frequency modulation

[PG1995]

Hi

In amplitude modulation both bands, lower and upper bands, carry the same information, and both bands are mirror image of each other. Likewise, in frequency modulation, two bands are produced which are mirror images of each other. Theoretically speaking, do both FM bands, upper and lower, carry the same information? Personally, I think, they do carry the same information but perhaps practically it's not possible to implement a single band transmission for FM like we do for AM. Please let me know what you think. Thanks a lot.

Regards
PG

 

[JimB]

Theoretically speaking, do both FM bands, upper and lower, carry the same information?

Yes

perhaps practically it's not possible to implement a single band transmission for FM like we do for AM

Never seen it done, there would be no advantage in doing so.

However, a little "trick".
It is possible to demodulate FM using an SSB Receiver, PROVIDED that the deviation is very narrow.
This is in no way a practical thing, but when you look at a very narrow deviation FM signal with a spectrum analyser, it looks very similar to an AM signal.
Before anyone says this is "slope detection", it is not. it is receiving the FM signal with an SSB receiver. As soon as the deviation is sufficient for the higher order side frequencies to become significant, the trick no longer works.

JimB

 

[PG1995]

Thank you, JimB.

Q1:

there would be no advantage in doing so.

Why wouldn't there be any advantage? If you are using only a single sideband transmission then you are effectively saving half of the usually required bandwidth. Please help me with it.



Q2:
It says here that noise is interference generated by by lightning, motors, automotive ignition systems, and any power line switching that produces transient signals.

During lightning a static charge is built up and ionization of the atoms take place. In case of ionization, electrons are transferred from one atom to other (electrons jump from shell to another, and from one atom to another) and this is accompanied by emission of electromagnetic radiation and this radiation is what interferes with the signal. If my understanding is not correct, please help me.

Something similar happens in case of automotive ignition systems where the spark of a sparking plug functions as a mini lightning bolt.

But what really happens in case of motors and power line switching. Perhaps it has something to do with inductive effect.

For an inductor, when the switch is closed, no current passes through it for a moment but maximum voltage appears across the inductor.
Then the voltage starts decreasing and current starts increasing. Magnetic flux would also increase along with increasing current. When voltage across inductor has reached zero value, the current and magnetic flux have reached their maximum values. The inductor stores its energy in its magnetic flux.

Could you please help me with this? Thanks.

Regards
PG

 

[steveB]

PG,

If you go back to the other thread where I pointed to a reference ( here it is again **broken link removed**) that worked out the math for FM, you will see that an infinite number of sidebands is produced, assuming you implement something physical that mimics that math. However, allowing all those sidebands to actually be transmitted would not be practical because you would interfere with other channels, hence additional filtering is needed. So, there are two sidebands, but I wanted you to be clear on how those "only" two are obtained.

 

[JimB]

Why wouldn't there be any advantage? If you are using only a single sideband transmission then you are effectively saving half of the usually required bandwidth. Please help me with it.

In telephony you do not use FM where you are looking for a minimum bandwidth solution, you use SSB.
SSB is a filtered AM signal. It works because the sideband which remains after filtering contains all the information required to re-construct the audio signal except the carrier which is re-injected at the receiver.

FM is totally different, if you were to filter the FM to leave one sideband and then re-inject the carrier, you would probably end up with some form of AM.
Viewed in the time domain (on an oscilloscope) an FM signal has a constant amplitude.
Viewed in the frequency domain (on a spectrum analyser) you will see that as the signal is modulated, producing the sidebands, so the amplitude of the carrier is reduced. Now comes a real brain squeezer, at certain deviations the carrier actually dissappears!
So coming back to our FM signal which we have filtered to be like SSB, to be in with a chance of modulating it correctly we need to re-inject the carrier with the correct phase and amplitude (which is varying with the modulation). How you would achieve that I have no idea, and neither has anybody else as it is not worth the effort.

JimB

 

[PG1995]

Thank you, Steve, JimB.

@Steve: Thanks for the link. I understand that two sidebands are created but this time I was wondering that why there is never a SSB FM. But you see here that Google does have some entries on it.

I'm not sure if I fully understand the reason for not generating SSB FM signal but I will give it some time before making few follow-on queries. But the fact is that generating a SSB FM signal is not worth the effort, in practical terms.

This text is helpful to understand JimB's statements from post #5. Thanks.

When you have time, please also help me with Q2. Thank you.

Regards
PG

 

[steveB]

@Steve: Thanks for the link. I understand that two sidebands are created but this time I was wondering that why there is never a SSB FM.

Yes, i understood your question and I understood you knew that there are 2 sidebands. What wasn't clear to me, and what still is not clear to me, is whether you understand that the math in that link shows that in theory an infinite number of sidebands is produces by the mathematical operation of frequency modulation. This is an abstraction, of course, and not exactly what gets implemented, but it is important to understand this point.

 

[PG1995]

Thanks.

Yes, i understood your question and I understood you knew that there are 2 sidebands. What wasn't clear to me, and what still is not clear to me, is whether you understand that the math in that link shows that in theory an infinite number of sidebands is produces by the mathematical operation of frequency modulation. This is an abstraction, of course, and not exactly what gets implemented, but it is important to understand this point.

To be honest, I didn't check the math there carefully but this text provides you with enough mathematical detail so that you get the overall understanding. I can see that Bessel coefficients extends to infinity. If you want me then I will definitely go through that math in more detail in next couple of days. Thanks.

If you have time then please also help me with Q2 from post #3. Thank you.

Regards
PG

 

[steveB]

Hi PG,

You can go through that math in detail if you want, but I'm not suggesting that. Just be aware of the general approach and the end result that multiple side-bands are produced.

For Q2 in post #3, I find that hard to answer. Power line interference basically is likely to get to the power terminals of any devices in the circuits. We try to filter the power with caps and sometimes even caps and coils, but filters only reduce noise, - and don't eliminate it. Hence, a very strong noise source on a power line can sometime get through. There is also the possibility of inductive coupling in some cases, and even electromagnetic coupling sometimes.

The case of distant lightning causing noise on an AM receiver is basically the impulse-like pulse of the lightning generates noise over a broad spectrum, and some of that noise happens to be in the bandwidth of the received channel. But, that's different than the power line interference.

 

[JimB]

But what really happens in case of motors and power line switching.

There are many things happening with power lines.
Whenever some device is switched on/off there is a change in current. Unless the switching is done at a zero volt/current crossing point in the AC cycle, the change will be abrupt.
As we have previously discussed abrupt changes create high frequencies (think sharp edges on square waves).
Another undesirable condition with a power line is where there is tracking across the insulators.
If the insulator is contaminated with industrial pollution or salt spray from the sea, small currents will track across the surface of the insulator. Again this can cause horrendous HF noise.

A motor such as an induction motor when it is running would probably create no noise.
However a motor with a commutator, where there are many switching events per turn of the rotor has scope for creating lots of high frequency noise, and if the commutator is dirty or there are faults in the motor, the commutator can look like a fireworks display and the HF noise is horrendous.

Poorly designed or faulty equipment such as motor speed controllers can inject lots of HF noise back into the power lines.

JimB