Ring modulator froma few op-amps?

[brett]

Hi. Another topic went a bit off-target and wound up here:

http://www.pat2pdf.org/patents/pat2890832.pdf

Figure 12 looks like a very nice modulator (with an extra "sweep" function for extra modulation ??). (The weird little valve-like guys are diodes and the triangles are inverting amplifiers.  This is pre op-amps)

Although more complex than your passive ring mod, it might be achievable.  

cheers

[lovekraft0]

Looks like a good app for a 4049 and a few germaniums - might be interesting! Are the resistors all the same (ie, is it just a passive mixer, more or less)? Is the sweep input for forward-biasing the diodes, or something else?

[puretube]

OT ?

that one actually multiplies the input with non-linear ("dioded") derivates...


AND: since s.o. complained the other thread getting "OT",
may y`all look here, where that whole diode-stuff originates from (1924 !).

Sorry, Brett,
for stearing off target again...
:wink:

[brett]

The sweep just pans between the X and Y inputs.  Whoever designed this definately was thinking of a Univibe.   :wink:  
I *think* that the resistors associated with the negative feedback and the diodes would have to be the same, and chosen to get particular curve shapes (parabolas) out of the amplifiers.
The mixing resistors could probably all be equal, unless there's something I can't see or don't understand (a fully normal situation).  
The 4049 idea is IMO extremely smart.  Six amps, all inverting!
cheers

[Paul Perry (Frostwave)]

That patent revolves around 'dithering', that is to say, adding a small high frequency signal to the input of a function generator, to give a smoothed rather than stepped output. Think of it this way, imagine that you drawan approximation to a curve from a bunch of straight lines, now imagine that you do this a bunch of times on the same page, but stating slightly before or after the usual starting point. And then average the lines out, you will see that the result gives a curved approximation, possibly closer to what you wanted.
What has this to do with ring modulators? Back in the analog computer days(when this patent came out) one was to generate a multiplication of two voltages (what a ring mod is!) was the "square law multiplier". What you did, was to take (X+Y) and (X+Y), feed each of these thru a squarign function generator (like the patent is discussing!) and then, take the difference of the results. Do the math.. !!
In a 'classic' diode ring multiplier,(the one with the transformer) the diodes are acting as approximate squaring circuits.
How crude a multiplier can you get away with? Well, the Korg MS20 synth generates a ring mod signal from two SQUARE WAVES!! using an XOR function, now that is really rough.. but musically useful.