Topband: Flag Xfmr Measurements End Game

[K9AY]

> We have to be very careful adding reactances in series with the 
> common-mode impedance of a system to reduce unwanted current. 
> In many cases, we can make the system worse by adding reactance.
> 
>     73, Tom W8JI
-----

I can verify Tom's above assertion. When designing my TRX-16
16:1 transformer, I tried making a two-stage design -- a low-loss
(tightly coupled) 16:1 XFMR plus a choke 'balun.' It did not work
well at all -- the choke had only small effect on isolation. The final
design is a loosly-coupled tranformer with separate primary and 
secondary windings, wound on a binocular core made of two 0.5"
OD x 0.4" long cores.

Two transformers back-to-back have a total loss of <1 dB, or 
about 0.4 to 0.5 dB per XFMR. The match to a 820 ohm carbon 
resistor is always under 1.2:1 throughout an 'assembly line' batch.
For comparison, my TRX-9  9:1 XFMR uses tightly coupled
windings (but still separate primary and secondary) and has about
0.2 dB loss.

73, Gary
K9AY
k9ay@k9ay.com
www.aytechnologies.com