Certainly, here’s a brief summary…
Maths Ch1 and Ch4 when triggered (on the trig input) will generate an envelope with a specific rise and fall time configured by the ‘rise’, ‘fall’, and ‘vari-response’ knobs. You can trigger the envelopes with a square wave running at audio rates. Though Maths is still generating an envelope at these audio rates, we’ll call it an impulse train, which is the terminology used in the Mangroves Technical Map.
What makes this particularly cool and unique in comparison to a traditional VCO, is when triggering an impulse train on Maths with a square wave VCO that’s being sequenced to a particular melody, the pulse width of the envelopes (or single impulse) do not change, thus formant synthesis is achieved. Conversely, on a typical VCO generating a sawtooth for example, the width of the sawtooth will scale along with the pitch. This pulse width scaling behavior can also be achieved on Mangrove through the ‘Constant Wave’ switch, and can be approximated on Maths by sending the v/8 sequence not only to the square waves that are triggering Maths but also to Maths ‘both’ input.
What’s rad about formant synthesis in this context, aka generating impulse trains from a square wave, is that when the wavelength of the triggering square wave is less than the rise time of the envelope, the function generator will ignore all rising edges of the square wave occurring after the initial one that triggered it until it the envelope has reached the end of it’s rise and is now falling. This results in subdivision of the impulse train, making it jump around and sound really wild.
On the MW link FluffyDuck posted I included a couple scope screenshots showing the subdivision which helps demystify what’s happening.
