It's a fine way to make a sine wave in certain circumstances. But the LC circuit can only be tuned to one frequency, so it won't work when you want a variable frequency sine wave. Also, if you want that sine wave to drive a load then a LC circuit will have trouble because its behavior can be impacted greatly by a load.
It's a fine way to make a sine wave in certain circumstances. But the LC circuit can only be tuned to one frequency, so it won't work when you want a variable frequency sine wave. Also, if you want that sine wave to drive a load then a LC circuit will have trouble because its behavior can be impacted greatly by a load.
The whole project had serious flaws IMO, but the method of generating variable frequency sine is interesting. Since you need low frequency sine, you can do this that way.
What they did was to use the MAX7480 ic, which is variable frequency, switched capacitance type, 8th order lowpass filter.
I'd say use a large electrolytic capacitor, take out the intestines and hide attiny13+max7480 inside.....
But seriously: does potentiometer count as resistor? That's would be a way to do that. In general if you want variable frequency RC filter then you need to vary either resistance or capacitance. Variable capacitors are out there, but of low value and big size, also usually much more expensive than potentiometers, so it's more convenient to vary resistance.
The general principle of convertine square to sine is that square consists of some fundamental frequency and it's odd harmonics. By using a filter with corner frequency equal to the fundamental frequency of square wave you filter out harmonics. This leaves only the fundamental frequency. And it so happend that frequency without harmonics means a sine wave.
By using a filter with corner frequency equal to the fundamental frequency of square wave you filter out harmonics. This leaves only the fundamental frequency. And it so happend that frequency without harmonics means a sine wave.
Yes, that´s one can learn in school.
However, the reality is not so simple - you cannot "filter out harmonics". You only can suppress harmonics down to a certain and acceptable level. And this determines the filter degree that is necessary to do the job. I think this forum is advanced enough not to simplify things to much.
And - the corner frequency should NOT equal the fundamental of the squarewave. It should be larger depending on the filter degree.
If two stages of integrators are connected in series, the first one converts the square wave
to a triangle wave, and the second stage converts the triangle wave to near a sine wave.
If the input frequency varies widely then the output amplitude will vary.
If amplitude needed to vary less then an AGC stage must be added on the output.
Duty cycle of the square wave needs to be near 50%, or must modified so.
Simply integratng triangle wave doesn't produce sine wave, it produces quadratic function wave (parabolic shape). I mean simple math tells you that integrating a linear function doesn't get you a sine or cosine, but rather x^2. This may look similar to sine, but it's not.
There is a method of generating quite good quality sine from triangle wave, but it involves feeding the tirangle into a network consisting of diodes and resistors which creates sine exploiting the nonlinear regions of diode characteristics. This method is used in IC's like XR2206 and ICL8038.
There is a method of generating quite good quality sine from triangle wave, but it involves feeding the tirangle into a network consisting of diodes and resistors which creates sine exploiting the nonlinear regions of diode characteristics. This method is used in IC's like XR2206 and ICL8038.
Another simple solution for triangle-sinusiodal conversion is the following:
Feed the triangle wave into a classical diff. amplifier (long tailed pair) with an amplitude that uses some part of the non-linearity of the transfer characteristics (tanh-curve causes "soft clipping" of the sinus). This gives rather good and frequency-independent sinus signals:
THD app. 1% for a triangle amplitude around 80 mV.
Switched capacitor is not something you usually do with discrete components. This is a way used in integrated circuits to obtain precise resistance. Manufacturing precision resistor in silicon is hard, while in case of capacitor it is not. By letting chatrge in and out of capacitor in controlled manner you gain controlled current flow which simulates resistance and can be very precisely controlled. In discrete circuits you usually don;t do that.