AW: Saw-to-Tri converters

[Haible Juergen]

The one-transistor method isn't that bad. 
Roland used a second stage (emitter follower) to compensate
for the Vbe temperature drift.

One funny converter is in the MS-20. Looks increadible stupid at first,
because it doesn't have any compensation for the diode drops.
But when you realize how it works, it's quite clever: It does *not*
produce a triangle from a saw wave, but one of the triangle's
corners is quit rounded, because of the diode thresholds. Now, when
you consider that a tri wave is mostly used as a poor man's sine
in VCOs, the Korg method isn't bad anymore !

JH.