4

It is said that in frequency modulation the normal wave represents a '0' (False) and a more frequent wave represents a '1' (True).

  1. Is there any particular rule that the normal wave's frequency should continuously be the same? How actually is this "normal frequency" found out?
  2. Consider an example of a wave that has first 3 waves in a time period, then 2 waves and finally one wave. How would it frequency modulate it?
Brian Carlton
  • 13,312
  • 5
  • 44
  • 65
  • 7
    When modulating from a digital signal, this is referred to as "frequency-shift keying". "Frequency modulation" usually applies to analog signals, I believe. – endolith Nov 05 '10 at 21:40
  • But the word modulation,itself means the process of conversion of digital signals to analog signals.Applying this logic,it should be frequency modulation and the opposite should be frequency demodulation –  Nov 28 '10 at 06:10

2 Answers2

5

This is not a perfect explanation, but I hope it helps.

When something is frequency modulated there are 2 frequencies, I would not define one as "normal".

There is a frequency that is a digital 1, there is a frequency that is a digital 0.

So, when you are receiving, lets look at a conceptually simple detector. You have a bandpass for each frequency, and on the output of these, you have a rectifier which measures your relative power.

When you are receiving, you need to know the bitrate, or you need a shared clock (which requires a second connection to clock the data, or a second communication channel that cycles between its two frequencies to send the data). Often you read off of the two rectified signals as if you were reading a UART line.

Normally your data rate is significantly slower than your modulation frequency. Normally by a factor of infinity, which in engineering we can get by as approximating as 10.

Please let me know if there is something I can expand on to make this make more sense.

Daniel Grillo
  • 7,677
  • 18
  • 51
  • 70
Kortuk
  • 13,412
  • 8
  • 62
  • 85
  • 2
    you just went over my head. –  Nov 05 '10 at 17:57
  • 3
    Wouldn't this be FSK? I suppose that's what fahad was getting at, though. – Jesse Nov 05 '10 at 18:43
  • By "more frequent wave" I figured that meant a wave at a higher frequency. I initially thought fahad may be learning about modulation in class and just needed a push, I think I may have guessed wrong though. – Kellenjb Nov 05 '10 at 19:36
  • Fahad, I can try to write something that is a bit simpler. This is FSK, but it makes more sense to me to teach that. My bad if that does not help. – Kortuk Nov 05 '10 at 20:36
  • Actually I am learning it from the very basics and I dont have the idea of bitrate etc..I read in my notes how frequency modulation is done and it was just that the more frequent waves(more waves in a time period) represents a '1' and normal frequencies(less waves) represent a '0'.As its very basic so please be to the point. –  Nov 06 '10 at 06:29
  • Yes, That is called FSK if you decide you want to look into it at all. I am going to have to look at this a bit later and see if there is a way I can get you a real simulation that should make more sense. – Kortuk Nov 06 '10 at 13:20
  • 2
    "which in engineering we can get by as approximating as 10." LOL! – akohlsmith Nov 08 '10 at 18:51
  • 1
    @andrew, You wish I was joking. – Kortuk Nov 08 '10 at 19:46
  • @Kortuk.waiting.. –  Nov 09 '10 at 03:02
  • @fahad, This is on my list of todos, but I have not had a chance to write it up. – Kortuk Nov 09 '10 at 04:29
  • 1
    @Kortuk nope, as an engineer I know you're 100% correct. Approximation is the heart and soul of engineering. – akohlsmith Nov 09 '10 at 15:25
  • @andrew, keeps life interesting. – Kortuk Nov 09 '10 at 15:36
0

As for (1): before modulating a carrier, data are usually encoded - one reason is to guarantee that there are no long trains of zeros or ones. Other reason is to add some data redundancy, so that the receiver can decode original data even if some of them were demodulated incorrectly due to channel noise or interferers.

As for (2): the transition time is (should be) negligible comparing to the data rate. Transition times may in fact be further extended by some filtering (band-pass or low-pass) often required to remove interferences.

Andrzej
  • 341
  • 1
  • 3