Op Amp sine wave oscillator at 50KHz is not easy to do?

[FvM]

the opamp integrator gives a phase delay of 270 degrees........................then the low pass filter opamp gives another phase delay of 270 degrees............thats equivalent to a phase delay of 180 degrees..........and so when ever a loop gain of unity occurs.we have oscillation at the fitler corner frequency...............

In fact, the +90° phase shift of the inverting integrator adds to the -90° of the second order low-pass, resulting in overall 0°, as required to fulfill the oscillation condition.

....the oscillation frequency is not affected by changing R1 and C3, as long as the loop gain is > unity

In the given dimensioning, the loop gain exceeds unity only by a few percent. Making it larger than 1.1 or 1.2 results in considerable distortions and frequency increase.

Any better and simpler and more robust quadrature oscillators gratefully accepted please

I think it's a good solution for a fixed frequency oscillator. For variable frequency, I'd prefer a classical double integrator.

By the way, does your book explain the considerations for the low-pass dimensioning?

 

[treez]

By the way, does your book explain the considerations for the low-pass dimensioning?

...well, it gives its corner frequency on page 82,83, but not the actual transfer function.........so to speak.

................also, i thought requirements for oscillation was 180 degrees phase shift and gain of 1?..................not phase shift of zero and gain of 1?

I thought opamp filters added -180 degrees to the -90 degrees, giving -270 degrees delay for the filter? (though i see where youre coming from, the book itself quoted the same phases as you yourself did)

 

[LvW]

Hi treez,

I think you have picked a circuit that also works well.
It simply consists of a classical unity-gain Sallen-Key lowpass (quality factor 0.5) and an inverting integrator.
The lowpass provides -90 deg at the pole frequency and the integrator +90 deg (equivalent to -270 deg) in a relatively broad frequency range.
The whole circuit belongs to the class of lowpass phase-shift oscillator. It requires more passive parts and one opamp less than the three-opamp circuit I have recommended.

 

[treez]

surely any op amp based filter always has at least a 180 degree phase lag due to the op amp, which is connected with negative feedback?..........so how can the opamp filter be just -90 deg?.......................WOOPS, my mistake, the opamp is connected as a follower..see what you mean now.............anyway, i think i'm going to have to get the complex numbers out, and go through the circuit of post #16 so as to proove that the gain is unity at the resonant frequency...........and pull out an magnitude from the overall closed loop transfer function.

.....also see what you mean about putting in an initial disturbance........the cct of #16 of LvW now works with an initial disturbance, and i like this corcuit because its clear to analyse with complex numbers.................

Anyway, this is the simplest and best quadrature oscillator that produces a sine wave so far............as per LvW post advice............

https://i48.tinypic.com/jhbvhd.jpg