high freq vco

[Magnus Danielson]

From: Scott Gravenhorst <chordman at earthlink.net>
Subject: Re: high freq vco
Date: Sat, 12 Aug 2000 09:49:59

> Just yesterday I played with this very thing.  I was not able to
> make a 4046 track more that a couple of octaves.  However, I
> had R1 at 10K and R2 at infinity giving a large range.
> I am thinking that linearity may improve if I restrict the 
> range some???

In my lovely RCA 1983 CMOS databook (I use it far more than any other!) there
are a pair of graphs for how the linarity of the CD4046B depends on R1 and C1.
Similar graphs exists for the CD4046A. THE GRAPHS DIFFER!!!

For the 4046A we see that a larger C1 and a higher R1 yields lower linearity
deviation values.

However, I run under the suspicion that it is really the input MOSFETs
nonlinearity that causes problems. If one pulls the expo current directly on
pin 12 that current run directly into the builtin current-mirror and gets
directed to the CCO core.

Hmm... where did I put my 4046s?

> Anyway, I need it to track a FatMan accurately
> enough for music, about 5 octaves.  This would be an ideal
> VCO for my purpose since what I really need is to use the 
> square output to drive counters for division and/or clock
> a pseudorandom noise generator.

Ah, yeah... the 4046 should be able to do that, just needs to check the
tunability then.

Cheers,
Magnus