AM/FM on one carrier

Andy replies to Joerg,

Thanks, I'll see if I can get a copy.... But remember, I'd rather be proficient with an old tool than attempt to learn every new one that comes out...

I still write my programs in BASIC. I can do anything in BASIC that can be done in Fortran, Cobol, C, Java, ADA, and whatever the hell the "latest craze" is...... Having learned, and abandoned, most of them long ago when the fad changed... ... not saying that LT Spice is a fad.... just an unnecessary step for me ....

I've used the HP harmonic balance simulator, called SYMPHONY or something -- I don't remember . I guess it works good, but I would have had to practice with it for a year before I could have gotten as comfortable as with PSPICE. At Texas Instruments, I used their in house version called T I SPICE. It worked OK. Not as self apparent as PSPICE, but that was probably because I resented having to learn a new language that did exactly the same as one I already knew...Like going to Mexico and having to speak German when I already know Spanish... .....

.... like having to solve a ciruit with differential equations after getting good at LaPlace transforms.... just seemed like 'busy work'....

Any way, I'll take a look at LT SPICE .... sometime..... Thanks for the tip....

Andy

If you define "carrier" as the amplitude of the central spectral line, excluding the sidebands, then an AM carrier is constant irrespective of AM modulation. The carrier amplitude of an FM signal varies with modulation index according to a Bessel function, and has nulls at various points.

So if you FM modulate a carrier with the right sinewave signal at the right deviation, the carrier disappears completely, which has got to confuse an AM detector.

I'm guessing that AM and FM modulation can be used fairly orthogonally at very low modulation levels.

John

--- Considering what happens in amplitude modulation, that is that the amplitude of the carrier is forced to vary with the modulating waveform, I'd say that the error was yours.

Consider CW, where the carrier is either on or off. Is CW amplitude modulation? Definitely. Does the amplitude of the carrier vary during the process of modulation? Definitely.

Let's look at a slightly more involved example/thought experiment. Imagine you have a rig set up so that you can plate modulate the final but that instead of varying the final's B+ by speaking into the microphone you vary the amplitude of the plate supply by using a pot. Rotate the pot counter-clockwise and the plate voltage decreases; rotate it clockwise and it increasese.

Will rotating the pot CCW decrease the amplitude of the carrier? Yes. Will rotating the pot CW increase the amplitude of the carrier? Yes.

If a modulating waveform causes the plate voltage to decrease will the amplitude of the carrier decrease? Yes.

If a modulating waveform causes the plate voltage to increase will the amplitude of the carrier increase? Yes.

-- John Fields Professional Circuit Designer

"John Larkin" wrote in message news: snipped-for-privacy@4ax.com...

You are mixing domains here. "Constant" refers to a time domain concept. "Spectral line" refers to a frequency domain concept. If you are saying the set of AM spectra starting with the same unmodulated carrier all have the same line height at the carrier frequency, I would agree with an exception for the results of DC modulations.

Sure. The crux is how low the modulation levels need to be given the OP's requirement and receiver attributes.

--
--Larry Brasfield
email: donotspam_larry_brasfield@hotmail.com
Above views may belong only to me.

Of course. Time and frequency are just different interpretations of reality.

Can't a spectral line vary as a function of time?

"DC modulations"? Oh oh, trouble coming up I think.

Just get his specs and do the math.

John

You avoid a lot of debate if you define DC to be the average value over the period of observation.

John

Here is my 0.02 - worth every penny ....

I thought this would work so I decided to try it in the lab:

1) Racal-Dana model 9087 AM/FM RF generator 2) "Ideal" demodulator - see below

The demodulator consists of an I.F. AGC amplifier feeding a Linear tech

14-bit ADC running at 50.4 Ms/S, feeding an FPGA computing amplitude (sqrt(I^2+Q^2)) and arctangent. Arctangent is differentiated to give FM

- without pre-emphasis or de-emphasis.

fc = 10.8 MHz, Rx Analog BW = 500 kHz, digital filtering to 22 kHz Modulation: 1kHz sine AM, 1.5 kHz sine, FM. FM = 5kHz deviation, AM =

90% modulation.

Result: at 10 mV carrier input, the demodulated signal looks ok in the time domain, both FM and AM. When I decrease the signal level to 40uV (-75 dBm), I see a little "fuzz" on the FM waveform that seems synchronous with the AM modulating waveform negative peaks.

Now, a question for the OP: Why? there are experts here (somewhere) who can help you do this a better way (if it's not homework). Tell us more about the application.

Frank Raffaeli

I do believe we have been through all this before, and I didn't change your mind then, nor probably now.

There are probably many methods of "AM" that will result in carrier voltage modulation while also creating sidebands, they are not a usable form of AM, as they result in godawful distortion with wideband splatter. Meaningful discussions of AM by inquiring minds should be oriented toward meaningful processes, not processes that generate unusable signals. My dabble into this was in 1954, and was short-lived because the result was what I anticipated.... But I just had to see it. If someone starts a thread with their interesting modulation method test results, I will post mine to it also.

Getting to the bottom line, if you use a selective frequency voltmeter to observe an unmodulated carrier, and then you (correctly) AM modulate the carrier, you will observe that the carrier amplitude does not change.

And the modulation power is all in the sidebands.

Don

Don

DC modulation is not AM modulation (in it's classic, intended sense). It is coded changes; turning the carrier on and off as in international Morse code. I imagine one could use a code with multiple levels of carrier amplitude along with signal on-off code, but it would not be AM either.

Don

You'll get FM edge detection out of the AM IF filter as the FM modulation moves the carrier frequency up and down.

Lukas Louw

Yup.

John Perry mentioned stereo and MUSAK, though he got the SCA subcarrier mixed up with the (newer) RDS subcarrier.

In case the Cached link (Google's 64.* server) bogs down (as it did for me): http://66.102.9.104/search?q=cache:EyL-PsUJ5WEJ:

http://66.102.9.104/search?q=cache:EyL-PsUJ5WEJ:

Hell no, just the opposite: I'm an engineer.

John

They got injected through the miracle of modulation. :-)

Look at your standard class A, B, or C plate-modulated transmitter.

You've got the carrier going in through the grid and coming out the plate going into the tank circuit.

In series with B+, you interpose a source of the modulation signal, like the secondary of the modulation transformer.

Now, it has been said that the instantaneous amplitude of the carrier hits zero at the most negative excursion of the modulating signal, at 100% modulation. Well, yeah, maybe, the instantaneous value in the time domain, but _that does not make the fundamental go away!_ What we are seeing is the instantaneous value of the entire modulated wave, at which point the phase of the carrier and the phase of what's on the sidebands cause a resultant of zero - at that instant.

In fact, if the AM carrier goes away, why do they go to such great lengths to suppress it for SSB or DSBSC? ;-)

Now, if you try to frequency-modulate it simultaneously, well, you'd get the set of sidebands that that particular frequency/deviation cause - the way this was explained to me is "It's simply the Fourier transform." OK. ;-)

But demodulating them - well, Frank Raffaeli reports in another branch, that he's done it with DSP:

-------- Here is my 0.02 - worth every penny ....

I thought this would work so I decided to try it in the lab:

1) Racal-Dana model 9087 AM/FM RF generator 2) "Ideal" demodulator - see below

The demodulator consists of an I.F. AGC amplifier feeding a Linear tech

14-bit ADC running at 50.4 Ms/S, feeding an FPGA computing amplitude (sqrt(I^2+Q^2)) and arctangent. Arctangent is differentiated to give FM

- without pre-emphasis or de-emphasis.

fc = 10.8 MHz, Rx Analog BW = 500 kHz, digital filtering to 22 kHz Modulation: 1kHz sine AM, 1.5 kHz sine, FM. FM = 5kHz deviation, AM =

90% modulation.

Result: at 10 mV carrier input, the demodulated signal looks ok in the time domain, both FM and AM. When I decrease the signal level to 40uV (-75 dBm), I see a little "fuzz" on the FM waveform that seems synchronous with the AM modulating waveform negative peaks.

Now, a question for the OP: Why? there are experts here (somewhere) who can help you do this a better way (if it's not homework). Tell us more about the application.

Frank Raffaeli

-------

Cheers! Rich

--
You\'ll notice that the _average_ carrier amplitude doesn\'t change, but
if you look at it with something a little more fine-grained, like a
scope, you\'ll notice that it does, I believe.

Sheesh.

The carrier amplitude coefficient ("A") is a constant for LC-AM, but cos(wc*t) obviously isn't. Also obvious is that constant amplitude coefficient transforms into the f-domain as a constant too. So no, it is not "time domain" concept. It is a signal definition concept that equally exists in both t- and f-domains.

LC-AM:

[A + m(t)]*cos(wc*t) ; by definition

A*cos(wc*t) is the carrier, and is of constant amplitude.

m(t)*cos(wc*t) is in practice near the carrier, but isn't the carrier.

"DC" is basically eternity, because that's how long you have to wait to make sure it never turns off or changes.

So if the "DC" is not eternal, then it is not "DC" but "AC."

That means it may be a low frequency, but low isn't zero. So the sidebands may be quite near the carrier, but they are not the carrier.

Transformation to the f-domain tells the same story.

In short, the term "DC modulations" is at best confusing and at worst simply wrong.

snipped-for-privacy@4ax.com,

modulation.

modulated.

this is a bit of a confusing way to think of it to my mind, the carier is modulated by varying its amplitude wich must mean it is definatly changing in amplitude, otherwise how else do those sidebands apear, unless you actualy just add them individualy, the sidebands are an inevitable part of a sinewave changing in amplitude, ... yet the single spectra of the carier displayed on an analyser does as you say stay constant.

If the FM sidebands are within the flat part of the IF response the detector should pick up only the AM.

it al depends how much crosstalk you can get away with, if its a stereo diference signal then this wont have as much impact as if its two seperate audio chanells, also if the AM signal is needed to be received on AM only receivers.

i gues you could do a fair job with predicting the crostalk and canceling it out, certainly with a DSP anyway.

Colin =^.^=

You sure are an optimist. Laughs.

"DC modulations" is an oxymoron.

Hello Jeff,

Thanks. SCA, couldn't remember that name. In Europe they also have traffic alert signals overlaid. At least since the 80's or 90's. I bought an Audi at a dealer there and when I drove off the lot the radio came on out of the blue with some info about a jam on an autobahn nearby. It almost gave me a scare since I didn't know they had that. I mean, you drive off in your new car all alone and suddenly a loud voice is heard from the back.

Regards, Joerg

That back-seet driver is the built-in wife...