Sine wave oscillator

dl324

Joined Mar 30, 2015
10,976
In the meantime, you can change the title to something that members won't be interested in reading. Or prefix it with "[solved]".
 

Thread Starter

Zeeus

Joined Apr 17, 2019
597
This is the loop gain (i think)

Please how to make the phase -180?
Sadly, TS has forgotten his mathematics

Thanks

IMG_0806.JPG
 

crutschow

Joined Mar 14, 2008
25,122
someone kindly make this work on spice
Here it is in LTspice:

Current source I1 injects a small current pulse into the loop (simulating normal circuit noise) to start the oscillations.

The frequency is 2Hz instead of the stated 1Hz.

Not sure why the sine and cosine signals have different amplitudes.
The sine output also has high distortion.

The values for R3, R5, and R7 (my designators) are unusually high for an op amp circuit.
Would suggest values at least an order of magnitude lower values for a real circuit (with corresponding adjustment of the associated capacitor values).


1594178315730.png
 
Last edited:

Thread Starter

Zeeus

Joined Apr 17, 2019
597
Here it is in LTspice:

Current source I1 injects a small current pulse into the loop (simulating normal circuit noise) to start the oscillations.

The frequency is 2Hz instead of the stated 1Hz.

Not sure why the sine and cosine signals have different amplitudes.
The sine output also has high distortion.

The values for R3, R5, and R7 (my designators) are unusually high for an op amp circuit.
Would suggest values at least an order of magnitude lower values for a real circuit (with corresponding adjustment of the associated capacitor values).


View attachment 211596
God bless
Please the .asc
 

LvW

Joined Jun 13, 2013
1,056
Zeeus - why are you asking for 180 deg?
Only in case of an inverting amplifier we need a feedback network that can produce a phase shift of -180 deg.
But your circuit is quite different!
Of course, this is NOT a quadrature oscillator. Such an oscillator would consist of two integrating circuits.
Therefore, no surprise that both amplitudes are (a) not equal and (b) not in quadrature (90 deg).
This is a specific form of a phase shift oscillator - it consists of an inverting integrator (+90 deg) and a 2nd-order active lowpass in Sallen-Key-unity gain configuration (-90 deg at the pole frequency).
Hence, the loop gain can fulfill the oscillation condition (Barkhausen) at the pole frequency of the lopwpass.
 
Last edited:

LvW

Joined Jun 13, 2013
1,056
Also thought the output of integrator is +90 but book keeps it at -90. The 2nd order can give up to -180 given total of -270
* The phase of an (ideal) inverting integrator is always +90deg. Any other statement is wrong. (Don`t blindly follow any book statements when they are obviously false).
A real inverting integrator (with a real opamp) has a phase shift of +90 deg at one single frequency only.
* The 2nd-order (non-inverting) lowpass has a phase shift of -90deg at the pole frequency (in your example app. 0.5 Hz) and approaches a phase shift of -180deg for infinite frequencies only!
 

ci139

Joined Jul 11, 2016
1,561
if the simulation is not the high frequency or square wave component one the LM301 & LM308 act best with external compensation of around 60pF
the "minimum" (fastest-stable) compensation is as below (i don't have required meters to evalutae the real life case for)
/// as an addition to http://www.mit.edu/~6.301/LM108.pdf , http://www.mit.edu/~6.301/LM101A.pdf , https://www.analog.com/media/en/technical-documentation/data-sheets/lm101a.pdf , http://radio-spravochnik-byru.fddvoron.name/pdf/lm/lm101.pdf ///
(below might-be not the optimum - but is ok for simulations)
OpAmp comp-test - a1.png
 

LvW

Joined Jun 13, 2013
1,056
@ci139 - did you notice that the given circuit oscillates at ap.. 0.5 Hz (as expected in post 14 and shown in post 17) ?
So - why are you speaking about high-frequency properties?
 

ci139

Joined Jul 11, 2016
1,561
about high-frequency properties?
1-st of all i was noting about my experience in compensating the LMx01/LMx08 for not the upper bandwidth edge operation

the 2-nd (for the sensitive or high speed op amps) such as and the LMx01/LMx08 appear to be (without the external compensation) . . . i was suggesting the minimum compensation scheme (otherwise they tend to oscillate around the expected output (median)) ...
... that appears to be both as the minimum-for-stable ((((and thus also the fastest for that particular compensation scheme))))
 
Top