toggle quoted messageShow quoted text
Hi -- take a look at the PDF for better description. Basically, the way we did ours is an AUTOtransformer; so this refers to having 28 total turns, and taking a 50 ohm tap at only 4 turns up from ground (common) The 28 turn end will reflect 2450 ohms (theoretially).
A 7:1 turns ratio causes the voltage to go up by 7, and the current to go DOWN by 7 (conservation of power) and the result is the impedance goes up by 49. (49*50 = 2450 ohms).
I hope that better explains it. If you don't have enough inductance in the 50 ohm part, you have losses....if you have too many turns you have other losses (?capacitance? i'm not sure) and there are losses in the ferrite. Unfortunately I'm not enough of an expert to do much more than build two devices, put them back-to-back across a spectrum analyzer and measure the total loss, divide by 2. Getting the right number of turns experimentally is the best that I can do. There may be a better core than type 43, but that is the best I have done so far. Forget iron cores for this particular application....for the lower bands.