Master-Slave VCO (was: RE: on oscillators)

[Haible Juergen]

	>:::Wasn't Juergen Haible building a triangle core vco with a second
timing cap
	>:::(with either half the charge current or twice the capacitance)
to generate a
	>:::glitch free sawtooth, synced to the start of the tri wave cycle
?
	>:::Juergen ?
	>
	>This would basically mean a second slave oscillator with hard sync.

Yes - works nicely. I still have the veroboard somewhere.
I discarded it for a different reason: Tracking problem of the (master)
triangle
VCO. I thought it was something about the expo current source (I think
I tried something different here as well), but meanwhile I'm almost sure
that's
because the OTA overdrive problem we just discussed a few days before.
But the master / slave core works ok.

	>Wouldn't a controlled slew rate limiter eliminate the saw glitches
	>in a tri - saw converter? I.e. slew rate slow in the charging
slope,
	>and fast in the discharge.

That's an excellent idea.

BTW, it's possible to do a pretty good glitch-free triangle-to-saw
conversion
without this trick, if only you're using the right components. At least I
was quite
satisfied with the circuit that went into the MOTM VCLFO (;->). You will
understand that I cannot go into details in this case (sorry), but I just
wanted
to say it's possible and there is no black magic involved. 
(And the CEM3340 did a nice tri-to-saw conversion too, didn't it ?) 

JH.