Exploring FM & AM: Questions & Answers

[mearvk]

Hi guys.

Two questions:

1. AM, as opposed to FM, is closer to what FFT is naturally used for right?

2. Is there a physical reason why FM is used in higher frequency bands and not AM? Could, for instance, FM be in the AM band and vice versa? Would transmitting, hypothetically, AM signals in the FM band mess up the FM signals on the same frequency?

Thanks.

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

 

[davenn]

Commercial FM broadcasting as in what is used in the 88 - 108MHz 20MHz bandwidth
is stereo FM and uses ~ 150kHz per channel
the AM band 576kHz to 1600kHz is only ~ 1MHz in bandwidth

there would only be room for maybe 6 - 8 FM stereo radio stations

see the problem ?

Dave

 

[mearvk]

Yeah, I see the problem.

Thanks.

 

[sophiecentaur]

There's a lot of history here. AM was all that was available and lf / mf were the only channels available with the early technology. A bit later, frequencies up to about 30MHz were established for AM and a bit of SSB for comms. Since then, there has been no incentive to change the band usage on those bands; compatibility has always been a problem.
In an attempt to improve sound quality, FM was adopted, involving fairly hefty frequency deviation, in order to improve on signal to noise and immunity to interference. Band 1 was already taken up for early TV so the next available band was around 80 -100MHz (the present FM bands are around here). That's a comparatively massive bandwidth and provided room for 50kHz channel spacing with high deviation FM giving the 'FM Noise Advantage'.

You can actually use narrow band FM in little more than the required Double Sideband AM channel bandwidth but there is no 'FM Noise Advantage'. Transmitters are cheaper and modern receivers are just as easy for FM as AM so narrow band FM is often used for narrow band, low quality comms. Around 150MHz, you find both AM and FM in use.
It goes on and on but you will have got the broad picture - it's all very pragmatic engineering stuff.

1. AM, as opposed to FM, is closer to what FFT is naturally used for right?

You would need to expand on that. Modulation and FFT are two separate issues. Are you referring to the fact that the FFT (the spectrum) of the baseband signal is identical to the (reflected pair of) AM sidebands but the FM spectrum is different?

 

[mearvk]

I'm not sure what sidebands are.

My original thinking was along the lines of taking a carrier signal and modulating into it an audio signal (the music). So, for instance, if I was listening to 96.1 MHz then I might take the spectral input from the antenna, remove all frequency data except that from 96.1 to 96.12 and then remove the carrier signal via some sort of FFT process and be left with just the audio data.

I was saying that this, to me, at the time, seemed closer to AM than FM in terms of the (de)modulation processes being similar. Generally though, I'm just trying to understand how radio transmissions work.

 

[rbj]

AM and FM have nothing to do with the FFT. the FFT is a "fast" method of calculating the DFT.

FM could be used in the AM band if it were "narrow-band FM". narrow-band FM takes up no more bandwidth than does AM. but the wide-band FM used in commercial broadcasting takes much more bandwidth.

 

[sophiecentaur]

I'm not sure what sidebands are.

My original thinking was along the lines of taking a carrier signal and modulating into it an audio signal (the music). So, for instance, if I was listening to 96.1 MHz then I might take the spectral input from the antenna, remove all frequency data except that from 96.1 to 96.12 and then remove the carrier signal via some sort of FFT process and be left with just the audio data.

I was saying that this, to me, at the time, seemed closer to AM than FM in terms of the (de)modulation processes being similar. Generally though, I'm just trying to understand how radio transmissions work.

This is too complicated for you to hope to 'invent' what modulation and demodulation involve, all on your own. The fact that you don't know what sidebands are implies that you need to do some serious reading around about the basics - rather than trying to get there by question and answer. That method can work OK when you're sitting with someone in a one-to-one session but the time lag on a forum makes it fail.
Suffice to say, you are conflating two ideas where it isn't really appropriate. Of course the two terms are to do with 'frequencies and things' but your connection between them is not really valid. Start at the beginning with Amplitude Modulation on Wiki and you will find out about sidebands and how demodulation can be achieved. The Fourier transform is just a way of 'looking at things' - nothing more and nothing less (very important, though).