Assessing Modulation Index in AM Demodulation: Techniques and Limitations

[watertreader]

hi,

Just want to have some comments on the AM Demodulation steps that I have taken.

A AM modulated signal is given by AM(t) = DC + m cos(wt)

For the general demodulation technique, I understand there is two general form - (i) Synchronous Detection whereby you mix the input wave with an local oscillator signal or in more advance form by using a 90 degree phase delay signal to get the quadrature as well (ii) Envelope detector - wherby we will just detect the envelope of the input with an diode and cap.

Now if we would just want to have the knowledge of the modulation index instead of recovering the whole wave. I am thinking of the following method

1) compute the FFT of the signal and measure the spectrum height at the modulation frequency (as a representative value of modulation)
Or
2) Compute the difference between the maximum and minimum signal off the signal at time axis

Would this be appropriate? I can only think of the problems would be that both techniques usage would be restricted to low frequency and low distortion cases.

Like to have your comments

Thanks

 

[Naty1]

Using the definition of modulation index here:
http://en.wikipedia.org/wiki/Amplitude_modulation#Modulation_index

you only need two constants M,A.

 

[watertreader]

Technically, I am trying to find M only. I am just wondering if my technique has any other drawbacks or whether would it be sufficient

1) Using FFT method essentially separate the A and M. A is in the DC while M or rather M/2 is sited with the modulated signal frequency

2) Using the subtraction method, effectively I cancel out A... What is left is the M index...

That is assuming the condition are pretty ideal.