Rbe over temperature?

[Terry Michaels]

Message text written by "Ian Fritz"
>Terry --

Ooops, Terry, not right. Remember that the scaling factor includes a
multiple of dVbe. There is no temperature dependence at the point where the
two Vbe's are the same, and opposite tempco signs (increasing in magnitude
as the magnitude of dVbe increases) on either side.

  Ian

----- Original Message -----
From: "Terry Michaels" <104065.2340 at compuserve.com>

>  Also, it can't be because
> of the difference between 3300 and 3500 ppm, that is a scale factor
issue,
> so if that was the cause you would see an incorrect scale factor setting
at
> all pitch ranges, not just at C7.
<

Hi Ian:

Sorry, I respectfully differ with you.  I said that Grant would see an
incorrect scale factor setting, not necessarily incorrect pitch (or
frequency).  Incorrect compensation for scale factor will cause a VCO to
produce an incorrect frequency at all frequencies other than the point
where the Vbe of the two expo transistors are equal.  It has exactly the
same effect as the 1 volt/octave trimmer being set incorrectly.  All
pitches will be off except at one point, the amount of error will increase
as you go farther either side of that point.  That is a linear deviation
from the correct 1 volt/octave VCO response.

Grant said he had proper tracking at the lower frequency ranges, which to
me indicates scale factor is properly adjusted and/or compensated there. 
Since tracking was off only on the high end of the range, his problem is
due to a change in the Rbe/flyback time compensation.   That is a
non-linear (or non straight line) deviation from the correct 1 volt/octave
response.

Of course this really depends on what Grant meant by "although the C2-C3
range still looks OK. "  If tracking is perfect in that range, the problem
has to be a change in the high frequency (Rbe) compensation.  OTOH,
incorrect scale factor compensation will produce pitches that are very
close to where they should be around the lower ranges of most VCO's which
is usually where the Vbe's are equal.  Very close is not the same as
perfect tracking, however.  The only way to be sure about this is to do
very, very careful measurements of frequency vs. control voltage.

Terry Michaels