Below is the LTspice sim of a 4-stage, phase-shift Bubba oscillator which uses a quad opamp package,

The diodes soft-clip the signal to limit the amplitude of U1, and the three following RC stages, along with providing the needed phase-shifts, act as low-pass filters to give a low distortion output:

Plot of node 1 to show the clipping?

ak

 

Plot of node 1 to show the clipping?

ak

Where is node 1? I simulated with a dual power supply and see no clipping. I got 0.02% harmonic distortion

 

How can rain affect internet speed?
I am internetless

 

Plot of node 1 to show the clipping?

Here you go:

Plotted V(1) (U1 output) referenced to Vb to show the symmetrical, very soft, clipping (to minimize distortion).
The small glitches I think are op amp crossover distortion, which is common in the LM324 op amp, (thus I added R10 to minimize that distortion in the output stage).

 

How can rain affect internet speed?
I am internetless

Who is your internet supplier?

 

Who is your internet supplier?

ROGERS.
BTW how can distortion in early stages be relevant? Here is the plot of my nodes but the green is the output wich is the only that matters

 

ROGERS.

Not familiar with them.
How is the internet delivered to your house.

BTW how can distortion in early stages be relevant? Here is the plot of my nodes but the green is the output wich is the only that matters

Because the distortion is not perfectly removed by the low-pass filtering effect of the RC phase networks.

Incidentally you appear to be using a different circuit then mine, as I see no clipping.

Show your circuit along with the plots.

 

ROGERS is Canadian provider. Comes with cable (optical) but everybody has the same issue. Weather affects the Internet speed

For the plot, I used the RC chain that I posted #11. But thanks to your sim, I discovered yet another hidden gem in LTspice: Error log!.

 

Comes with cable (optical) but everybody has the same issue. Weather affects the Internet speed

The weather wouldn't affect the cable speed, so it must effect however Rodgers is getting their internet signal from the rest of the world (perhaps terrestrial microwave).

 

The weather wouldn't affect the cable speed, so it must effect however Rodgers is getting their internet signal from the rest of the world.

There are two big names in Canada: ROGERS and Bell. together they make a good pair for bullock cart

 

here
that one is ROGERS and the other one is Bell

 

Is the idea similar to post #17?

Similar, yes.
But some key differences.
The Integration of the squarewave to a trianglewave significantly reduces the harmonic content.
As I mentioned previously, there are many sinewave generator sub-variants.

 

Where is node 1?

The net named "1" is the output of U1. There is a rotated numeral "1" near the output pin. He has a similar net name for each of the other three opamp outputs.

ak

 

Hello,

Have a look at page 4 of the attached PDF for a comparison between the different types of sinewave generators.

Bertus

 

Hello,

Have a look at page 4 of the attached PDF for a comparison between the different types of sinewave generators.

Bertus

Thanks! So JFET is the approved sub for light bulb according to TI.

 

Incidentally you appear to be using a different circuit then mine, as I see no clipping.

I finally put together the sim but cannot get LTC show me the "harmonic distortion" How do I do it?

 

I finally put together the sim but cannot get LTC show me the "harmonic distortion" How do I do it?

NM. Different versions of LTS have it in different places. In old version needs R-Click on the plot and go to "view" from there

 

cannot get LTC show me the "harmonic distortion" How do I do it?

See sim below and note the addition of the .four command and changes to the transient command:
Also R7 and R8 are not needed in this circuit.

I pulled the exact Fourier frequency to use from the FFT plot using the FFT command in the plot View pulldown.

 

Thanks! So JFET is the approved sub for light bulb according to TI.

I'm not sure about "approved", but it is one that has been around for a long time and is well understood. It does have a problem that a light bulb does not, and that is its frequency response.

After the output sinewave is rectified and filtered to produce the gain element control voltage, there is some ripple at the output frequency on top of the dc control signal. That small ripple modulates the FET's channel resistance at the output frequency, so the resistance is changing constantly during each cycle, which means that the gain of the opamp circuit is changing. This shows up in the output signal as harmonic distortion. A light bulb filament's thermal mass has a way longer time constant than the sinewave cycle period. The circuit wanders up to its nominal operating amplitude much more slowly than the FET circuit, but with less distortion (all other circuit parameters being equal).

ak

 

A light bulb filament's thermal mass has a way longer time constant than the sinewave cycle period. The circuit wanders up to its nominal operating amplitude much more slowly than the FET circuit, but with less distortion (all other circuit parameters being equal).
ak

Hmm. Makes perfect sense. I do not think there is any way to slow down the response time of JFET. The issue with this Wien Bridge is it is too fussy. the two resistors that have 1:2 ratio must be exactly 1:2 or there will be the ugliest output that resembles nothing. Or the oscillation fades to zero. I have already started working on @crutschow circuit. That insignificant distortion is not going to be an issue in my project. I also thought about a light-sealed LED+LDR but that one will probably fail somehow.
In addition, i am looking for something straightforward so, a beginner or hobbyist can tackle.