Help with understanding modulation in AM, FM and PM radio

[onion3000]

I have a few questions about AM, FM and PM.
1. If FM (or PM) radio modifies the frequency of the signal, how is the signal picked up without having to change the frequency constantly?
2. If PM modulates phase, which modulates frequency, what is the difference between PM and FM?
3. Why is AM only used between 535-1700 kHz (and 148.5-283.5 kHz) and FM is only used between 86 and 108 MHz? Why not on higher or lower frequencies?
4. If a radio is tuned to a certain frequency, how come it can pick up other frequencies (sidebands)?
5. Why doesn't the carrier overpower the sidebands (how come we can't hear it, only the sidebands)?
Sorry for all of the questions, I'm just starting out with ham radio self-study, and am trying to make sense of it.
Thanks!

 

[sophiecentaur]

Hello and welcome.
That's a lot of questions and too much for one thread. I can give a short answer to some of them. Any radio signal consists of more than just one frequency; the signal will contain a range of frequencies (bandwidth) and a radio would normally 'tune' to the centre frequency but its input filter would admit the whole bandwidth of the transmitted signal..
The reason for the choice of the modulation method in a particular frequency band is largely arbitrary and many choices were historical, based on politics.
AM ("Ancient Modulation") was the first modulation used for broadcasting and early receivers and transmitters could only work at around 1MHz. The HF bands were developed for communications (long distance, mainly) and overseas broadcasting. AM and Single Sideband. Early (405 line) TV broadcasts used the next highest frequencies (VHF) and were basically AM. Next up are the higher VHF frequencies and they were used for FM sound broadcasting. Above that were TV (AM mainly, but now Digital TV occupies the UHF bands which were previously Analogue TV)

The difference between FM and PM? Very little, actually but there are some sophisticated differences which affect the performance. FM sound radio (and FM TV, too) uses a strong pre-emphasis (hf boost) and that in almost turns an FM signal into more or less PM. Apart from some pretty arm waving descriptions of FM, it is a much harder system to explain and the calculations rapidly get very complicated. FM can be demodulated in various ways but a basic FM discriminator circuit normally produces a varying voltage (the modulating signal) that tracks the frequency of the received signal.
5. The carrier and the sidebands are all part of the AM signal. The presence of the carrier makes it very easy to detect the modulation envelope (Crystal set is just a rectifier circuit, basically). But the carrier contains the majority of the transmitted Power and that is wasteful and occupies twice the bandwidth that's actually necessary. You can just transmit one sideband (SSB) and a suitable receiver can demodulate the signal perfectly. (It's just more complicated and expensive).

Read around about this stuff (wkik etc.) and come back with one question at a time, as a result of what you find out.

 

[onion3000]

Thanks so much for the information, and sorry again for the multitude of questions (I'm new here as you saw).

 

[nsaspook]

If you want to understand modern methods of modulation the study of I/Q is mandatory. The concept seem abstract at first but the math is pretty simple and is an intuitive way to explain (de)modulation of all kinds.

 

[meBigGuy]

Basically all modulation schemes do something to a carrier. That something always results in energy in sidebands, and there may or may not be energy at the carrier frequency. The modulation scheme is matched by a demodulation scheme that converts the energy in the sidebands and carrier into the original modulating signal. The matching demodulation scheme is generally where the "don't hear the carrier" magic happens.

In order for demodulation to work, the bandwidth of the receiver must be large enough to pass enough sideband energy to achieve the fidelity you require.

Regarding FM. While it is intuitive to think of FM simply as a carrier varying in frequency, it can also be modeled as a carrier and sidebands. In fact, the FM spectrum is infinite.
http://www.radio-electronics.com/in...y-modulation/spectrum-bandwidth-sidebands.php
For FM, it turns out that ( Carson's Rule) 98% of the signal power is contained within a bandwidth equal to the deviation frequency, plus the modulation frequency doubled.
For commercial FM they transmit a limited bandwidth.
(Note that FM deviation is based on amplitude. The frequency of the modulating signal presents another spectral dimension resulting in the bessel sidebands. Think of the rate of change of the change in carrier frequency)

So, every modulation scheme is match by a demodulation scheme, all modulation creates sideband energy, and the quality of the received signal depends on the receive bandwidth.

BTW, there are receivers that lock and track carriers, jump around, etc. There are different methods of spread spectrum transmission, etc

 

[davenn]

3. Why is AM only used between 535-1700 kHz (and 148.5-283.5 kHz) and FM is only used between 86 and 108 MHz? Why not on higher or lower frequencies?

you have a little bit of a mis-understanding of how it all works , you are not the first to make that comment and you won't be the last AM, FM and other modulation modes are used all over the radio spectrum
for example
1) old analog TV, on 48 - 60 MHz ( Channel 0 - 3) and in the VHF frequencies between ~ 170 - 300MHz ... AM was used for vision transmitter and FM for the audio transmitter
2) on the 118 - 130 MHz aircraft bands, AM is still standard
3) Amateur radio operators use AM, FM and many other modes across their many band allocations
and in that I have AM, FM and SSB transceivers that operate right up to 24 GHzcheers
Dave

 

[meBigGuy]

The commercial AM band is licensed by the FCC (or other countries regulatory agencies) to transmit in the ~535-1700 kHz. So, it is common to find receivers in that range.
The same is true for Commercial FM which is licensed to operate in the ~86 to108 MHz band.

Cell phones are licensed, TV station are licensed, Amateur Radio is licensed. Even your wifi is restricted to a band where it can operate unlicensed.

 

[Merlin3189]

1. If FM (or PM) radio modifies the frequency of the signal, how is the signal picked up without having to change the frequency constantly?
2. If PM modulates phase, which modulates frequency, what is the difference between PM and FM?
3. Why is AM only used between 535-1700 kHz (and 148.5-283.5 kHz) and FM is only used between 86 and 108 MHz? Why not on higher or lower frequencies?
4. If a radio is tuned to a certain frequency, how come it can pick up other frequencies (sidebands)?
5. Why doesn't the carrier overpower the sidebands (how come we can't hear it, only the sidebands)?

To me, your point 4 is the crux of your questions. What does it mean to have a radio "tuned" to a particular frequency?
If we put aside modern DSP techniques for the moment (*), this is achieved by using such things as inductors and capacitors to make frequency dependent circuits. These circuits have properties which generally change continuously with frequency. So you can make a circuit which responds most at one particular frequency, but it will also respond at nearby frequencies. Look at the response curve for an LCR circuit.

So for 1 and 4 the answer is that the frequency selective circuits need a bandwidth wide enough to accept all the required side frequencies and narrow enough to reject frequencies outside this band.

Bandwidth is also one of the factors affecting point 3. The first thing to note is that generally FM produces a wider range of side frequencies than AM. For broadcast AM uses about 30kHz bandwidth and FM 150kHz - 5x as much (though generally FM is better quality than AM.)
So how much badwidth is available between 535 and 1700kHz? 1165kHz is enough for 38 AM channels, but only 7 FM channels. This is one good reason for using AM rather than FM. At VHF you have 22MHz, enough for 146 FM channels and 440 AM channels, so it doesn't matter what you use.
If you looked at the old long wave band below 300kHz, you would get only one 150kHz FM channel and even AM had to reduce their bandwidth to 15kHz or less (by restricting the maximum audio frequency used.).
So AM is useable at any frequency (including light frequencies!) but FM is difficult at lower frequencies.

It is worth noting that bandwidth upto 5%-10% is easy to achieve in selective circuits while retaining good rejection outside that. But as you increase the % bandwidth beyond that, it becomes more difficult to get good rejection of frequencies outside the passband. So at medium frequencies like 500kHz, a 30kHz passband is 6%, but a 150kHz FM signal is 30% and it would be more difficult to make good filters.

AM on the MW band epitomises the heyday of radio broadcasting. Aside from theoretical limitations mentioned above, these two provide the simplest receivers. AM demodulation is so simple, requiring nothing more than a diode. MW circuits are easy to build with the required bandwidth and stability. Tuning AM is not critical: so long as the the carrier is somewhere within the passband, you can get intelligible audio.

5 What do you mean by overpower? For AM, the signal IS the carrier and the sidebands. 100% carrier and no sidebands is silence, full carrier and 25% in each sideband is maximumum modulation. If you have less carrier (or more sideband), the signal is overmodulated and the audio is distorted.
You could hear the carrier as a CW signal if you used a CW receiver setting.

(*) I would strongly disagree with NASpook : if you can get away with it, ignore I/Q and DSP in general, until you are more comfortable with things like AM, FM and general radio systems. I had been building and using radio Tx & Rx for 30yrs before I even heard of I and Q. Now that I do understand them, I find they do not help me nearly as much as the methods I used before. But I don't have any DSP equipment nor need to use it.

 

[sophiecentaur]

(*) I would strongly disagree with NASpook : if you can get away with it, ignore I/Q and DSP in general, until you are more comfortable with things like AM, FM and general radio systems. I had been building and using radio Tx & Rx for 30yrs before I even heard of I and Q. Now that I do understand them, I find they do not help me nearly as much as the methods I used before. But I don't have any DSP equipment nor need to use it.

Yes. I think I agree with that. However, I remember my Dad telling me about sidebands and phasors when I was nobbut a lad. It made sense at the time in explaining how AM works but the envelope and envelope detector diagrams were a lot easier.

 

[nsaspook]

(*) I would strongly disagree with NASpook : if you can get away with it, ignore I/Q and DSP in general, until you are more comfortable with things like AM, FM and general radio systems. I had been building and using radio Tx & Rx for 30yrs before I even heard of I and Q. Now that I do understand them, I find they do not help me nearly as much as the methods I used before. But I don't have any DSP equipment nor need to use it.

40+ years ago we were using I/Q for RF systems without fancy digital electronics (some used tubes) and there was no DSP involved. I'm sure you've heard it called the 'phasing method' or ‘Weaver`s method in the 60's for SSB.

Armstrong FM TX

http://traktoria.org/files/electronics/communication/A_Third_Method_of_Generation_and_Detection_of_Single-Sideband%20Signals__Weaver.pdf

My point was not to say the traditional way of understanding carriers and side-bands as analog signals was an unnecessary requirement because it is fundamental knowledge but things move on in technology with SDR platforms becoming the norm instead of a expensive piece of gear for professionals. So a clear understanding in the beginning of how these systems work will help a person with the circuits they see today and tomorrow.

 

[meBigGuy]

I would agree that in the olden days knowledge of I and Q was less important.

As soon as one can understand what I and Q represent mathematically then they should become the focus.
http://www.ni.com/tutorial/4805/en/
http://whiteboard.ping.se/SDR/IQ
https://en.wikipedia.org/wiki/In-phase_and_quadrature_components

All signals are about magnitude and phase, and the cartesian representations are important to understand. Especially if you start thinking about the frequency domain.

BTW, you may notice that I and Q are sometimes referred to as the "real and imaginary" components of complex numbers. Complex numbers can be represented as magnitude and phase (angular notation), or I and Q (rectangular notation) or in exponential notation. Which you prefer depends on taste and what you are trying to do.

 

[Merlin3189]

Thanks for the comments - more tempered than my rather chauvinistic one re. naspook!
Indeed I do remember reading of the Third Method. I think I even have an old Wireless World with such a design, if the paper hasn't disintegrated yet! I don't remember whether it used terms like I/Q. It would have meant nothing to me at the time. I didn't encounter that until I did a DSP course, which is why I associate the two. The main feature I remember was the large array of RC sections which tried to produce a 90^o phase shift across a decade audio spectrum. And that the whole idea was interesting but impractical.

On reflection I agree that someone learning radio now does need to understand I/Q signals. I just thought it was a sledgehammer to crack a nut for simple AM and just a bit abstract for giving an understanding of sidebands. I explain it with squiggly lines and show that adding the right combination of sine waves (a carrier and two half height side frequencies with appropriate phase) gives exactly the same squiggle as amplitude modulating a sinewave carrier (sin(c)*(1+sin(m)). These days Excel (or Matlab etc) does it easily, with matching numbers to back it up. And if more proof is needed, generate a couple of pure audio tones so that you can hear the single modulated tone that produces. Once you know and believe the reality of this, then you can take the maths on trust. Until then, however well you can do the maths, are you not vulnerable to the sort of misunderstandings in the OP?

Maybe I'm being egocentric. I learned to tune a radio by turning a capacitor and heard the signal coming into tune, passing through and turning back to centre it. On 208 from 8 til late I heard hetoerodyne whistles and could reduce them by sliding a little off centre. Even the TV where you had a channel switch, also had a fine tuning knob and frame and line frequency adjustment. These things gave a sort of visceral knowledge, so that when I learned about sidebands or the structure of a TV signal it almost seemed obvious: what they were saying just fitted in with what I experienced. Nowadays people come with a very different experience base, so perhaps need to understand in different ways.

Certainly now people have the tools to make this understanding easily available. A $10 USB SDR dongle and SDR# will let you watch the sidebands as you listen to the audio. You can narrow or widen the passband and slide the signal across it to see and hear the results. (Maybe there is other software which would also let you add notch filters or to see an I/Q display of a signal?)

Anyhow, let's hope Onion takes up Sophie's suggestion and sends follow up points, which we can all address in our own preferred ways.