opamp rescaling/variable saw-tri-ramp lfo

Rob cyborg_0 at iquest.net
Sun Nov 14 19:46:42 CET 1999

----- Original Message -----
From: Paul Perry <pfperry at melbpc.org.au>
To: <synth-diy at mailhost.bpa.nl>
Sent: Sunday, November 14, 1999 7:09 AM
Subject: Re: opamp rescaling/variable saw-tri-ramp lfo

> At 01:36 PM 13/11/99 -0500, "Rob" <cyborg_0 at iquest wrote:
> >Im actually considering using the old 7400 nand gate type oscillator as a
> >pwm square wave output source then feeding that into an integrator to get
> >what was discussed earlier as far as a variable saw-tri-ramp output so I
> >save components and panel space.
> >
> >The variable slope integrator idea was a godsend. Good that you all
> >discussed it! ;)
> If you try to make a variable slope triangle by integrating a variable
> rectangle wave, you won't get a constant amplitude.
> paul 'spoilsport' perry Melbourne Australia

Oh, yeah, I know about that. ;)) Thats merely cause you arent applying the
same amount of voltage over time... Also, when the frequency goes up, so
does the RMS voltage..

So, the two conceivable things that I have to overcome are as follows:

1) When the positive duty cycle decreases, so does the RMS voltage,
therefore less voltage is applied to the cap.. When the duty cycle
increases, more charge is stored

2) Same goes for frequency: More frequency = more stored charge, less freq =
less stored charge.

So, perhaps a voltage follower after the integrator?  Oh, this could be
fairly easy. I could just use a final stage buffer after the integrator and
"sense" the voltage that is used to set the input reference on the PWM amp.
Then, using this voltage, I could manipulate the feedback loop of the final
stage buffer using a fet and compensate for the changes in amplitude.

It wont be very temperature stable, but it should work. Any other ideas?

I have also thought, too, that it might be just as well to not add a final
stage and just manipulate the feedback loop of the integrator itself to


More information about the Synth-diy mailing list