To the fledgling electronics enthusiast who has just enough knowledge of electronic music theory to be dangerous, it makes sense that the simplest tone, the sine tone, should be the simplest/easiest to construct in the analog electronics domain. Of course, this is not the case. While geometrically simple to describe, the sine tone is not easily achieved using electronic components. Most designs found online indicate that the most common method is to “filter” a square wave created by an op-amp through a given set of components placed between the output, the non-inverting input, and the ground. This is the method employed by the Wien Bridge Oscillator.
From the Wikipedia article above:
“Hewlett used an incandescent bulb as a power detector, low pass filter and gain control element in the oscillator feedback path to control the output amplitude. The resistance of the light bulb filament (see resistivity article) increases as its temperature increases. The temperature of the filament depends on the power dissipated in the filament and some other factors. If the oscillator’s period (an inverse of its frequency) is significantly shorter than the thermal time constant of the filament, then the temperature of the filament will be substantially constant over a cycle. The filament resistance will then determine the amplitude of the output signal. If the amplitude increases, the filament heats up and its resistance increases. The circuit is designed so that a larger filament resistance reduces loop gain, which in turn will reduce the output amplitude. The result is a negative feedback system that stabilizes the output amplitude to a constant value. With this form of amplitude control, the oscillator operates as a near ideal linear system and provides a very low distortion output signal.”
The lamp used in modern versions of this circuit is not a common part and must be ordered. We tried a thermistor instead with limited success. Ultimately, we could not adequately determine the value of R3 for use with the thermistor to achieve anything like a sine tone so we could not test any of the distortion claims for the circuit.
We did eventually get something resembling a sine tone by using a diode-stabilized version of the Wien Bridge Oscillator which we found on this page. (for your edification and quick-downloading pleasure, here is a PDF copy of that website module on oscillators: Oscillators-module-03)
Below is the circuit for the diode-stabilized version of the Wien bridge.
We substituted a few components, namely a 741 op amp for the LM324 they use in the tut. We calculated our RC combos to give us 1000 Hz, but you can choose what you want by adjusting them according to the equation f=1/(2piRC).
Here’s some pictorial proof that this circuit works.
Here’s a different angle showing the circuit (and voltage divider…)
Here’s a hilarious video of us firing up the first version of the circuit using lamps instead of resistors in the voltage divider. They lasted about 10 minutes before burning out.
After the lamps in the video burned out we tried a couple 1 watt resistors but calculated the voltage wrong and got the magic smoke to appear out of them.
Here’s an annotated version of the functional diode circuit:
- A small resistor replaces the thermistor / lamp.
- Two N4001 diodes are placed in parallel with R3 between pins 2 and 6 (via a jumper.)
- These two resistors replaced the lamps originally used in the hilarious video above that burned out after 10 minutes or so.
In the above picture, the two alligator clips attached to R3 run to a potentiometer we used to “tune in” the resistance. Because of the nature of the non-inverting input, slightly too much resistance and the circuit just becomes a square wave generator, slightly too little and it will not oscillate at all. Using a pot with the resistor that was close to correct allowed us to dial in the required resistance. The difference between no oscillation and square wave is about 7 ohms in our circuit — really tight. My electronics mentor (and yours) Ethan has some ideas about stabilizing the voltage using a transistor. Perhaps we will have an updated post soon with exciting details. Additionally, I’ll add a video of the working circuit when I have a chance. Until then, keep letting the magic smoke out!