| Vocademy |
Frequency modulation varies the frequency of the transmitter instead of the amplitude. The sidebands produced by FM are much more complicated than those produced by AM. However, the same bandwidth requirements apply to FM transmission as with AM transmission.
FM has the advantage of amplitude noise immunity. Changes in the amplitude of the transmission caused by outside sources, such as lightning, light switches, fluorescent lights, etc., are not received by an FM receiver.
An FM receiver is virtually identical to an AM receiver from the RF stage through the IF stage. The difference is the FM detector. An FM detector must produce a DC voltage that varies with the changes in the frequency of the carrier. Several circuits do this. One that is discussed in detail elsewhere is the phase-locked loop (see digital electronics). Other FM detectors take advantage of how inductive reactance and capacitive reactance vary with frequency. For example, when the frequency of a signal placed across a capacitor increases, the capacitive reactance decreases. This lower reactance acts like a lower resistance. Therefore, the voltage across the capacitor decreases. As the frequency decreases, the voltage correspondingly increases. This causes the average voltage across the capacitor to vary at the rate which the frequency varies. The final signal is sent through a low-pass filter that blocks the high-frequency carrier, passing only low-frequency variations. This is a reproduction of the original modulating signal.
FM broadcasting had a hard time catching up in popularity with AM broadcasting. This was mostly due to the FCC (Federal Communications Commission) changing the allocated frequency band from around 50 MHz to around 100 MHz when television was developed. The entire system of FM transmitters and receivers had to be scrapped (some people think that David Sarnoff made a back-room deal with the FCC to put Edwin Armstrong out of business). It took a long time for FM to make a comeback. It wasn't until the 1960s that it became a significant part of broadcasting again. Broadcasters needed to increase the demand for FM reception to increase revenue. FM already had sufficient bandwidth to carry frequencies up to 20 kHz and thus reproduce high fidelity sound. The broadcasters wanted to add stereo (two-channel) sound to complement stereo phonographs that were becoming popular. The FCC demanded that proposals to do this be compatible with existing monauaral receivers.
Unlike AM broadcasting, where each channel has a 10 kHz bandwidth, FM broadcasting has a bandwidth of 200 kHz for each channel. This means that, not only can an FM transmission carry full-spectrum audio, but could carry ultrasonic frequencies to 100 kHz as well. FM transmissions use subcarriers to make use of this extra bandwidth.
A subcarrier is just another frequency that can be used to modulate the main carrier. You can modulate a carrier with audio frequencies. You can modulate it just as well with ultrasonic frequencies. Analog TV modulates a carrier with frequencies as high as 4 MHz. Also, a carrier frequency doesn't need to be in the megahertz range. The AM radio band centers on 1 MHz, but you can modulate 100 kHz just as easily as you can modulate 100 MHz. Let's say you have a 40 kHz sine wave. Now, modulate that with an audio signal just like you would any other carrier. What you get are sidebands above and below the 40 kHz carrier. If you do this with a full-spectrum audio signal (20 Hz to 20 kHz), you get sideband frequencies ranging from 20 kHz to 60 kHz (40 kHz plus and minus 20 kHz).
|
|
| First, modulate a 40 kHz carrier with a full-spectrum audio signal… |
 |
| and you get this. |
Let's pause for a moment and see what we have. We modulate a 40 kHz sine wave and we end up with frequencies ranging from 20 kHz to 60 kHz. If you can modulate a carrier with frequencies ranging from 20 Hz to 20 kHz, why can't you modulate a carrier with frequencies ranging from 20 kHz to 60 kHz? You can. Let's now modulate a 100 MHz carrier with all the frequencies we have after modulating the 40 kHz carrier, including the original audio frequencies.
 |
| Then modulate a 100 MHz carrier with that… |
Now we get sidebands created from the original signal and the original modulated carrier—a radio signal within a radio signal.
 |
| and you get this. |
We have the original baseband audio signal that was used to modulate the 40 kHz carrier. This produced an AM signal (carrier plus sidebands) that, along with the original baseband signal, became the new baseband signal. This new baseband signal is then used to modulate a 100 MHz carrier producing a new modulated signal. The original 40 kHz carrier now has counterparts in the sidebands of the 100 MHz modulated signal. These are at frequencies of 99.94 MHz (40 kHz below the 100 MHz carrier) and 100.04 MHz (40 kHz above the 100 MHz carrier). These carrier frequencies within the sidebands of the 100 MHz carrier are called subcarriers.
Now, to fully utilize this signal we need a receiver within a receiver. First, we demodulate the 100 MHz signal to recover the signal that was used to modulate it. We now have the original audio baseband and the 40 kHz modulated carrier. The receiver within the receiver demodulates the 40 kHz carrier.
What use is this? So what if we have a copy of our audio baseband riding on a subcarrier? Why not modulate the subcarrier with a completely different audio signal? Now we have a full-spectrum audio signal in the baseband and a completely different audio signal carried on the 40 kHz subcarrier. Modulate the 100 MHz carrier with all this and we now have two completely different audio channels on a single 100 MHz radio carrier.
This is how FM stereo is done, almost.
To start with, FM broadcasts only deliver frequencies up to 15 kHz. This is not full-spectrum audio, but it's good enough for the average listener. The narrower bandwidth allows a slightly lower subcarrier of 38 kHz Instead of a 40 kHz. Secondly, since the subcarrier itself is just wasted energy (see sideband radio above), it is filtered out (using a balanced modulator), leaving only the sidebands. To properly demodulate this double-sideband, suppressed carrier signal you need a perfect copy of the original carrier. To create this carrier, a low power 19 kHz sine wave, called the pilot carrier is transmitted along with the rest of the signal. This pilot carrier fits neatly between the top of the audio baseband signal and the bottom of the lower sideband of the subcarrier signal. Finally, the stereo signal is not sent by putting one channel in the baseband and the other on the subcarrier. The left and right channels are mixed together to create a monaural signal for the baseband. This allows monaural FM receivers to get a usable signal instead of just the left or right channel. The left and right channels are then passed through a differential amplifier to create a difference signal that can be used to separate the left and right channels from the monaural signal. This difference signal is used to modulate the 38 kHz subcarrier.
We now have the superbaseband[1] signal needed to modulate the FM carrier. The result is the following FM broadcast signal.
 |
| A broadcast FM stereo transmission. |
An FM stereo receiver is a receiver within a receiver; an FM receiver demodulates the main signal. After this, a sideband receiver within the FM receiver demodulates sideband signal that contains the left-minus-right difference signal. The receiver extracts the 19 kHz pilot carrier with a high-Q filter and doubles it to 38 kHz. This is used to demodulate the double-sideband 38 kHz transmission that contains the stereo difference signal.
The 19 kHz pilot carrier may be doubled using a phase-locked loop. Another possibility is a frequency multiplier.
A frequency multiplier starts with a non-linear amplifier. A non-linear amplifier doesn't amplify all DC voltages equally. For example, let's say an input of 0.1VDC produces an output of 0.1 VDC (a gain of 1) but and input 0.5 VDC produces an output of 0.4 VDC (a gain of less than 1). A sine wave passed through such an amplifier becomes distorted because the various voltages are not amplified equally.
 |
| A sine wave (left) and a distorted sine wave (right) |
Recall from basic AC principles that any wave shape other than a pure sine wave contains various harmonic frequencies (multiples of the fundamental frequency). Therefore, the 19 kHz pilot carrier, after non-linear amplification, contains harmonic frequencies of the original 19 kHz (i.e., 38 kHz, 57 kHz, 76 kHz, etc.).
The next characteristic of a frequency multiplier is a tuned circuit. Recall from basic AC circuits that an oscillator is an amplifier that has a gain of 1, combined with a tuned circuit such that it has positive feedback. The frequency multiplier for an FM stereo receiver would be tuned to 38 kHz. If it had a gain of 1, it would oscillate, but given a gain of less than 1 it will not oscillate but will be an amplifier that works only at 38 kHz. Combine this with the non-linearity and it will amplify the 2nd harmonic of the 19 kHz pilot carrier, which is 38 kHz. The final result is that 19 kHz in gives you 38 kHz out—frequency multiplication.
The sum of the left and right channels is sent along with the difference signal (left minus right) to a summing amplifier (a simple mixer). The output of the summing amplifier will be only the right channel. Mathematically it works like this:
(L+R)+(L-R) = (L+L)+(L-R) = 2L
The left channel gets canceled when the sum and difference are mixed together, leaving only the right channel. To extract the left channel the signals are mixed through a differential amplifier[2]. The mathematical result of this is:
(L+R)-(L-R) = (L-L)+(R-(-R)) = (L-L)+(R+R) = 2R
The double-sideband 38 kHz transmission extends to 53 kHz (the end of the 38 kHz subcarrier's upper sideband). This leaves 47 kHz of space before hitting the 100 kHz limit. Here there is room for more subcarriers. FM stations use other subcarriers to carry private transmissions to increase revenue. These might be pager signals, or private music or voice transmissions. Muzak® is a famous company that used to lease subcarriers to broadcast commercial-free "elevator music" (Muzak® now uses satellites). This subcarrier service is called Subsidiary Communications Authorization (SCA).
 |
| The upper sideband of a broadcast FM multiplex transmission with SCA services. |
FCC regulations limit the total amount of power allowed in the transmission. The SCA services use only a small part of that power. This is because power put into the subcarriers is power that is not in the main channels. Too much power in the SCA services would limit the range of the main broadcast.
A newcomer to FM broadcasting is the Radio Broadcast Data System (RBDS). This delivers a low-speed digital data stream to display data on the FM receiver, such as program information. This is a double-sideband suppressed carrier signal at 57 kHz, which is the third harmonic of the 19 kHz pilot carrier.
In summary, the final FM transmission is built first as a baseband. That is:
The baseband signal (the left-plus-right main audio, the pilot tone, the double-sideband transmission carrying the left-minus-right signal and any SCA subcarriers with their sidebands) is used to modulate the main carrier. This creates the upper sideband as illustrated above and its lower sideband mirror image.
—————————| Vocademy |