Wien Bridge Oscillator Waveform Shape

[eev_carl]

Hi,

I simulated a Wien Bridge Oscillator and implemented it on my breadboard.  I'm getting oscillation, but it looks like there's clipping at the higher and lower ends that make this look more like a square wave.  I'm using a 741.  I tried changing the source voltage but it just produced a similar shape at a higher voltage.

Is this expected for this old op amp or should the waveform be more rounded as in the sim?

Thanks,
Carl

[Andy Watson]

To obtain a sinewave with the Wein Bridge circuit the op-amp needs to have a gain of exactly three. Any less and it does not oscillate. Any more and the oscillations will build up until the op-amp clips - as you've witnessed. The usual method of stablising the gain is to incorporate a voltage dependant element into the feedback of the op-amp - like a thermistor/small incandesent lamp.

[Tim F]

The 741 isn't a problem. Instead, the sim doesn't match reality or you lucked out with the oscillation frequency and opamp bandwidth/slewrate specs of the simmed opamp. It would be bad practice to rely on the opamp being slow enough if that is indeed the case in the sim.

As above, if the gain is too low you get no oscillation and if it's too high it keeps amplifying until it finds the rails or other limitations of the opamp. You need a variable gain element to stabilise it. I wrote an article about bulbs for wien bridges once, you can find it here: http://feleppa.com.au/wienosc.html

It's also possible to add some diodes to 'soft clip' the output, similar to the waveform in your sim. Google image search 'wien bridge diodes' for ideas.

[eev_carl]

gain of exactly three

That's producing a better shape.  I replaced R4 (the negative feedback amp part) with 2k which gives me A=1 + 2/1=3.  Any insights as to the clipping at the bottom of the attached waveform? 

Also, When I used 2k for LT Spice, I didn't get any oscillation.  However, when I used 2.1k, the sim did produce a sine wave.  I went back and breadboarded 2.1k but that didn't oscillate.  (These resistors are 5%.)

[eev_carl]

I wrote an article about bulbs for wien bridges once, you can find it here: http://feleppa.com.au/wienosc.html

Thanks will check out it out and post an update.

[Benta]

Like others wrote, you need an AGC in the circuit. Believing you can hit 3 exactly with trim pots etc. is utopia. Years ago it was pretty standard to use a fast self heating glass-bead NTC in the feedback loop. IIRC, it had part number RA53. Seems to be obsolete, though.

[HoracioDos]

https://www.eevblog.com/forum/beginners/best-way-to-amplify-a-sine-wave-without-distorsion/msg1150871/#msg1150871

[jose347]

some control theorem I can't remember but basically your system is either stable, oscillating or unstable which is your case, your poles are located at the rights side of the graph making the unstable response growth to infinite(in this case your supply voltage). when you use the trimpot to match the 3 gain you are actually moving the poles to the axis of the graph
*but I may be mistaken, so just ignore me

[drussell]

You an also use a FET to stabilize the amplitude instead of a light bulb. 

I'm trying to remember what I used the last time, probably a 2N5457 or 2N3819 or something like that, whatever I had on hand, with good results.  I would have to dig up a schematic or see if I have one around on a proto board or something to double check exactly what I did... 

[drussell]

You could try something like this: