AW: mc1495 ic

[jhaible]

> Also MRM uses sinewave modulator, so RM needs
> lin input wich 95 has, and 96 not, so it'd use square (which is
> implemented
> in NE565 chip


I don't know anything about the Oberheim RM. But in general,
the 1496 can surely be used with sine carrier. It's just that the
carrier input is not linearized, so you either keep the amplitude
down to 20mV, or you feed a triangle wave into the input with
90mV, to have it crudely shaped to a sine wave by the 1496.
(If you can live with 3% THD)

[Just to avoid confusion, what I've done in my Frequency Shifter
is another (different) option: I did the sine shaping with the
*linear* input (by carefully selecting the gain and overdrive)
which results in lower THD. Then I used the carrier input
(the nonlinear one) for the signal (<20mV).]

What I described in the first paragraph, carrier input <20mV,
modulator input linear, is what I've sucessfully tried in the JH-3
ringmodulator and in the Synthi Clone. I think several Roland
synths used this simple scheme as well. And the original VCS3
RM is basically the same circuit built from transistor arrays.

I don't remember who brought up the rectangular modulator
stuff (and I'm sure there is such a background in RF devices
that do work with switching modulators), but it definitely
works with any kind of signals. It's just the usual stuff of linearized
vs. non-linearized inputs, like CA3080 vs. CA3280 for instance.
Noise, THD, offset voltage compromise.
In the JH-3 ring modulator, I used the nonlinear inputs of the 1496
for an (optional) overdrive feature. Resistor divider scaled to allow
much higher levels than 20mV, and input attenuator pot to tame the
levels for linear operation. Not as noise-free as a 1495 I bet, but
good enough for my needs.

JH.