positive feedback in an op-amp

Thread Starter


Joined Jul 26, 2004
Hello All,

I have a Twin-T oscillator, where the Twin-T (RC network) is connected between the negative input and output and the positive of the amp is grounded. The frequency of this oscillator is f = 1 / 2pi RC. (where R and C are the resistors and capacitors forming T-network in the feedback path).

However, with the same connections between the negative and output, if I short the positive and the output terminal (so now, I have a -ve feedback and a positive feedback), and connect a resistor between the positive and ground, the frequency of oscillation is tremendously reduced and does not follow the f = 1/2pi RC.

Can somebody explain to me the role of the positive feedback here and how it can cause a variation in the frequency. Or is there a way (any formula) by which I can exactly predict the frequency of oscillation with the positive feedback.

Any suggestions are welcome.


Joined Aug 10, 2005

Had to look myself since it has been some decades since I did any design work with them.
One of the best explanations I have seen (that didn't drone on for days) is right here on this site:


Since positive feedback in its most primitive form can make a latch of an op amp, the hysterisis will definitely drop the frequency from the point you might expect based on the RC time constants.
It kinda wants to stay lit - not relax.