simple opamp based function generator schem inside.

[nelson]

On my travels around the WWW I came across this, thought someone might get some use out of it.

Close enough for rock and roll?

[Steben]

Looks interesting enough for tremolo's.
I wonder how to use it in modulation fx (chorus/vibrato/flangers/phaser), since the triangle is a serious basic for these.

[stm]

Two comments on the circuit:

1) "Sine wave" output is not really a sine but rather a parabola.  The third opamp (bottom up) is an integrator, thus, integrating a triangle wave you get a parabolic waveform.  This might be important or irrelevant depending on your application.

2) The amplitude of the "sine" output varies in proportion to the oscillator's frequency. This again can be troublesome if the oscillator needs to be tuned (especially if it reaches clipping) or not if the oscillator will be fixed in frequency.

[R.G.]

Yep, variable amplitudes on the sine.

There is another variant that is almost as simple, and gives stable sine, triangle, and square amplitudes.

Instead of an integrator on the output of the triangle, you put a high Q filter. That filters the Triangle to a sine. That sine output goes to the comparator for square and then to an integrator for the triangle. However, before the triangle integrator, you have to put a variable attenuator that's driven by the sine wave's amplitude. This forces the sine to be the same size all the time, and it does so bv forcing the triangle to be the same size.

But it does add some complication.

A MAX038 is simpler.

[nelson]

ICL8038 is simpler again. Can still get them from futurlec for $14

For my purposes I think this quad opamp function generator will be fine.

[Paul Perry (Frostwave)]

Isn't that 'integrating' stage just acting as a low pass filter? taking out the upper harmonics.
In any case, it's a fixed frequency device, if you change the frequency of the triangle, then the amplitude of the 'sine' output will change.
I don't have the book, but it looks very much as though it is from a forrest mims handbook.. if so, at least we know it is 100% sure to work.

[gez]

Isn't that 'integrating' stage just acting as a low pass filter? taking out the upper harmonics.
In any case, it's a fixed frequency device, if you change the frequency of the triangle, then the amplitude of the 'sine' output will change.
I don't have the book, but it looks very much as though it is from a forrest mims handbook.. if so, at least we know it is 100% sure to work.

Yeah, it's easier to think of the second integrator as a filter.  Although not a mathematically pure sine wave, the output will be sinusoidal and it's possible to get a wave form that's almost identical (I can barely perceive a difference with the sine from my function generator).

Not only will amplitude vary with frequency (if made adjustable), but at lower frequencies the waveform will 'triangulate'.  You only have a limited range therefore.

It's possible to get round this, and the reducing amplitude with increasing frequency problem, by using a dual gang pot.  One half to control rate (by dividing down the square wave voltage), the other to change the resistance of the integrator's input resistor. 

I use a variation of the above circuit for most of my LFOs when I want a sine wave (plus options).  Hint:  a LM13700 used for the integrators gives an output with stable amplitude and is reasonably easy to implement (sorry, no schematic).

A similar circuit to the one posted can be found in A Flind's book on oscillators (Babani publishing).