| Electronics > Beginners |
| Wien bridge oscillator Design |
| (1/3) > >> |
| NASK:
i need to design Wien bridge oscillator with 50Hz sine wave. so i went through https://www.electronics-tutorials.ws/oscillator/wien_bridge.html tutorial and design in Proteus but didn't work at all. can i know the reason |
| David Hess:
You need to fix the gain control; it has to be regulated to produce a controlled output. This is easily accomplished with a small incandescent lamp. Starting on page 29 of Linear Technology application note 43 is a good discussion of Wien bridge circuits. |
| JS:
The amplifier must have a gain greater than 3 to get started but exactly 3 to run and output a sinewave. For this some gain control is needed, many options around that, a very small incandescent lamp is the easiest one. It should be placed where R3 is and turned on as soon as the oscillator starts to do it's thing. It shouldn't be too small that the lamp gets burned. The off resistance of the lamp should be as such you have a gain just over 3, so when it turns slightly on you get a gain of 3 and it keeps it happy and cool. You could add a resistor in series with it or even trim it to get the desired value. Low freq sinewaves with this configuration have higher THD than higher freq, as the thermal constant of the lamp is shorter or comparable with the period of the wave. JS |
| Benta:
Back in the 70s/80s, it was standard practice to use a fast glass thermistor (NTC) for gain control. The most common was a 5 kohm type called RA53 (this was really the type designation, not the silk-screen name :) You can still find RA53 NOS if you search the web. |
| LvW:
No - I do not think that the missing gain control would be the reason for your observation. Gain control is necessary - if the oscillator starts to work - for softly limiting the rising output amplitude before it will be hard-limited at the power rail. However, in your case (if I understood your problem right) the oscillator does not start at all at t=0. To solve the problem, you need something like a "kick-off" (remember a pendulum clock, which also cannot start by its own). If your amplifier model is a real model (including some non-ideal properties), there will be a small DC output which is equivalent to an unbalance - and the oscillator most probably will be able to start. However, this is not the case for an ideal opamp model. For an ideal model, you have several options: * Give one of the capacitors an initial condition (a voltage at t=0) * Switch one of the power supplies at t=0 (otherwise the simulator assumes that power is on since minus infinity) * Introduce a short voltage peak (impulse) into a grounded node. Remark 1: The most simple amplitude control is to use two (antiparallel) diodes across R4. Remark 2: Simulation duration should be at least 10...20 oscillation periods with a resolution of at least 50 points per period. |
| Navigation |
| Message Index |
| Next page |