Please is there anybody to direct me for a reasonable method to take the double integral of the audio spectrum or at least the speech?
Is it possible to do so by 2 op-amp integrators connected together in cascade mode?
I myself was thinking of something like the below pic. But I am not sure if such an op-amp integrator is able to be used to integrate the speech or audio spectrum, Is it?
I myself was thinking of something like the below pic. But I am not sure if such an op-amp integrator is able to be used to integrate the speech or audio spectrum, Is it?
That should work. Set the corner frequency of each integrator somewhere below the low end of the audio band. Your audio response will roll of at a rate of 40dB/decade. Is that what you want? Why do you want the double integral of audio?
Please correct me if I am wrong:
The 3dB frequency for such a circuit and of course for each stage is:
F3dB= 1/(2pi x Rf Cf)
I noticed to have a good integrator the Rf should be very large, I noticed the same for the Cf too, Now please tell me how to choose the values for the Rf and Cf so that could cover the audio range yet having a reasonable integrator? I can simulate it if you have any approximate values in your mind?
Your audio response will roll of at a rate of 40dB/decade. Is that what you want? Why do you want the double integral of audio
I just want to take the doubl eintegration of the audio, So it should be possible to select the commonents so that the roll of frequency happens to be far away from the audio spectrum so that have the minimum of attenuation, Or there is something which I do not know?
I would think you need the 3dB frequency to be no higher than 20Hz, if you want to cover the entire audio band. If you make Rf=1MegΩ and Cf=10nF, the corner would be ≈16Hz. At 160Hz, the response would be 40dB down, at 1.6kHz, 80dB down, and at 16kHz, the response would be 120dB down. The initial gain (below 16Hz) would be set by Rf/Rs in each stage.
I hope you are not planning on trying to listen to the resulting output.
If you make the roll off corner above the audio range, you won't be integrating audio.
What are you really trying to do? maybe a double integrator is not what you need.
The circuit we are discussing here is just a low pass filter with 40dB/decade slope, and a corner frequency of 1/(2πRfCf). In fact, a true integrator has its low frequency corner at 0Hz.
It won't work as shown since there is no DC path (or reset function) across the feedback capacitors. The output will just go to one rail voltage and saturate. If you add a high resistance between the output and the (-) input it should work.