Wien Bridge Oscillator

Audioguru again

Joined Oct 21, 2019
6,674
Hi Wendy, your latest circuit probably does not need the diode parallel with the Mosfet.
R7 should go to ground, not to the positive supply.
The rectifier diode probably needs a series resistor to prevent the opamp from trying to drive an extremely low resistance.
The distortion of a Jfet is much less when you feed half the amplitude at its drain to its gate, but that can be tried later.
 

Thread Starter

Wendy

Joined Mar 24, 2008
23,415
It is built into the MOSFET, I include it for accuracy. It is also a dandy marker for what is positive and what is negative on a MOSFET.
 

Thread Starter

Wendy

Joined Mar 24, 2008
23,415
Actually I select the wrong MOSFET. It is an n channel, an IRF540 I want the default biased to be in the on position I will use the signal from the oscillator to turn the FET off you will find it will find its own balance that way.
 
Last edited:

Thread Starter

Wendy

Joined Mar 24, 2008
23,415
Latest brain fart:
Wien Bridge oscillator 3.png
At first I need the high gain. Then the mosfet will adjust whatever resistance it needs to limit the signal.
 

Audioguru again

Joined Oct 21, 2019
6,674
Even though R6 eliminates the crossover distortion produced by the LM358, the distortion is much higher than other Fet-stabilized Wein Bridge oscillators. Maybe because the others use a Jfet not a Mosfet?
I expanded the signal at the Mosfet drain to see it:
 

Attachments

Audioguru again

Joined Oct 21, 2019
6,674
Yes, the Jfet works much better than the Mosfet for Automatic Gain Control. The amplitude bouncing is obvious at startup.

They say that if you feed half the drain amplitude to the gate of the Jfet then the even numbered harmonics are cancelled and the distortion is much lower.

A better opamp should be used for good sinewaves. I used a TL071 single and a TL072 audio dual opamp many years ago.
 

Thread Starter

Wendy

Joined Mar 24, 2008
23,415
I have not been feeling very good last few days so I took a vacation from this project. My problem seems to be in getting the voltage doubler to work. So I built a test circuit off to the side and here are the results after I got it to work( the doubler that is):

Wien Bridge oscillator 3 junk.png

this is the waveform with no gain control:

SDS00001.png
actually this is what I wanted to give the doubler full drive.

I was able to go down to 2.68 volts. Which method did turn the waveform into 9.32 volts. Hopefully that is enough drive to turn the MOSFET most of the way off. I use true ground( the negative power rail) for the MOSFET to keep it in the linear region and minimize distortion.
 

AnalogKid

Joined Aug 1, 2013
10,989
C7 should be reversed. CR1 clamps the C7 cathode near GND, while the opamp output can drive the C7 anode below GND for a significant, and almost constant, reverse bias.

ak
 

Thread Starter

Wendy

Joined Mar 24, 2008
23,415
So having gotten the voltage doubler to work, it still did not have enough drive to shut the transistor in into a low conduction state. This is the circuit I tried:
Wien Bridge oscillator 4.png
And this was the output waveform:
SDS00001.png
I'll move the ground reference to the virtual ground on the doubler. To give it a boost in driving the MOSFET off.
 
Last edited:

AnalogKid

Joined Aug 1, 2013
10,989
Note that in both post 53, the opamp output does not go below GND. I think this means CR1 never conducts, so there is no voltage doubling. I suggest that you return C5, CR1, and Qx to the virtual ground rather than GND. With this, the fact that C5 starts out with zero charge will work for you rather than against you.

ak
 

Thread Starter

Wendy

Joined Mar 24, 2008
23,415
So what are the odds this might work?

Wien Bridge oscillator  5.png

My thought is to drive the gate to a lower DC value. Where the voltage doubler might be able to do something.
 
Last edited:

Audioguru again

Joined Oct 21, 2019
6,674
1) R7 turns on the N-channel Mosfet resulting in massive opamp gain.
2) The opamp is overloaded trying to charge C6 with its backwards polarity into CR1.
There is no voltage doubling, the Mosfet might never have its conduction reduced enough to avoid the severe clipping from the opamp.
 

Thread Starter

Wendy

Joined Mar 24, 2008
23,415
A friend of mine pointed out that a grain of wheat bulb is an alternate name for a peanut bulb. After doing a quick search came up with this. These would have worked just fine . But being a stubborn cuss I will continue to work on the design concept I was working on until I make it work. I will post both projects when I have them done.
 

Audioguru again

Joined Oct 21, 2019
6,674
A 1.5V grain of wheat bulb at 100mA calculates to have a hot resistance of 15 ohms which is way too low for an opamp circuit even if it is cool at 150 ohms. The datasheet for an LM386 power amplifier shows a 3V 15mA (200 ohms hot) bulb in a Wien Bridge oscillator.
Google shows some 12V bulbs at 70mA with a hot resistance of 171 ohms that might work with an opamp.
 
Top