[sdiy] expo converter....clues for the sleuths

[harry]

No... its correct.

The most negative voltage is giving you the least current, which would
mean the
lowert LFO frequency.  The missing part is that this is the bias trim.
You'll set this
to some operating point, then vary the OTHER voltage input to get the
frequency
you want. It has a much bigger resistor, so the change per volt will be
a lot less.
With proper adjustment, you should be able to cover a wide range.

EFM (hi tom!) projects are usually a 0-5V or -5 to +5 range.  I like to
go further,
so I adjust mine to suit myself.

Your expo seems to work OK. You will find examples of a one transistor
expo
current source (sometimes used in EFM resonance control). This is the
cheapest dirtiest method. It does not work with voltages less than 0.7V
and that varies with
temperature.

The two tranny unit (this one) is much better... the diode connected
tranny compensates for that .7V drop, allowing full range. If they are
thermally connected
(like a 2SC1583 must be...) the drifts of each transistor cancel out.
This still does not
make it OK for a VCO.... you would still need a tempco to correct. But
it would be much better than #1 above.

The best expos have two trannies AND an op-amp. Still need a tempco, but
the
opamp will keep the trannies biased at exactly the right point.

Yhen you'll find some really esoteric ones... like the new PAiA and
similar OTA based expos that don't need a tempco resistor at all. I
haven't tried them yet... National semiconductor published this maybe 20
years ago... but called it a log converter by mistake... so we all
ignored it. John Simonton of PAiA somehow noticed it was really an expo
converter, and whizzed us all !!!  Way to go John...

H^) harry

Benjamin Smith wrote:

>
>
> first thanks to all who replied, harry your drawings were very
> helpful. i have tested the circuit with the output of the transistor
> pair connected via a 470k resistor to the negetive supply, and made
> the following observations;
>
> var res at 100%,  -15V on the wiper, -660mV after 10k res, -13.61V at
> collecter of output tarnsistor, therefore 1.3uA output.
>
> var res at 0%,  15V on the wiper, 660mV after 10k res, 0.65V at collecter
> of output tarnsistor, therefore 31.5uA output.
>
> var res at 25%,  2.66V on the wiper, 1.10mV after 10k res, 0.65V at
> collecter of output tarnsistor, therefore 31.5uA output.
>
> the relationship is opposite to the one expected but otherwise seems
> to function correctly. from watching the dynamic response as i turned
> the pot the relationship appears exponential with the asymptote at
> 0.65V. i placed the transistors into sockets so that i could compare
> different devices, the only difference seems to be in the input
> sensitivity, with some the full range of the variable resistor is
> needed to reach the extremes, with others the max output is reached
> before the full travel of the pot is used.
>
> any suggestions?
>
>
> on a similar topic i intended to build the ca3046 circuit described by
> RA Penfold at
>
> http://aupe.p
> ys.andrews.edu/diy_archive/schematics/synths/penfold/penvco.html
>
> any problems replacing the BC559 with one of the 2n3906 that i have so
> many of? any considerations i should make powering it from 15V instead
> of 12?
>
> thanks very much
>
> ben
>
>
> -----------------------------------------------------------------------
> Get Your Private, Free E-mail from MSN Hotmail at
> http://www.hotmail.com.