[sdiy] filtering out a saw-reset glitch in a triangle shaper

[David G. Dixon]

> A simpler solution might be to try a "crappy" op-amp like a 741 or
> 1458. Built in slew limiting, and they perform beautifully in
> waveshapers like that. You'll also see these also used in tri to saw
> shapers and output buffers for older wavetable VCOs. In fact,
> sometimes people get in trouble replacing EVERY op-amp in vintage
> synths with higher spec parts, only to find that the synth now sounds
> horrible!

I've tried both 741 and 1458 in the simulation.  The 741 was definitely
better, and the 1458 was marginally better, but both needed the LPF to
completely eliminate the glitch (although the 741 just about did it without
it).  I've decided that 1k + 2.2nF is the way to go.  The glitch almost
completely disappears, even with TL072 (13V/us).  I've also got two 22pF
caps in the rectifier circuit as well, and both are important, no matter
which opamp I try.

If the circuit behaves anything like my triangle-core VCO's saw-shaper, then
the actual circuit will be less glitchy than the simulation.  I think it
will be fine.