How to mock a sine wave out of triangular? (LFO)

[jrc4558]

Say I've got a 3v rms triangular wave from lfo. if i use shunt diodes to clip it, like two pairs of 1N914 to Vref and parallel them with a .22 uF capacitor, will the resulting wave look more like a sinewave, rather than a trapeziod? Is such filtering applicable to DC signal at all?
Thanks.

[Paul Perry (Frostwave)]

It won't be perfect, but yes, parallel clipping with diode/resistors will get it more smooth (and, the impedance from the precvious stae has a bearing on the resistor values you need). You will lose signal amplitude though.

You can get a very good sine by overdriving a CA3080 amplifier.
Now that 3080s are scarce, use 1/2 a LM13700.
It is also possible to make the front end of a 3080 out of transistors & overdrive that.

And look at the bottom right of the last pic here:
http://home.att.net/~synthsource/ens76/vco2.htm
how to convert a triangle to a sine using a fet!

[jrc4558]

Nice! Thank you! Could you possibly elaborate on the CA3080 apllication for this purpose? I have some of those chips, so maybe...

[Paul Perry (Frostwave)]

http://www.musicfromouterspace.com/analogsynth/vco.html

There is a 3080 in there, with a triangle input, and a sine output.
But, note it is for a +-12v rail system... so you need a few mods.
Don't let R37 get smaller than 10K!

The LM3080 or CA3080 data sheets & appliction notes will show it as a 'straight' amplifier somewhere. You just reduce the input resistor until you get a sinewave!

[Mark Hammer]

The Hypertriangular Clock circuit from James McConkey's old POLYPHONY article that I have posted at my site shows how to make one side of a 2-op-amp triangle wave sinusoidal using an FET.  IT makes reference to a prior article in POLYPHONY from Jacques Boileau which used a related but more complicated technique where BOTH half waves were made sinusoidal and then rectified to produce a waveform that was triangular at one end and sinusoidal at the other.  

As McConkey cleverly notes, the original used BOTH gate-drain and gate-source  for sine shaping, and you could keep the triangular form of one half wave by simply dropping one diode on the FET.  Of course, stick the other one back in and you're back to sinusoidal on both sides.

You can see the circuit at:  http://ampage.org/hammer/files/hypertriangleclock.gif

[Paul Perry (Frostwave)]

That fig2 is pretty neat, thanks for pointing it out Mark! at first I was thinking, that can't work, where's the first leg of the voltage divider? then... :idea:  :shock:  
Note these triangle to sine distortion type shapers only work for a particular input signal swing.. but you can turn that to your advantage, on my SpaceBeam optical theremin oscillator, I feed a variable amount of triangle to a 3080, and as you increase the drive the output goes from triangle to sine to rounded square, with very litte change in output signal level. Plus a DC offset would give another variable in the mix.

[jrc4558]

Thank you gentlemen! My endeavor here is to use a converter to be applicable with Craig Anderton's phaser design, to make the phaser sound like a uni-vibe... Just as Mark once recommended - feedback elimination, capacitor value shange and LFO waveshape alteration. That should be enough, I assume. But what will be beneficial is the amount of control one can have from the existing design. Additive-substractive mix, variable feedback or no feedback, LDR based - no FET-associated distortion and bipolar 9V power conditions. One would never have such controls on a Uni-Vibe. That's my idea of a good, well controllable phaser. And I really like the sound of four stages. Hehe. For now.

So thanks again and best to you.
cn

[Transmogrifox]

Sedra/Smith text, "Microelectronic Circuits" has a relatively nice circuit for converting a triangle wave into a sine wave.  It's pretty much a glorified tubescreamer with about 3 different clipping thresholds per polarity.  It generates a pretty good approximation for LFO purposes.

[toneman]

then...in an electronic galaxy long ago,
PAIA made a 2720-14 sine converter module.
triangle in & sine out & also PWM out.
triangle-2-sine shaper uses the ol' piecewise-diode function technique..
what's *your* triangle?
staysined
tone

[ExpAnonColin]

And look at the bottom right of the last pic here:
http://home.att.net/~synthsource/ens76/vco2.htm
how to convert a triangle to a sine using a fet!

That IC is a 556... Is it mislabled or are dual timers also op amps?   :?

-Colin

[toneman]

further down, on page4, it mentions:

"The relaxation oscillator is Terry's original with a few modifications. As a general rule, for the entire circuit, you should be able to substitute LF351's or similar for the 556 , 301, and 307 op-amps. You can also use an LF351 for the CA3140."

i've never heard of an opamp named 556.
was thinking it was a VCO chip....(?)

with the + and - inputs, it's got 2 B an OA.
see the ref to Ian Fritz's VCO....same one...drawn diffenently

staytuned
tone

[Paul Perry (Frostwave)]

Must be a mipsrint, there used to be a LM356 (low noise 741 equiv).
The 556 is a dual timer, sure enuff.

[Peter Snow]

... and the VCO chip is a 566.

Peter