Hi Peter,
Thanks for this forwarded message!
So I noticed my first mistake: crystal filter has ~200 ohms Zin/Zout - not 50, as I thought before.
Ok, so that means, attenuator has right Zin/Zout, but there is an impedance mismatch between 1st AMP (just right after LPF) and mixer.
This AMP translates 50 ohms of the LPF to ~200 ohms, but mixer expects 50 ohms. Maybe that's why we can notice much improvement in the way BITX works, when we remove that 1st AMP completely - then mixer has 50 ohms on all ports.
But I don't understand two things:
Second the terminal impedance of the crystal filter and the various
RF/IF amplifiers were optimized for 200 ohms the mixers are better at
50 ohms and there is no impedance matching to correct for that.
Why mixers are better at 50 ohms?
I thought there is not much difference in mixing for different impedances on mixer ports, but the point is to keep them EQUAL on each port (nevermind if it is 50 ohm or 200 ohms - mixer will work fine as long as those impedances are equal on each port). Is that correct?
And second thing: why RF/IF are optimized for 200ohms?
As it was explained by the author of that project, each AMP translates impedance: 50ohms -> 220ohms and 220ohms -> 50ohms (for resistances used in the circuit).
So I think there is no mismatch anywhere except this one between 1st RF AMP and mixer. Is that correct?
Additional question at the end: is that a good idea to use ADE-1 (SMD level 7 double balanced mixer) instead of those 1n4148 diodes and "manual" transformers?
Regards,
Rafal SP3GO
