[Elecraft] [BL1] Seems lossy in all useful configurations

[Vic K2VCO]

N2EY at aol.com wrote:
> In a message dated 10/13/07 4:58:04 PM Eastern Daylight Time, vic at rakefet.com 
> writes:
> 
> 
>> There is no difference if the balun is on the input or output side of an 
>> unbalanced tuner.
> 
> In theory, no.
> 
> In practice, there can be a big difference.
> 
>> See
>> <http://fermi.la.asu.edu/w9cf/articles/balun/balun.html#SECTION00050000000000000000> 
>>
> 
> He's not looking at the big picture.

His point is that if you use an unbalanced tuner with the balun on the 
input, then there is a common-mode reactance that is equivalent to the 
differential-mode reactance that is being canceled by the tuner which is 
reflected back to the balun. So the problems that the balun has on the 
output do not go away when you move it to the input.

Using a *balanced* tuner followed by a balun or a link-coupled tuner 
like the Matchbox solves this problem. But what many of us want to do is 
to feed a balanced line from a radio with a built-in automatic (but 
unbalanced) tuner, like the KX1-K3, without adding an additional tuner.

> If the balun is ideal, or close enough to ideal, or if the shack-end 
> impedance of the balanced load are within a certain range, the 
> unbalanced-tuner-with-balun-at-the-antenna-end idea works fine.  Thousands of hams use it with no 
> problems and good results.
> 
> But in some cases the shack-end impedance of the balanced line can be very 
> high, very low, and/or highly reactive. Under those conditions some baluns don't 
> work well, and all sorts of odd things happen. Sometimes the end result works 
> well enough that the ham doesn't notice anything wrong, particularly if s/he 
> has nothing else for comparison.

Absolutely correct, and I recommended exactly what you do: measure the r 
and x on the balanced antenna and then either change the line length, 
add reactance to compensate for the x, and choose an appropriate balun 
ratio to get as close as possible to the r. Of course, you can't always 
do this easily for all bands!

> You can't just blindly increase the number of turns on a wound-core balun to 
> increase the impedance because you may set up self-resonances that cause all 
> kinds of fun.

Although if you are only interested in a few bands it's easier.

> The best approach IMHO is to model the antenna-feedline system and see what 
> the actual shack-end impedances are. Or measure them. Then decide what tuner 
> setup is needed to do the matching job.

Modeling it is harder than it looks because all kinds of things affect 
the real antenna, and sometimes the impedance at a particular distance 
along the line from the antenna changes rapidly. My antenna was an 88' 
inverted V with the apex at 40 feet, so the ends were not that far from 
the ground, causing incorrect modeling results on the lower bands.

I have an Autek VA-1 antenna analyzer which is small and has a 
self-contained battery, so I just put it on a large carton and hooked it 
directly to my parallel feeders. An MFJ with an AC supply got incorrect 
results because of capacity to ground. I was lucky, and on the bands of 
interest, 80 and 40 meters, I got r ranging from 15 to 160 ohms and x 
was inductive in both cases. So I chose a value of capacitance which 
canceled the inductance on 80 (the band on which tuning was the 
sharpest) and which reduced the reactance on 40 significantly (and also 
incidentally changed the sign).

Once I did this I got good results from a 1:1 balun. I tested it by 
attaching a thermocouple to the core and measuring the temperature rise 
before and after adding the reactance compensation. Without the 
capacitors, the temperature went up several degrees when I transmitted 
at 100 watts for one minute. With the capacitors, there was no 
measurable temperature rise.

The SWR seen by the radio was below 8 to 1 throughout the CW bands on 
both 80 and 40 (as low as 2:1 at one point on 80) so the internal tuner 
had no trouble matching it.

I also checked for balance by using an RF sniffer on the feedlines at 
the exit of the balun and got good results.

Performance seems to be no worse than with the Johnson Matchbox -- and 
now I can operate on 80 and 40 meters with this antenna without tuning 
anything, something that makes life easier in contests.

The balun I used was a huge home-made choke with six (!) ft-240-31 cores 
wound with 8 turns of RG-58/u. This is only intended for the low bands 
and would probably be poor on the high bands, but the Elecraft balun 
should work well on all HF bands.

There's an article in the latest QEX that describes a method of using a 
combination of a voltage balun and a current balun to get better 
performance in situations characterized by wide ranges of impedance. I 
would have tried this if I hadn't been able to meet my needs as 
described. The Elecraft balun is a current balun, and one could add a 
simple voltage balun. Someone should try this!
-- 
73,
Vic, K2VCO
Fresno CA
http://www.qsl.net/k2vco