A fundamental problem of exponential VCOs?

[Eli Brandt]

Don Tillman wrote:
> A third possibility is to build a seperate module that performs the
> Vout = log(exp(Vcv)+Vdetune) operation.  But that's not elegant at
> all. 

Unless you're already doing digital control, in which case it's just a
table lookup on the pitch to get
        deltaPitch = ln(exp(Pitch) + deltaHz) - Pitch
                   = ln(1 + deltaHz/Hz)
I believe Juergen has mentioned doing this in the Matrix-N family...

On the other hand, linear FM is good to have for other purposes.

-- 
     Eli Brandt  |  eli+ at cs.cmu.edu  |  http://www.cs.cmu.edu/~eli/