[sdiy] Temperature Compensated Exponential ConverterUsingSSM2164

[David G. Dixon]

I think that the temperature dependence of the voltage source to the tempco
voltage divider is basically a non-issue.  Even if it is 3300ppm/K, by the
time the voltage is divided down (particular if it comes from a 15V rail),
this temperature effect will be very small.  For example, if we divide 10V
down to, say, 300mV, then the temperature coefficient is divided from
3300ppm/K to 100ppm/K.  This is on par with the temperature coefficient of
the metal film resistors.

> Maybe you could use a voltage reference that has a negative temp
> dependence ? And then adjust the tempco in the expo circuit itself to
> account for that?
> I.E. the tempco in the expo circuit goes half of the way, the voltage
> reference's 'temp compensation' does the other half ... not sure how
> that would work though :) but it seems it's easier to balance out the
> imperfections of a circuit, than it is to find parts that are
> absolutely perfect.
> 
> D.
> 
> On Sat, Sep 5, 2009 at 7:31 PM, Jerry Gray-Eskue<jerryge at cableone.net>
> wrote:
> >
> > <<So, is it best to divide the tempco voltage input to the buffer
> directly
> > from the rail, or to drop it across a zener first?>>
> >
> > <<<I also know that the effectiveness of the tempco relies on
> > coming within about +/- 250 microvolts of the optimum tempco voltage.
> >>>
> >
> > Don't forget that common Zeners have a Temperature Coefficient too. You
> can
> > however get the composite (Zener/Silicon set) type (like the LM329 used
> in
> > the Ian Fritz Dial-a-Temp co converter) that has temperature
> compensation
> > built in, or use a true voltage reference part.
> >
> > If this temp co voltage is this twitchy you want as stable and
> temperature
> > independent reference voltage as possible.
> >
> > - Jerry
> >
> > -----Original Message-----
> > From: synth-diy-bounces at dropmix.xs4all.nl
> > [mailto:synth-diy-bounces at dropmix.xs4all.nl]On Behalf Of David G. Dixon
> > Sent: Saturday, September 05, 2009 1:17 PM
> > To: 'cheater cheater'; 'synth-diy'
> > Subject: RE: [sdiy] Temperature Compensated Exponential
> > ConverterUsingSSM2164
> >
> >
> > That is an excellent point!
> >
> > Incidentally, I have looked at the temperature sensitivity specs of
> metal
> > film resistors (~1.5 ppm) and 18-turn trimmers (~100 ppm max), and as
> long
> > as the bulk of the voltage drop is taken across the metal film resistor,
> the
> > effect should be minimal.
> >
> > Also, we mustn't forget that the control pin of 2164 has a 5k input
> > impedance, so the tempco voltage must be buffered, particularly if
> multiple
> > expos are to be hooked up to it, as in a polyphonic application.
> >
> > So, is it best to divide the tempco voltage input to the buffer directly
> > from the rail, or to drop it across a zener first?
> >
> >
> >> But what you get in exchange is the possibility to have multiple
> >> expo's which drift in the same way. Which makes it perfect for poly
> >> synths.
> >>
> >> D.
> >>
> >> On Wed, Sep 2, 2009 at 1:26 AM, David G.
> Dixon<dixon at interchange.ubc.ca>
> >> wrote:
> >> >> I think you need to trim it since values varies and repeatability of
> a
> >> >> design needs to be achieved. Also, how do you repeatably achieve
> >> >> 0,2678345615807 V or whatever we crank out as the optimal voltage?
> >> >
> >> > Yes.  What I was arguing was that the adjustment could be confined to
> >> one
> >> > trimmer, specifically, the one which gives the tempco voltage.  The
> >> > appropriate feedback resistance could simply be calculated and
> applied
> >> to
> >> > the nearest 0.1k.  I believe that I have demonstrated successfully
> that
> >> the
> >> > trimmed values are more or less insensitive to the VCA gain
> temperature
> >> > coefficient, and that the optimal feedback resistance at the expected
> >> value
> >> > of the VCA gain coefficient will still serve admirably for tempco
> even
> >> if
> >> > that gain coefficient is several percent off spec, which is highly
> >> unlikely
> >> > for 2164s.
> >> >
> >> > Having said that, I also know that the effectiveness of the tempco
> >> relies on
> >> > coming within about +/- 250 microvolts of the optimum tempco voltage.
> >>  Even
> >> > being 1 mV away from the optimum puts the first octave about 0.25%
> out
> >> of
> >> > tune even at the reference temperature.  This extreme sensitivity is
> >> > probably the "Achilles' heel" of the design.  Drift in the supply
> >> voltage,
> >> > and even temperature sensitivity of the resistor and trimmer, could
> >> become
> >> > serious issues.
> >> >
> >> > _______________________________________________
> >> > Synth-diy mailing list
> >> > Synth-diy at dropmix.xs4all.nl
> >> > http://dropmix.xs4all.nl/mailman/listinfo/synth-diy
> >> >
> >>
> >> _______________________________________________
> >> Synth-diy mailing list
> >> Synth-diy at dropmix.xs4all.nl
> >> http://dropmix.xs4all.nl/mailman/listinfo/synth-diy
> >
> >
> > _______________________________________________
> > Synth-diy mailing list
> > Synth-diy at dropmix.xs4all.nl
> > http://dropmix.xs4all.nl/mailman/listinfo/synth-diy
> >
> > _______________________________________________
> > Synth-diy mailing list
> > Synth-diy at dropmix.xs4all.nl
> > http://dropmix.xs4all.nl/mailman/listinfo/synth-diy
> >
> 
> _______________________________________________
> Synth-diy mailing list
> Synth-diy at dropmix.xs4all.nl
> http://dropmix.xs4all.nl/mailman/listinfo/synth-diy