How AM Radio Works
Return to Home
Return to Resources
1. Introduction
2. Principle of Amplitude Modulation
3. Tuned Radio Frequency Operation
4. Superheterodyne Operation
5. Demodulator
6. Conclusion
Introduction
Principle of Amplitude Modulation
A signal can be understood in time domain or frequency domain. The figure above is in the time domain because it shows a sinusoidal over time and the modulated signal has changing amplitudes over time. However, other concepts like bandwidth is difficult to visualize in time domain so we enter the frequency domain. It was mentioned earlier than audio frequencies range from 0 to 20KHz and we could represent the spectrum of audio frequencies as a triangle with more strength at lower frequencies than high. On the other hand, a constant sinusoidal signal produced by an oscillator is simply represented as a spike. For example, a good 50KHz oscillator produces a frequency that is exactly 50KHz and does not have a range below or above 50KHz.
In the frequency domain, the notion of bandwidth is clearer because it represents the strengths of each frequency in a broad range of frequencies. Audio frequencies range from 0 to 20KHz so we want a circuit that can easily transmit this range so it has to have a bandwidth of at least 20KHz. A circuit with a narrower bandwidth than the range of audio, say 1KHz, would attenuate or cut out the higher frequencies in audio and distort the output. On the other hand, a circuit with narrow bandwidth suitable for audio is useful for selecting a station as explained next.
In addition to antennas being incapable of transmitting audio frequencies, the audio spectrum itself would prevent the possibility of many radio stations. If all audio frequencies have the same range of 0 to 20KHz then picking out the desired audio is impossible; this is analogous to listening to many people talking at once. Amplitude modulation basically shifts the center of the audio range to a fixed RF station. Every station has their own frequency and the receiver can choose which station to select and extract the audio. In the frequency domain, the figure below explains the action performed by modulating and demodulating. Traditional demodulation does not return all of the audio strength to the 0 center because it creates a sum and difference effect that also creates a copy at two times the carrier frequency but all the math behind this will be quietly ignored. However, the audio stage has a low pass filter to pass only the audio centered at 0 and to eliminate the higher frequencies. Also note that the strength of the signal when modulated is equally split on the positive and negative carrier frequency fc to indicate that signal power is neither created nor destroyed.
In theory, a station can be selected by using a narrow bandpass filter centered at the desired station. The filter would effectively eliminate other stations, as demonstrated in the animation below. Note that each triangle is color coded to represent an unique radio station and all of the triangles exist somewhere in the RF range of the frequency spectrum.
So what is the problem with this idea? Nothing really, but in practice an adjustable narrow bandpass filter is difficult with generic inductors and capacitors. Inductor-Capacitor (LC) circuits can resonate at a specific frequency and basically tune in the station, but they do not eliminate adjacent frequencies very well and are termed broad bandpass filters.
Tuned Radio Frequency Operation
Assuming a narrow bandpass filter was practical, the simple approach is to use a single RF tuner and amplify the weak audio signal through several stages. In reality, a simple LC circuit passes through several stations at once and it is too complicated to design an adjustable narrow bandpass filter that operate at RF frequencies, so many of the early designs used a multistage TRF circuit to tune in one station. The animation below demonstrates how three broad bandpass LC tuners could be adjusted to select one station (green) or another station (blue).
Many 1920s radios such as Atwater Kents come to mind as multistage TRF radios because they had several tuning knobs in a row but often required a diligent radio operator to tune in anything. Later TRF radios had all the tuning knobs ganged together but each stage was preset to be slightly offset from each other to tune in only one station. Below is a basic block diagram of a two-stage TRF receiver.
In short, the principles behind the TRF circuit explains why it is often associated with poor selectivity.
Superheterodyne Operation
One subtle detail behind the principles of superhet operation is the image band. Earlier it was mentioned that the mixer produces a sum and difference. In mathematical terms, the mixer multiplies the radio frequencies with the local oscillator frequency, but both signals are sinusoidal so the output of the mixer is sinusoidal with two frequencies: one that is the sum and the other the difference of the radio frequency and the local oscillator. For example, if the local oscillator is at 1455KHz and the IF is set to 455KHz, then two stations could be created by the mixer, one at 1000KHz and another at 1910KHz! A simple LC tuner at the front of the mixer can easily eliminate the image band. Note again that the LC tuner will pass through a few radio stations, but it most certainly can eliminate radio stations way out of its range. For the previous example, a LC tuner set at 1000KHz might pass through a range from 800 to 1200KHz, but it definitely filters out 1910KHz. As a result, the IF stage will select 1000KHz out of the small range passed through by the LC tuner.
Below is the basic block diagram of the superheterodyne receiver. Notice the first stage is a RF amplifier and mixer block. This is basically the mixer stage, but in most superheterodynes the circuit also doubles as a RF amplifier. Furthermore, the local oscillator circuit often consists of a tuneable LC circuit. The tuning capacitor for adjusting the oscillator frequency is ganged with the other variable capacitor for the broadband LC filter on the input of the RF tuner stage to allow simultaneous adjustment of both capacitors.
When the superhet was invented, it took some time to take off because the circuit required another tube for the local oscillator and the concept was difficult to explain to radio technicans at the time. Nonetheless, the superhet design was quickly adapted in all commerical radios by the 1930s and is still the common method for tuning in AM stations today because of its superior selectivity.
Demodulator
The most primitive crystal radios could function with just an antenna, a crystal, and the earphones because the construction of the earphone itself was a crude form of low pass filtering with capacitance between the windings of wire. But for improved quality, a low pass filter is used to supress the high frequency noise. By using a capacitor and resistor on the output of the diode, the capacitor charges up to the peaks and discharges slowly between peaks of the RF signal thereby creating an output that resembles the original audio signal that was modulated onto the carrier.
The signal from the detector is either fed into earphones or the audio stage for amplification to drive a speaker. Audio amplifier designs vary widely and will not be discussed here.
Conclusion
Amplitude modulation is one of the earliest methods for transmitting a signal over a distance. Audio frequencies range from 0 to 20KHz but these frequencies do not radiate off metal antennas as well as radio frequencies (RF) in the high KHz range. So a RF oscillator is used to create the carrier frequency that is subsequently amplitude modulated with the audio, hence the name AM. The receiver is responsible for tuning in a station then demodulating the signal simply by extracting the audio from the carrier. Tuning in a station may sound simple but there are several challenges and methods that will be discussed here.
The amplitude of the signal is basically the vertical lengths of a sinusoidal and the amplitude can be changed by modulating the audio onto the carrier over time. The figure below demonstrates this concept.
The Tuned Radio Frequency (TRF) radio receiver is one of the oldest receiver designs that was popular in the 1910s to 20s. The basic idea was to simply use a LC tuner to select a station, but there are many problems with this approach. At the time, triodes did not have very good amplification gain so several amplification stages were needed. Second, as mentioned earlier the LC circuit in practice is generally a broad bandpass filter and will pass through a few adjacent stations as demonstrated in the animation below. However, many early radios at the time were able to use the simple LC tuner when there were not so many radio stations adjacent to each other so all they had to tune in was the desired frequency, hence the name Tuned Radio Frequency.
A fine invention by Edwin Armstrong, the superheterodyne exploited similar principles behind demodulation of a signal. As mentioned earlier, demodulation will result in a signal centered at 0 and 2 times the carrier frequency, but rather than demodulating the signal, it is mixed with a different frequency. This article will not divulge how mixers work, but the basic idea is that the mixer produces frequencies that are a sum and difference of the RF station and the other frequency. To keep the explanation simple, the superhet uses a local oscillator for tuning rather than a LC tuner. Precise adjustment of an oscillator is far more practical than precise adjustment of a LC tuner or several tuners. The superhet uses the difference between a local oscillator frequency and the radio station frequency to create an intermediate frequency (IF). The IF frequency mathematically is the frequency of the local oscillator subtracted by the radio station frequency. Also it is easier to create a fixed narrow bandpass filter rather than an adjustable one. The IF is a fixed narrow bandpass filter that is often centered at 455KHz. If a radio with 455KHz IF is playing a station at 1000KHz, then its local oscillator is operating at 1455KHz. Similarly, if the radio is playing a station at 1200KHz then its local oscillator is at 1655KHz. The following animation demonstrates the concept behind the superhet.
The earliest radio receiver was simply made with an antenna, a detector and earphones. The detector serves the purpose of extracting the audio from the modulated carrier and often does so by allowing current only in one direction. The diode is the simplest device that allows current in one direction. However, only allowing the positive edges of the modulated signal to pass through to the audio stage will produce high frequency noises and can be visualized in the following figure.
Hopefully this article provides a basic timeline of the evolution of radio receivers and insight on various receiving strategies. Note that the TRF and superhet sections only explained the RF tuner end of a radio and how a specific radio station was tuned in from the multitude of stations in the airwaves. Once the hard part of tuning in a radio station is done, the rest of the radio is simple. Demodulation is often done with a diode detector and the final stage is amplified with some sort of audio amplifier. The overall block diagram of most AM radio receivers is detailed below.
Back to Top