Doesn't work basically. The suggested circuit (as any other possible 555 circuit) supplies a squarewave. Depending on the specification of allowed harmonics respectively sine distortion, you can use a filter (low-pass or band-pass) to convert it into a sine. Unfortunately, the filter has to be tuned when you intend to vary the sine frequency.
It should be possible to do this without using a tracking band pass filter, but rather a wideband shaping circuit. For this the 555 should be modified to produce 50% duty cycle output (can be done by adding 1 extra resistor in discharge pin) to keep the shaped sine wave distortion low. Next the charge/discharge ramps on the timing cap should be made linear (also to minimize dist) by implementing current sources. This triangle wave can then be applied to a sine wave shaper through a buffer. This shaper can operate from Hz to few 100KHz. The result will be a sine wave output with about 1-2% THD over the applied frequency range.
You will have to control the current source/s to alter the frequency.
If you are allowed extra opamps you could integrate the square output of the 555
If you design a function generator with a sine shaper, I don't see a particular purpose of a 555. The suggested integrator also needs to be gain tuned over the frequency, and it doesn't provide a defined zero bias without additional circuitry. It would be easier to build an OP triangle generator directly.
Generally, using a XR2206 is a convenient way to design a sine function generator, not really an answer to the original question however.
From his post I took it that he must use a 555, which appeared to be a home work problem or project. There are obvious much easier ways to accomplish the task if the 555 is not a requirement.
I doubt, if this can be seriously a homework or a project. Basically, because the effort for the supporting circuit would be a multiple of the 555 generator itself. And I don't think, that any feasible solution actually fits the specification "only 555".
It can be a meaningful project to generate a sinewave synchronized to a reference input, typically involving a PLL or another selftuned circuit. But the specification would be obviously very different from what's said in this thread.
Here I attach a circuit solution of my original idea (without the current sources)
When the frequency gets higher the non-linear slopes are less of a problem, and the resulting sinewave distortion is not that obvious.
Short description:
U1A buffer for triangle over C timing
U1B amplification + offset adj (sine shaper needs about 4vpp)
R8 adjusts for minimal distortion.
I think that the 50% duty cycle with varying the frequency can be maintained by having mechanical ganged pots for R1 and R2. Duty cycle for this is about 51%. This can be trimmed by an extra resistor in the discharge path.
May I propose to the community another simple and good working circuitry for a sine shaper ?
If you have a triangle wave with a fixed amplitude of 80 mV you can use it as an input for a classical long tailed pair (differential amp) - and as a result you get a sinusoidal form of approx. 1.2% THD.
Of course this shaper is not frequency selective, it only uses the soft clipping feature of the tanh function.
May I propose to the community another simple and good working circuitry for a sine shaper ?
If you have a triangle wave with a fixed amplitude of 80 mV you can use it as an input for a classical long tailed pair (differential amp) - and as a result you get a sinusoidal form of approx. 1.2% THD.
Of course this shaper is not frequency selective, it only uses the soft clipping feature of the tanh function.
Yes, can basically work this way. Instead of a the required ganged R1/R2 pot, you can use a much more simple 555 circuit with a single RC combination connected to Output instead of Disch.
Here I have used a floating current source idea from a old (1973) Siliconix application note to linearize the ramps. Now the frequency is adjustable with only one pot (200k) over a wide range. Resulting sine wave distortion is less than 2% as shown by dist meter.
I am sure fine tuning can give lower distortion. For higher frequencies a faster opamp should be used.