Which Oscillator type is this?

[John Lyons]

I can't find other examples of this.
What is this type called?
Do I need to change all 3 caps to
change the range?

[PRR]

Do I need to change all 3 caps ...

I think so.

[anotherjim]

I think it's a phase shift oscillator. There are 3 RC phase shift elements with the two opamp integrators and the passive RC of R12 & C11. The LEDs are limiters. I don't know why the job would be done this way.

[iainpunk]

The leds are opto's. Left one is a square wave, since its coming off of a Schmidt trigger. The 2nd opamp is an integrator. Think of the integrator as a slew rate limiter. Its voltage rises steadily, until a threshold is reached, the Schmidt trigger flips and the integrators output starts to go down at the same steady pace, until another threshold is reached, and it repeats again. Creating both square and perfect triangle waves

Cheers

[John Lyons]

iain
The LEDs are a bi-color common cathode LED.

[PRR]

Is it open on your bench? Try C10 twice the size. Since IC1B is an 'infinite' gain integrator it should still oscillate. All three caps have effect so changing just one by 2:1 makes about 1.26:1 change of frequency. Still it *might* stand a 10:1 change on one cap for 2.15: change of freq.

If it works, let us know.

Since it is a 3-phase "quadrature" (triature?) oscillator, taking a bi-color LED at two equal points should give not one not two but three colors in a cycle. (But C10 way oversize will distort the cycle?)

[iainpunk]

iain
The LEDs are a bi-color common cathode LED.

Owh, id helped someone with a very similar LFO with 2 opto's like that, for modulation purposes. The rest of my Comment was inaccurate as well, as i assumed it was the same osc. I helped someone else with.

The left opamp is not a Schmidt trigger, its more like a 'slow comparator'.
The right most opamp is still an integrator. Id put tis in the category of phase shift oscillator.

Cheers

[anotherjim]

I think the left-hand opamp is still an integrator but non-inverting. You need one inversion in a PSO and that's the right-hand amp.
And yes, the LED's are just a fancy colourful rate indicator.

[PRR]

Both stages integrate (not switch). It's all negative feedback and linear. And yes it is a Phase Shift.

In the ideal world of SPICE, it won't start without a kick, and when it builds-up to clipping the math blows-up. I'm sure with a real LM358 it will self-start, build, and limit more or less smoothly.

If you change C10 larger, as I suggested, the output level falls. Duh, of course. The level at IC1A out is more constant, but offset caps upset the lovely three-fold symmetry.

I do not know why there is 100k and 150k. Normally all three are the same, but of course here one must pot-vary.

It won't be the cleanest sine in the book, between hard-clipping and the nasty 324/358 crossover. Sometimes this is OK in an LFO.

[John Lyons]

I forgot to thank you for this Paul.
THanks for running that sim.
I'm  still fooling with the circuit.