# AM modulation/demodulation ???

Discussion in 'Wireless & RF Design' started by Mathematics!, Dec 3, 2009.

1. ### Mathematics! Thread Starter Senior Member

Jul 21, 2008
1,022
4
Ok , for AM demodulation their are 2 ways of doing it envelope/diode detection. (i.e where a diode , capacitor , and resistor is used to obtain the envelope of the AM signal )

And product detection where a local osscilation (with same frequency, and phase of the AM carrier wave ) is mix together with the incoming carrier and filtered to get the original modulated message.

My problem is I don't understand AM sidebands.
I thought when you modulate a carrier wave using a AM you are just varying the amplitude so the frequency would stay the same for the carrier wave. The bandwidth is defined as the highest frequency minus the lowest frequency but with AM the frequency is the same so you should have zero bandwidth???

Are they defining AM bandwidth as the max amplitude - min amplitude and not the traditional max frequency - min frequency ???

Either way where are these sidebands coming from ???
Unless they are talking about the upper and lower envelope that can be produced with the diode detector to be these sidebands.

Are the sidebands just the envelope of the wave???
If not then is the positive part of the wave the upper sideband (i.e the sine wave parts above the y >=0 axies )
Lower sidebands y <=0 axis )

I just am unsure where these sidebands come from???

2. ### Wendy Moderator

Mar 24, 2008
20,735
2,499
This is almost the definition of epiphany. The amplitude concept is taught because it gets the point across easily, and is easy to understand. Strictly speaking, it is also wrong. But as an analogy, it also explains diode detection, the model (which is what it is) is just too good to ignore.

Fundamentally, when you get into radio, it is all about sidebands. These extra frequency are what carry the information. The diode detector is actually heterodyning the carrier and sidebands together to create the audio, but that would be a tough pill for a person new to the field to understand.

I don't know how computer savvy you are, but a BASIC experiment I tried a long time ago was to take a 1Mhz wave form (display too tightly to see individual wave forms) added algebraically with a 1.001Mhz wave form. You will see the classic AM waveform pop out.

3. ### Mathematics! Thread Starter Senior Member

Jul 21, 2008
1,022
4
I don't understand AM bandwidth would be zero unless the AM carrier varies the frequency as well in someway ?

This Sideband stuff is confusing me.

I know the AM wave has the same info in the upper part y >=0 axis as it does in the lower part y<=0 axis (if you where to graph it). So using the diode this just cuts off the lower part of the AM wave and the capacitor/resistor trace the envelope of the AM wave.

But I don't understand what these side bands are coming from or where they are in the graph of a AM wave? (for detecting )

I also read that SSB cann't be detected by the diode/envelop detector??? Why???

Any help would be great.
AM is varing amplitude not phase or frequency !
If it is perfect AM modulation only the amplitude will change not the frequency when the message is modulated on to a carrier at a fix frequency , I just don't get where the bands are someing from???

Last edited: Dec 3, 2009
4. ### Wendy Moderator

Mar 24, 2008
20,735
2,499
The carrier is a reference, it doesn't move, nor change amplitude (hopefully).

I understand sidebands being confusing, it is why the classic amplitude modulation model (CAMM) exists. But sidebands are the real scoop, the CAMM is just training wheels for beginners. I strongly advocate using CAMM as a teaching tool, but at some point it has to be left behind.

In SSB you are using either the upper or lower sideband to carry the audio. This has the effect of decreasing (dramatically, by half) the amount of RF spectrum that is used.

In suppressed carrier SSB you eliminate the carrier altogether. You can only do this if you have some tight tolerances on frequency control, which happened with phase locked loops, to resynthesize the carrier at the receiver. This allows a really major reduction in transmitting power, which also means what's left seems to travel further. If you are broadcasting silence there is nothing being transmitted!

I strongly recommend the graphing of the 1.000Mhz and 1.001Mhz together for yourself. It will illustrate my point more than any number of words anyone can use.

5. ### Mathematics! Thread Starter Senior Member

Jul 21, 2008
1,022
4
Ok , plotted the waves like you said

But I don't see anything special poping out.

I just don't get this ssb stuff.
AM the info is held in the changing amplitude is it not?

Also I am reading about some suppression of carrier stuff I thought data can only be transmitted if their is a carrier wave?

Also can these so called sidebands occur in FM or PM modulation modes or are they strictly for AM modulation?

File size:
11.6 KB
Views:
56
• ###### sine_1_1MHZ.PNG
File size:
11.8 KB
Views:
50
Last edited: Dec 3, 2009
6. ### Wendy Moderator

Mar 24, 2008
20,735
2,499
Try doing it as a single pass, with 3000 cycles of 1Mhz. You will see an AM modulation pattern.

7. ### lmartinez Active Member

Mar 8, 2009
224
6
Take a look at this link:

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

Digest what yo can out of the math and if you do not find it clear as mud, we will make every attempt to help you. It is not completely intuitive to the mind but the math explains it all.

8. ### Mathematics! Thread Starter Senior Member

Jul 21, 2008
1,022
4
I get the math I have my masters in pure mathematics so that is not the problem.

But when I first started learning about the analog modulation methods AM , FM , PM . I was under the impression that am varied just the amplitude of the carrier and fm varied just the frequency of the carrier ,...etc keeping everything else fixed.

But if carry = C*sin(wt +phase)
and the message = M cos(wt + phase) ( or for that matter any trig product that can change the frequency will do)

Then the product of the 2 sinisodal waves can change the frequency as well as amplitude. Using the trig product formulas you can write the sidebands as they do on your link.

So then my question is why call this modulation AM really doesn't just modulate the amplitude it also modulates the frequency as well. But then this really shouldn't be called AM modulation but AM and FM modulation all together?

Maybe they call it AM , FM , or PM based on what parameters of the formulas are being varied in the carrier not on the actual output result which could have many varying factors. If this is true then that totally changes my outlook on what AM , FM , PM are . Any one of the 3 could be made to change Amplitude , frequency , or Phase and any combination by multiplying by a suitable function.

Thanks for any clarity on this.

But why then won't the diode/enevlope detector not detect SSB ???

Last edited: Dec 3, 2009
9. ### lmartinez Active Member

Mar 8, 2009
224
6
Ok, are you familiar with the fourier transform?

10. ### Mathematics! Thread Starter Senior Member

Jul 21, 2008
1,022
4
Extremely familar.

11. ### lmartinez Active Member

Mar 8, 2009
224
6
You might want to review it once again. The answer is there.

12. ### Mathematics! Thread Starter Senior Member

Jul 21, 2008
1,022
4
I am not sure how you want me to use the fourier transform?

If you trying to get my to notice that the side bands are all sine function and the entire thing would be represented by an purely odd fourier sine function instead of the cos terms in the general fourier transform equation then ok.

But how is the envelop detector not at least allow you to hear part of this? Assuming that we are missing the cosine elements of the envelop curve ???

I guess I am missing something.

Clarity in what you are refering to would be nice

Thanks

13. ### Wendy Moderator

Mar 24, 2008
20,735
2,499
The thing is, the wave form you think of as AM is formed from the side bands. This is why I was proposing the experiment of the two frequencies. It will show this to be true.

14. ### Mathematics! Thread Starter Senior Member

Jul 21, 2008
1,022
4
Yes , I get that the modulated AM carrier can be decomposed into the sums of sine waves of different frequencies => The 2 sidebands but why won't the diode/enevlope detector detect just SSB waves???

15. ### Wendy Moderator

Mar 24, 2008
20,735
2,499
It does........

Remember, a diode is a nonlinear device, as such it heterodynes the two frequencies, the sum and the difference. The difference in this case being 1Khz.

16. ### Mathematics! Thread Starter Senior Member

Jul 21, 2008
1,022
4
what?

I am not debating on the sideband issue anymore just why a diode/envelop detector cann't be used to demodulate SSB ?

And if the 1khz issue is a big deal couldn't you just have the tuner 1khz out of tune to make up for this.

I just don't see where the major failure in demodulating comes to be?

Last edited: Dec 4, 2009
17. ### Wendy Moderator

Mar 24, 2008
20,735
2,499
You are trying to treat them as separate entities. They aren't. They work together to form an AM signal. The term sidebands is the description of the frequencies next to to the carrier. With standard AM the carrier is still required. If you remove one side or the other of sidebands it will still be AM, and the diode detector will work. It works by using the carrier and the sidebands though.

With other modulation schemes, such as suppressed carrier SSB the diode can't work.

Your question is a bit obscure, so just to repeat myself, a single sideband will work with the diode detector. Both sidebands also work with a diode detector.

If you were to graph out the scenarios I suggested you would get an AM waveform with 1Khz modulation in both cases.

18. ### beenthere Retired Moderator

Apr 20, 2004
15,815
282
You're still missing something here.. FSK = Frequency Shift Keyed. That sort of implies a shift in modulating frequency with the input keying. Morse code sent by a speed bug is one example. The CW output is one frequency with the bug on one contact, and shifts to another frequency when the other contact is made.

A modulation scheme that might fit your idea is QPSK - http://cnx.org/content/m10042/latest/

The mechanism that allows compression of data in digital transmission is the use of groups of bits to determine the modulation of the carrier. The Navy used to use pairs of bits to select one of 4 tones, and then combine 14 other selections into a composite that got transmitted. The result sounded like cats being skinned alive, but allowed 30 bits to be transmitted in parallel.

19. ### Mathematics! Thread Starter Senior Member

Jul 21, 2008
1,022
4
Questions

Ok , so let me get this straight the diode/enevlope detector will work with
SSB and DSB with carrier but won't work if the carrier is taken out in either one of these. Yes/No?

If this is true then for detecting SSB without removed carrier would you have to offset the tuner by the sideband frequency shift or keep it the same as orginal DSB not removed carrier orignal AM diode detector?

If the carrier is completly taken out i.e SSB removed carrier then is the only way to detect/demodulate it by product detectors (i.e where you use a local ossilation and a mixer/filter ,...etc etc)???

And are standard AM receiver that you buy at stores normal built for orginal AM (DSB with carrier) , or (SSB with carrier) , or SSB with out carrier or DSB without carrier?

Thanks

20. ### beenthere Retired Moderator

Apr 20, 2004
15,815
282
If you think about it, removing the carrier leaves the modulation, which is simply the original analog signal.

I might be mistaken here, but receiving a suppressed carrier signal only works if you can inject the carrier frequency in the receiver. So
is correct.

Broadcast AM has both sidebands and the carrier. It's the only way the receiver could be made so cheaply. I believe someone has already pointed out that cost controls all.