AW: [sdiy] Re: OTA nonlinearities

[Czech Martin]

Ok, let's see, I have a network that offers 0- -40dB attenuation (say).
Now I include that into the feedback path.
What will hapen? 0dB will turn into unity gain. -40dB forces the opamp
to have a +40dB singnal out the output in order to stay linear.

So, I think the dynamic range will not be changed, only the control curve
will be mirrored.

Am I missing something?

But I think if you cascade two ideal buffered -40dB networks, you can get
-80dB.
You have to accept the control curve that the device has by law of pohysics,
however, and this won't be linear, in the case of LDR it may be expo.

FETs can only deal with relatively small signals (linearity), even with the
prefered
circuit. LDR get very slow when very dark.

I'm not shure if these methods are good for a general purpose vca.
Limiters or compressors is another story. 
As long as no variable resistor action is required, LDR or fet can be used
for voltage controlled Q in filters. Even in clones this would not change
the sound,
at least I hope so.

m.c.

> -----Ursprüngliche Nachricht-----
> Von:	Anderson, Robert O [SMTP:RobertOAnderson at eaton.com]
> Gesendet am:	Mittwoch, 9. Mai 2001 16:56
> An:	'synth-diy at node12b53.a2000.nl'
> Betreff:	RE: [sdiy] Re: OTA nonlinearities
> 
> From: Czech Martin [mailto:Martin.Czech at Micronas.com]
> >FET and LDR solutions have usually low dynamic range.
> >I don't know what is required here, but even 60dB attenuation may be too
> much crosstalk.
> 
> I usually put the FET voltage divider in a feedback loop.  This provides
> enough dynamic range.
> Regards, Bob A