Could someone give a simple high level explanation of what is meant by AM and FM?
To keep this as simple as possible, let’s first consider transmitting a simple sine wave. What would I do to that signal to transmit it by AM and FM?
My intuition is that if I did a simple addition of the signal to a single frequency carrier signal, that I would be doing an AM transmission. Here is my guess as to how FM would work (don’t laugh). I would transmit a pure sine wave that is the sum of the frequencies of the sent signal and the carrier signal.
If my AM intuition is right, or at least close, it is easy to see how the same technique could be applied to any signal. If my FM description is anywhere near to close, I do not see how it could be used to carry a complex signal. Does it make sense to speak of an instantaneous frequency?
Observing members:
0
Composing members:
0
12 Answers
That does help. My AM guess seems to be right. I was completely off in my FM guess. What it looks like is that for transmitting a sine signal, the FM signal has a constant amplitude and varies the frequency according to the amplitude of the input signal. I am going to guess (maybe this time I will come closer) that the amplitude of the FM signal varies with the instantaneous frequency (however that is determined) of the input signal.
Yes, AM is A mplitude M odulation, FM is F requency M odulation.
My guess about changes to amplitude is probably completely wrong. Since a signal is completely determined by its amplitude as a function of time, it would be sufficient just to encode the amplitude by varying the frequency and keeping the amplitude constant.
FM is old-tech. PCM is where it’s at.
@LostInParadise Yes, what you said in your last post. That’s why FM is immune to things like noise spikes—changes in amplitude are ignored. That graphic put up by @ragingloli says it all (without words).
@jerv, Do you have an explanation for how PCM works?
We should be converting from the spacial domain to the frequency domain. Each transmission slot should contain information about the largest amplitude sine wave which should be added to all previously transmitted sine waves. MP3/JPG “kind of” does this but fails in that they “chunk” the data. There are a couple other things missing like some checksum from which errors can can be corrected and a prediction algorithm. The data transmission should be the delta from the prediction instead of the data itself. If the prediction algorithm was decent the delta-prediction data should be smaller than the original data. We should also have spread-channel communication so that variable-bitrate communication can allow for higher quality on channels which would benefit. Spread-channel communication combined with error correction also prevents jamming of a certain “station”. Ideally there should also be a list of “commonly transmitted sequences” or some sort of backward looking cache so that the same information isn’t transmitted again and again. Then you need some sort of out-of-band way to transmit other data such as closed-caption, song title, etc.
Start with a carrier wave. That is a sinusoidal waveform at the frequency that your radio is going to tune. Imagine the baseline wave at a constant strength. Notice that the frequency of the carrier wave is order of magnitudes higher than the frequency of voice or a typical human audio signal.
AM changes how strong the carrier wave is, literally making it more powerful where the signal is stronger and weak where the signal is soft. Remember that the information signal is at such a low frequency that when you superimpose its waveform as the strength of the broadcast carrier each peak-to-trough of the signal covers many many cycles of the carrier.
Your tuner will strip out the carrier frequency and what’s left is the amplitude of the signal, i.e. your original signal.
FM uses a carrier that is at a steady amplitude. This carrier frequency stays centered at the frequency you tune to but will change its frequency slightly according to the signal waveform. Your tuner will compare the broadcast wave to an internally generated wave that is stable at the tuned frequency. The difference in frequency between the reference steady wave and the broadcast wave is a ‘beat-frequency’ that is the same as your original signal.
These two work sufficiently well to send/receive human-oriented audio signals but is relatively slow compare to data signals that we like to use. That’s why more fancy ways to modulate a carrier (e.g. PCM) have been developed.
P.S. So, no your AM model is not correct, the signal is not added to the carrier. The signal is used to change the strength of the carrier.
And P.S. no your FM model is not correct, either. The amplitude of the FM signal is kept constant. It’s frequency is changed according to the signal. It stays centered at the carrier frequency but becomes slightly faster and slower according to the signal.
Answer this question 
This question is in the General Section. Responses must be helpful and on-topic.
Composing members:
0
