Op amp oscillator operating point

Jul 24, 2008 3 Replies
S

I simulated this op amp oscillator with LTSPICE but was a bit surprised that the output was a square wave rather than a sinusoid. It is a dual supply op amp with 3 RC networks between the output and the inverting terminal.



To make it I simulated the RC network cascade and saw where the transfer function phase crossed 180 degrees. I was expecting oscillation to happen there at 3.9 kHz, but it actually occurred at



3.04 kHz. I had assumed that the op amp doesn't add phase, but maybe I should look at the phase transfer function in the circuit with the caps removed.

Perhaps I need to shoot for a higher frequency if I want a sinusoid?



R1 Vout 3 10k C1 3 0 10n R2 2 1 10k C2 1 0 10n R3 3 2 10k C3 2 0 10n XU1 0 1 VCC VEE Vout LT1001 V1 VCC 0 2.5 V2 0 VEE 2.5 .tran 2m startup .options method=trap .lib LTC.lib .backanno .end


Vote
Twitter Facebook LinkedIn
A

That would be like balancing a pencil on its point, which is sometimes possible in spice.

A sine wave is available at the inverting input. To get one at the output, you need a way to stop the amplitude of oscillation growing to the point where it's limited (clipped) by the available output voltage swing.

Checkout

formatting link

Vote
Twitter Facebook LinkedIn
B

Check out Jim William's low distortion Wein bridge sine-wave oscillator - this is fairly close

formatting link

1,P2440,D7461

The gain of the circuit is carefully controlled (by the bias on the FET) so that it controls the amplitude of the sine wave, rather than relying on letting the output clip.

-- Bill Sloman, Nijmegen

Vote
Twitter Facebook LinkedIn
M

For oscillators, you need to force the time step on the .tran card. Use all 4 parameters.

The very fact that the op amp is of finite bandwidth means it will add phase lag.

Vote
Twitter Facebook LinkedIn

Join the Discussion

Have something to add? Share your thoughts — no account required.

Didn't find your answer?

Ask the community — no account required