Wave Wraper revisted #1

[Thierry Rochebois]

>Here are some files that I'm playing with. The first two labled WWrap1
WWrap2 are
>the Thierry Rochebois wave wrapper that he posted.

Thanks very much for the files.

> I've cascaded 10 sections
>because as you exceed the limit of foldover for onr stage, the signal
"breaks out"
>in amplitude.

I agree, you can implement four stage with a quad op amp. Thus it is quite
OK.

>It either suffers (or benefits) from the
>soft knee of the diodes. Making the load resistor arbitrarily small helps
stiffen
>the knee, but at the expense of circuit loading.

- Well, soft knee transforms the "triangle" into a sort of "sine" thus the
wave wraper
behaves like a "zero frequency" FM carrier !
- As it is a symetric distorsion, it only adds odd harmonics.
- Maybe, it is possible to add a final waveshaper to compensate the soft
knee.

>I decided on an active diode approach. This allows you to set the diode
knee to any
>level you want (not just .7V),

Maybe, you can achieve almost the same thing by adding amplifiers between
two stages.

>and makes the knee very sharp. You can see this
>circuit in HWrap1 and HWrap2. It is much larger in component count... but
hey...
>today quad op amps cost "chump-change".

Well, yes, but you have to build it...

>I only simulated six
>stages because the sim takes a lot longer... Also notice that the amplitude
>envelope of this circuit is very flat... (again this could be good or
bad...)


Well, I agree, as I said, it is much like "zero frequency carrier" FM. Thus,
the amplitude
of the incoming signal corresponds to the FM depth. Maybe, We have to add an
envelop detector and an output VCA.

Maybe, adding a ring modulator to multiply the output by the input will be
interesting
(it will translate the spectrum)...

>This circuit produces waveforms that llok similar to some of those I saw in
the
>discussion of the Serge wave multiplier. But I still don't understand the
>LM3900....


I do not too. It was also used by ARP for their filters. It seems to be
interesting for
low noise designs...
Is there anybody knowing how to use these "Norton amplifiers ?".

>Varying DC offset may be fun, too.

It will certainly add even harmonics (dissymetric distortion).

>Saw and Square are not useful.

Well, if you add an allpass filter at the input stage...

>Thanks for the cool idea Thierry.

Thanks for the beautifull schematics.

Well, it is time to go back to work (I design DIGITAL effects !)

Thierry Rochebois
DSP engineering
Canam Computers  (Quartz D2D softwares)
http://www.canam-comp.fr