[sdiy] OTA nonlinearities

[Jim Patchell]

    A subject I am fairly familiar with....The primary cause of distortion in an
OTA is the differential input stage.  If you look at the transfer function,
(voltage in vs current out), the curve is a hyperbolic function (sinh, or was it
tanh....can't remember).  The wider the range on the inputs, the more distortion
you get.  If you look on this web page:
http://www.oldcrows.net/~patchell/archives/ca3280_2.html, you will see a
comparison I did of the CA3280 vs the CA3080.  The CA3080 is your normal OTA,
and the CA3280 has a built in linearization network.  Even with the
linearization, I would not call the CA3280 HiFi.  For low signal levels, both
are about the same in distortion.  The CA3080 also exhibits soft distortion,
where the CA3280 is harder.

    Also, I did an OTA analysis.
http://www.oldcrows.net/~patchell/archives/ca3280_1.html, I cannot vouch for the
accuracy, as this was done over 20 years ago, and frankly, I can not figure out
what it was I did.....

    -Jim

Nils Pipenbrinck wrote:

> Hi there.
>
> Just wondering where the non-linearities of the CA3080 OTA are.. I mean.. as
> long as I keep the current on the current-input pin constant, can I expect
> the amplification of the OTA in "hifi"-quality, or does it distort the
> signal?
>
> Or put it different: Under which circumstances will the OTA work nonlinar?
>
> And now back to work...
>     Nils Pipenbrinck