"ralph sansbury" wrote in message ...
George, I see that if the effect of one gate was to reduce the
current by 1/2 before passing through the region controlled by
the other gate which also reduced the current passing under its
control by the same fraction that the resulting current would
then by 1/4 or the product of the two gate voltages. Is this what
you meant by the multiplication coming first?

Sort of, but remember charge is conserved so there cannot
be a difference in current in different parts of the channel.
It is like a pipe with two taps, one after the other. To get
water out you have to turn on both but the rate depends on
the angle you turn each tap. In a FET, that relationship
includes a square term which we can use to contrive a
multiplication function.

But I still have a
problem of understanding how a filter circuit after this could
output one of the orginal sequences of voltages on one gate?

It doesn't. The original sequences are the antenna signal
and the reference sine wave. The multiplication manufactures
the sum and difference frequencies out of those, and the
filter then discards (say) the sum but leaves the difference
unaffected.

What I meant by saying multiplication comes first is that
until you multiply the voltages, there are no sum and
difference terms for the filter to work on. Look at the
block diagram:

multiplier filter(s)
+---+
a(t) ----| | +---+
| * |---------| ~ |-- d(t)
b(t) ----| | c(t) +---+
+---+

Signal c(t) is created by multiplying a(t) by b(t). That
sequences of voltages is actually the product of a and b,
but it can also be considered the superposition of two
signals at the sum and difference frequencies. The filter
then reduces the amplitude of the unwanted sum term to a
negligible level leaving only the signal at the difference
frequency.

George