[sdiy] Low Frequency Square to Sine Waveshaper

[David Moylan]

Paul Perry wrote:
> Just overdriving the input diff pair, I believe.
> 
> Perhaps you would be able to use an OTA as a variable gain stage, and make a
> square to sine thusly:
> 
> 1. run a square wave to an integrator. 
> This gives a triangle (of unspecified amplitude).
> 
> 2. run the triangle into an OTA, detect the level of the output, compare it
> to a reference, and use the result to adjust the gain of the OTA so that the
> triangle is amplified to just reach the amplitude of the desired sine.
> 
If you wanted to try this route maybe check out the LM13700 datasheet. 
There's a circuit in there for an AGC amp that uses the buffer 
transistor to replace the diode of an ordinary peak detector.  The peak 
level is fed back into the input diodes which lowers the gain.

I forget what the initial application of this thread was, but there are 
probably better ways to get the sine than trying to ride the gain. 
Triangle to OTA overdrive is probably the most typical for a reason, but 
some of the other methods seem good too.

Dave




> Paul perry Melbourne Australia
> 
> -----Original Message-----
> From: David Moylan 
>> Is the overdrive characteristic of the OTA just an extension of 
>> overdriving the input differential pair or is there a different 
>> phenomenon which makes them good for creating sines?
> 
> 
> _______________________________________________
> Synth-diy mailing list
> Synth-diy at dropmix.xs4all.nl
> http://dropmix.xs4all.nl/mailman/listinfo/synth-diy
>