OpAmp based Ring Modulator thread over at electro-music.com

[frequencycentral]

I thought some of you noise mangler types might be interested in this, which is based on a Harry Bissell design. The topic-starter is a knowledgeable guy, and added an VCA each to input. I'm thinking these could be LFO or envelope follower driven for stompbox use, or even omitted, just leaving the heart of the RM, which is based around a dual opamp. The design runs off a bipolar supply, but again, this could be adapted to 9 volts with a vref for stompbox use.

http://electro-music.com/forum/viewtopic.php?t=39762&sid=03e85c3c86fa501a5850d5e4535ce263

[Cliff Schecht]

One of my first designs was an op amp based real ring modulator. I posted it on here years ago and I actually still use the pedal with my Whammy. Running the carrier and modulation input with the wet and dry signals from a Whammy gives some very useful up and down octave effects, along with all sorts of crazy sound in between the pedal being rocked all the way back or forth. I can post the design if anyone is interested.

I'm also working on a new compressor that is based off of one of my friends new envelope detector design. The thing that separates our design from others is the complexity and accuracy of the envelope generator, which we've been working on for a good 6 months now. Perhaps I can convince him to let me post the design on here when we finish.

[slacker]

I saw Fonik's post over there earlier and breadboarded it, just the modulator part without the VCAs  and its pretty cool.
I haven't tried it with guitar, just some synth modules, but you get all the usual ringmod/nyquist aliaser  sounds if you feed it 2 audio signals and it makes a nice slightly dirty VCA as well if you drop the carrier down to sub audio speeds.
Well worth messing about with if that's your bag.

[soggybag]

Looks like it's based on an OTA to me.

[slacker]

The OTAs just let you change the level of the signals going into it and add DC offsets. The actual ringmod part is the bit on the right hand side made up of the TL074.

[thehallofshields]

I tried breadboarding the Opamp section and hooked it up to a simple oscillator.
Unfortunately the effect didn't work. I heard both the guitar signal and the LFO, but no sum or difference.

Did anyone have any luck with it?

[thehallofshields]

After a little research I understand that what we're looking at is a 4-quadrant Multiplier.

I guess the question that follows is what the operating conditions are, and whether or not it can replace the 4-quadrant Multiplier IC's in circuits like the Maestro RM-1.

[ElectricDruid]

No, this isn't a 4-quadrant multiplier. Whatever else it's doing, it's not doing multiplication as far as I can see.

Instead, I think the circuit is much closer to the original "diode ring" ring mod, the design that gave ring modulators their name. These don't really do multiplication either, instead they're what's called a "switching modulator", I think.

There's confusion over the terminology, since most modern "ring modulators" are actually 4-quadrant multipliers, and they don't use a ring of diodes, but instead are mostly OTA or VCA based. Some people argue that "ring mod" should be reserved for the genuine article, whilst others argue that the name "ring mod" has come to mean something that does 4-quadrant multiplication by whatever means.

I've simulated it and whilst the results are *similar* to what you might see from a 4-quadrant multiplier and might well sound pretty similar, it isn't the same thing at all. I'm still trying to work out exactly what it's doing since it's a big pile of precision rectifiers and mixing.

HTH,
Tom

[thehallofshields]

Thanks for the clarification. Points of curiosity:

-- The precision rectifiers are have the diodes in opposite directions.
-- One input stage is just getting the carrier signal, but the other is getting signals.
-- The resistors are all a convenient 100k, except one in the NFB path with is 200k.

[ElectricDruid]

Thanks for the clarification.

Well, I dunno how *clear* I managed to be...

Points of curiosity:

-- The precision rectifiers are have the diodes in opposite directions.

They do. One gives positive half-cycles only, and the other gives negative half-cycles only. Note that that op-amp circuit with the diodes on its own is only a half-wave rectifier. You need an additional mixer to make a full wave rectifier.

-- One input stage is just getting the carrier signal, but the other is getting signals.

Yes. This is one of the things that's confusing me. A genuine 4-quadrant multiplier is "symmetrical", in the sense doesn't matter which input is modulator and which is carrier. A * B  is the same as B * A after all.
But this circuit seems to treat one differently from the other, which seems odd.

-- The resistors are all a convenient 100k, except one in the NFB path with is 200k.

Not quite - there's two 200Ks, one for the straight path of the upper op-amp, and one as the feedback for the final mixer. This gives the output from that upper rectifier a gain of 2 which, when mixed with the straight signal, makes the upper op-amp a full wave rectifier. Compare with this:



Except that it's also got a bit of the lower signal mixed into it. Odd.

Tom

[thehallofshields]

I've read through the original thread over at Electro Music.

A person there complained that he was only getting mixing, not sum/difference (same as me). They countered that it woudln't work properly without both signals at the same dc-offset.

If you have any suggestions about how much gain to apply to the guitar signal, or a simple sine oscillator that will center around 4.5v, it would help.
I'm eager to see if I can get it to work without the VCA portion of the circuit.

[thehallofshields]

-- The precision rectifiers are have the diodes in opposite directions.

They do. One gives positive half-cycles only, and the other gives negative half-cycles only. Note that that op-amp circuit with the diodes on it's own is only a half-wave rectifier. You need an additional mixer to make a full wave rectifier.

.

Well, I was able to put that part together. But how that creates sum/difference modulation is way beyond me.

[ElectricDruid]

-- The precision rectifiers are have the diodes in opposite directions.

They do. One gives positive half-cycles only, and the other gives negative half-cycles only. Note that that op-amp circuit with the diodes on it's own is only a half-wave rectifier. You need an additional mixer to make a full wave rectifier.

.

Well, I was able to put that part together. But how that creates sum/difference modulation is way beyond me.

Yeah, agreed. I'm not at all sure that it does. It looks to me like it makes a waveform which *sounds a lot like* a ring-mod sound, without actually being one, or producing the strict sum and difference frequencies necessary. Looking at the output on the sim, its very reminiscent of a ring mod, but it's also clearly not a ring mod (and here I'm using the "loose" terminology and meaning a 4Q multiplier).

T.

[thehallofshields]

ElectroDruid, do you think that final inverting Opamp is doing anything special?

I'd like to just replace it with a standard buffer (like a JFET or Non-Inverting Opamp).

[ElectricDruid]

It's the typical inverting op-amp mixer. So its mixing all the bits back together again. Aside from that, it doesn't do anything special. It's not just a buffer though. I suppose you could replace it with a JFET or non-inverting buffer but then you'd basically have a passive mixer followed by a buffer.
What's the need? What's wrong with the perfectly good and very simple op-amp mixer? </curious>

Tom

[thehallofshields]

What's the need? What's wrong with the perfectly good and very simple op-amp mixer? </curious>

I hadn't read up on 'Summing Amplifier'. I thought I might save an OpAmp and use a JFet.