If you wind a 4:1 Guanella balun as two identical 1:1 chokes on a common
core, you force the CM voltages across the two chokes to be the same
because they share the same flux. If we call the input voltage V, the
balun then drives the output terminals to +3V/2 and -V/2 with respect to
ground. If you had a perfectly balanced load that would force 3 times
the current into one leg compared to the other. Analysis here:
http://www.karinya.net/g3txq/temp/4%20to%201%20current%20balun.png
and here's the experimental demonstration on a 'scope using a perfectly
balanced 200 Ohm load:
http://www.karinya.net/g3txq/temp/bal1.jpg
http://www.karinya.net/g3txq/temp/bal2.jpg
Because of Sevick's pre-occupation with differential-mode performance he
never spotted that effect. In fact he made lots of DM measurements - DM
impedance, phase delays etc - but with a *floating* load; andas we all
know, if you have a floating load you don't need a balun to maintain
current balance :)
The simple fact is that a 4:1 Guanella balun wound as two 1:1 chokes on
a common core will force voltage and current *imbalance* with any
non-floating load; I don't call that "working" !
For a 4:1 Guanella balun to "work" properly and drive equal currents
into any load, the CM voltages of the two chokes must be able to take up
different values; they can't do that if they share the same magnetic
circuit.
73,
Steve G3TXQ
On 24/04/2013 21:30, Rick - DJ0IP / NJ0IP wrote:
Steve, Sevick is right. You "can" wind a 4:1 Guanella balun on a single
core. I guess the question is, how well does it work?
I have tried it both ways. In fact I was using dual core's for a couple of
years until I read that there is very poor tolerance in the cores... up to
20 % variance. Since I have no way to measure this and don't know where to
buy matched pairs, I begin to wonder how do I maintain symmetry when there
is 20% variance. So I can't decide whether it's better to use 1x FT-240 or
2x FT-140. Any ideas on this?
73
Rick, DJ0IP
_______________________________________________
TenTec mailing list
TenTec@contesting.com
http://lists.contesting.com/mailman/listinfo/tentec
|
