Ringmodulator (magnetic)

[Martin Czech]

I'm currently trying out different ring modulator (ie. four quadrant
multiplier) designs. First was the good old transformer/diode-ring
thing. Now, the transformers have good common-mode supression (?) and
also good bandwidth. I also tryed to match the (Schotky) diodes. But I
couldn't get an acceptable carrier supression.  I could try Germanium
diodes (though I live in Germany, they are hard to get ;->>), cause I
think one reason for the misbehaviour could be the very strong nonliearity
of the Schottky d.. I tryed the two and three transformer configuration,
well, the two transformer thing is cost saving and one can apply a dc
offset into the diode ring for better carrier suppression, so I think
that the carrier transformer can be ommited (with the implication that
an op-amp might be needed, no passive cicuit any more). Mmhh, I should
check the trafo common mode suppression again, but I still think that I
should do something with the diodes. If the carrier ihas one polarity,
two diodes are open, in the other polarity the other two. Now, I could
add two trim pots, one for each path, and maybe two fixed resistors for
the other two diodes and see, if symmetry can be optimised.

The diode ring seems to introduce a lot of distortion, this can be a
nice effect. So I think it's still worth to stay magnetic.

Any ideas ?

m.c.