Twin Tee Sine Wave Oscillator

How does a twin-tee sine wave oscillator work?

The schematic I have shows an op-amp with a twin-tee between the output and
the - input, and a voltage divider formed by a 10K resistor from the output
to the + input, and a 1K resistor from the + input to ground.

I understand that at the frequency w= 1/ RC the twin-tee produces a -180
degree phase shift. But at w= 1/RC, wouldn't the - input of the op-amp
become 0V because of the twin-tee's notch effect? Then the loop gain AB =0?

zero wrote:
> How does a twin-tee sine wave oscillator work?
>
> The schematic I have shows an op-amp with a twin-tee between the
output and
> the - input, and a voltage divider formed by a 10K resistor from the
output
> to the + input, and a 1K resistor from the + input to ground.
>
> I understand that at the frequency w= 1/ RC the twin-tee produces a
-180
> degree phase shift. But at w= 1/RC, wouldn't the - input of the
op-amp
> become 0V because of the twin-tee's notch effect? Then the loop gain
AB =0?

The notch has finite depth, typically 60dB i.e. B = 0.001

The potential divider on the +ve input provides a little positive
feedback to make sure the oscillator starts reliably.

If B = 0.001 would I have to change the voltage divider on the + input of
the op-amp to produce a gain of A = 1000 to keep AB = -1 at the desired
frequency?


"Andrew Holme" <(E-Mail Removed)> wrote in message
news:(E-Mail Removed) ups.com...
>
> zero wrote:
> > How does a twin-tee sine wave oscillator work?
> >
> > The schematic I have shows an op-amp with a twin-tee between the
> output and
> > the - input, and a voltage divider formed by a 10K resistor from the
> output
> > to the + input, and a 1K resistor from the + input to ground.
> >
> > I understand that at the frequency w= 1/ RC the twin-tee produces a
> -180
> > degree phase shift. But at w= 1/RC, wouldn't the - input of the
> op-amp
> > become 0V because of the twin-tee's notch effect? Then the loop gain
> AB =0?
>
> The notch has finite depth, typically 60dB i.e. B = 0.001
>
> The potential divider on the +ve input provides a little positive
> feedback to make sure the oscillator starts reliably.
>

zero wrote:
> If B = 0.001 would I have to change the voltage divider on the +
input of
> the op-amp to produce a gain of A = 1000 to keep AB = -1 at the
desired
> frequency?

No.

For one thing: B is not 0.001; that's just a typical "ballpark" figure.
It will vary with component tolerances, temperature e.t.c.

Another thing: you can't control gain like that with POSITIVE feedback.

Sustained oscillation requires zero (or 360 degrees) phase shift and
unity gain. Due to temperature coefficients and component tolerances,
you can't get exactly unity gain, so oscillators are designed with
gains greater than unity. The question then is: what is the amplitude
limiting mechanism?

The amplitude may be allowed to grow until limited by the power supply
rails, the designer may use the non-linear characteristic of an active
component to limit the amplitude, or some form of automatic gain
control may be employed - whichever way it is done, the gain is
effectively unity!

The amplitude limiting mechanism affects the purity of the sine wave.

On Thu, 17 Mar 2005 09:17:33 GMT, "zero" <(E-Mail Removed)>
wrote:

>How does a twin-tee sine wave oscillator work?
>


The positive feedback path wants to make it oscillate; assume it wants
to oscillate at all frequencies. The negative feedback through the t-t
overcomes the positive feedback and kills the oscillation. But at the
tt notch frequency, there is no neg feedback, so that's the only
frequency it can oscillate at, so it does.

John