[Elecraft] how did I calculate the loss?

[Charles Greene]

John,

I sorta backed into the design.  I first made a balun; the final design was 
a little different.  I tried it on my W3EDP antenna at 100 watts.  The 
W3EDP antenna is known to have high impedances on some bands, and the SWR 
measurement was higher than 5:1 on some bands, but the balun did not become 
hot as it would have if it had any but minor losses.  So I figured out how 
to measure the efficiency by placing the balun in a thermos and measuring 
the temperature rise during a timed operation.   I measured the efficiency 
to 1% accuracy.  Subsequent improvements in technique improved the accuracy 
to 0.1%.  I tried the 4:1 balun on my G5RV where it worked fine at 100 
watts, but replaced it with a 1:1 balun as that was a better match.

The balun offered as a kit by Elecraft is a better design than the Low 
Power Balun which is flat only through 35 MHz.  It is flat enough to be 
useful on 6 meters.  It is unique, and I highly recommend it.  I lately 
have learned how to design a 4:1 balun using a single core, and I built one 
in a tiny plastic box weighing slightly over one oz. in a 135' OCF (Windom) 
that gets hoisted to the top, with a RG58 feed line.  It works fine with a 
K2/100 and antenna tuner on most bands.  When I get around to it, I will 
put this design on N0SS's web site

A balun usually does not decrease the losses in a system.  It enables you 
to use low loss twin lead then transform it to unbalanced near the antenna 
tuner and eliminate lossy coax.  The twin lead will carry RF without 
excessive loss and radiation when the SWR is high on the feed line without 
radiation; the coax will not.

Some one mentioned that the fact that a current was balanced or not on a 
twin lead or open wire feed line was dependent on the antenna and not the 
source.  If balanced, there should be no radiation.  Hooking a twin lead to 
a KX1, one side to ground does not change this, therefore there should be 
minimum radiation from a 300 ohm ladder line, one side connected to ground 
and one side to the business end of a KX1 antenna tuner (or any other 
antenna tuner).  To expand this a bit, this holds true for an end fed zepp, 
whether or not the 300 ohm ladder line is elevated or just laying on the 
ground, as in a portable installation.  This gives some possibilities for 
antennas for the KX1 involving getting a wire into the shack without losses 
and radiation as it departs from the part up in the air.

Concerning saturation.  I have a whole mess of baluns I have been able to 
saturate in certain situations, even at powers on the order of 15 
watts.  The situations involve high differential mode current on a 
coax.  It has never happened on feeding an open wire feed line, or on an 
dipole or OCF antenna.  When saturated, they lose effectiveness, and the 
wire and core get hot.  Unless heated to over 135 deg C, the ferrite 
material will recover after the RF current is removed.  More likely, the 
balun will fail due to over voltage.  The magnetic wire varnish insulation 
is good for about 500 volts peak, and the insulation on the ferrite 
material the same, so this should give you some idea of the safe operating 
impedance and power.   Be mindful that there is always the possibility of a 
nick in the insulation during construction.  Larger baluns capable of more 
power use glass tape and teflon insulated wires.

I realize this line has deviated from the original subject (sorry Eric), 
but some suggested feeding the 8030 with open wire line and a balun, and 
one thing lead to another.

At 02:15 PM 12/16/2003, John Buck wrote:
>Hi Charles,
>
>I enjoyed reading the documentation you referenced at 
>http://www.njqrp.org.  This is great construction and test documentation 
>of an excellent design.  It is recommended reading for anyone considering 
>building a balun.
>
>As I read the documentation, the balun you studied was terminated in 50 
>ohms on the input side.  When used in a typical K2 application with 
>internal antenna tuner, the input and output impedance may be much higher 
>(or lower).  What can be said about the performance when driven by and 
>antenna tuner to match impedance's such as found with a G5RV type dipole 
>fed with various lengths of 300 or 450 ohm twin line?  My problem is with 
>the varying conditions for multiband use.  Yes properly applied Baluns are 
>not very lossy.
>
>In the case you quote, it appears that you were looking at the efficiency 
>of the 4:1 balun when terminated with 50 ohms on the input and 700 ohms on 
>the output.
>If used with the K2 KAT2 or KAT100 the input impedance might be nearer 200 
>ohms.  What impact does this have on the loss and on the saturation 
>characteristics?
>
>Aloha,
>John KH7T
>
>Charles Greene wrote:
>
>>It's another of those bits of folk lore, that a balun is lossy except at 
>>the design impedance.  I did some extensive loss measurements on a 4:1 
>>current balun working into 700 ohms at a high reactance instead of 
>>200+j0, and found the efficiency was 94% (it was 97% on a 200 +j0 ohm 
>>load).  Find a place in the open wire line where the SWR is less than 5:1 
>>and place your balun there.  My tests are documented at: http://www.njqrp.org/
>

73, Chas,  W1CG
K2 462b, 3571