multiplier: was Re: 3280 and AD533 availability, EN frequency shifter

[Martin Czech]

This is one of my projects on the long list:

How good are the available multipliers? Noise, inter modulation,
linearity, drift, carrier feed through?

There is:

RC4200 (exp/log, expensive, and one quadrant, can be extended to 4)
MC1495 (linear inputs, why isn't this available in Europe??)
MC1496 (nonlinear input for modulator, ie. expects square modulator)
AD633  (crap, if you ask me)
diode ring (original ring modulator, with or without iron)
diy    (using dual transistors)

and then there are very expensive AD chips.

There are basically three circuit topologies: symmetrical controlled gain
cells, exp/log structures, and diode rings.  exp/log  being only one
quadrant (without circuitry tricks).  I must say that I trust more in
the gain cell thing. Most chips work that way.


I have tried the complete diy route, without dual transistors.
Excellent noise and bandwidth, but drift.

I have tried AD633 and found it too noisy.

I've bought 4200, but expect problems due to the unsymmetry of the
circuit. It is basically one quadrant.  We'll see.

The best thing would be a 1495, same topology as 633 or my diy, but
unavailable for some mysterious reason.

The diode ring suffers from distortion, noise and feed through.
Could be interesting for lofi sound experiments. I've
tried a version with audio transformers. That's pretty expensive
and you get low frequency problems as well. It is possible
to go without iron, using differential op amp schemes.
Found that in an old "Funkschau". Certainly a way to explore,
sounds really different than a clean , linear multiplier.

The 1496 is available and cheap. It seems to be used for radio.
Unfortunately the linearizing circuitry is missing, because
it is intended for square wave carriers. But, I think it is possible
to add a linearizing stage with dual transsitors. One dual as diff
amp, the other as loging diode.

m.c.