that your 989C
might have similar characteristics. All I will say at this point is that I
believe that my test protocol and instrumentation are correct.

I used the AEA CIA-HF impedance analyzer Model 5012-5000, Serial No. 0157. I
have used this analyzer since about 1998 and have found it to be
exceptionally useful and accurate. Contrary to Frank's comment in his QEX
paper that "instability of the readings made it impossible . . ." , my
instrument is rock steady in its indications once it has gone through two or
three computational cycles.

For each band, the lower band edge frequency was used. The test fixture
provided a 1% precision thick-film substrate 50 ohm resistor and two such
resistors in parallel for 25 ohms. A 1/4-watt 100 ohm metal film resistor
rounded out the set.

I used return loss as the variable for computation since the AEA analyzer
provides this parameter with three digits of precision. And frankly, the
equations were easier to use working with return loss than with SWR or rho
(!).  <:}

The table below summarizes the results of the tests. The power output is
relative to an input of 5 watts.

Band    Loss(dB)    Loss(%)    Output(watts)
------    ---------    ---------    --------------
80            0.91            18.9            4.06
40            0.58            14.3            4.29
30            0.73            15.5            4.23
20            0.53            11.5            4.43
(20)         (0.58)         (12.5)         (4.38)
15            0.56            12.1            4.34
10            0.81            16.9            4.15

The tuner was adjusted for the 50 ohm load with the capacitors set at
mid-range and the inductor varied for lowest SWR. Then the input capacitor
was adjusted until the SWR was reduced to 1.01:1 for a return loss of 46.1
dB. The values shown in parentheses for 20 meters were obtained with
different tuner settings as explained below.

The increased losses at 80 and 10 are to be expected, due to more inductance
being used on 80 and overall higher circuit losses on 10.

In order to determine the effect of other tuner settings which also provide
a 1.01:1 SWR, the output capacitor was placed at maximum capacitance - the
setting generally recommended for minimum loss - and the inductance and
input capacitor adjusted for 1.01:1 SWR as before. This test was done only
on 20 meters and the results are as shown in parentheses in the table.  It
is interesting that the loss was slightly higher with these settings.

The latest QST contains in the Technical Correspondence section a piece by
Frank discussing further the accuracy of his indirect method. My position
after making these tests and studying Frank's articles is that the method's
inherent errors are significantly less than my AEA analyzer probably
contributes  to the process.

My overall conclusion about tuner loss is that these measurements confirm
what I have always thought about the level of tuner loss: around half a dB,
maybe a little more or less. The effect "at the other end" is clearly in the
noise level as it were. I suspect that other losses in the antenna system
exceed those of the tuner unless a very inefficient tuner is used or it is
maladjusted. Again, note that few if any commercial tuners are made with
ventilated cases or fans to provide cooling. Tales of "3 dB tuner losses"
are probably just urban legends.

I use a "3 KW" tuner for QRP largely because of the inherent lower losses in
the large-gauge wire in the inductor and the quality capacitors. While small
tuners using small components are convenient for field use, etc. there is
the possibility of added loss. It would be of great interest if someone
would make a series of tests like these using a BLT Z-match or other popular
QRP tuner.

With the large tuner, the case is likely to lower the Q of the coil less
than a smaller case would. An interesting test that I will make next is to
remove the top cover of the 989C to determine if its presence is, in fact,
increasing inductor loss.

I hope that these data will be helpful and that they will encourage others
to make these measurements and to post the results for comparison. Note that
Frank determined that the MFJ 259B is more than adequate for this task and,
in fact, is his preferred instrument.

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