Fundamentally, the operation of amplitude modulation is a multiplication of the two signals.

The square wave has an infinite series of harmonics and those are carried through the multiplication process.

 

you can generate the spectra of the two components (the sinusoidal carrier and the square signal) by calculating the Forier transform of them.

Then, using this

http://en.wikipedia.org/wiki/Convolution_theorem

you can show that the spectrum of the AM signal is the convolution of those two spectra.

the convolution doesn't remove the harmonics- they are still there even if you choose the frequencies and phase so that the transitions of the square wave correspond to zero crossings in the sine wave.

 

I'm waiting for studiot to accept that, but I'm not sure he understand the problem.

 

Now the point I made earlier is that the spectrum of a sine wave amplitude modulated by a sine wave is the same as the spectrum of a sine wave frequency modulated by a square wave.

 

There's no way to tell those signals apart so they would obviously interfere with each other.

 

There's no way to tell those signals apart so they would obviously interfere with each other.

Receivers demodulate signals; thus, this statement seems untrue sometimes. I must misunderstand your point.

"sometimes" doesn't enter into it.

I'm talking about a specific pair of signals.

One is created by FM, the other by AM, but the output is the same.

Either decoder would give an ouput, but the outputs would be different.

Which one is right?

Oh, OK.