Also found in: Dictionary, Thesaurus, Medical.
A device for recovering information from an amplitude-modulated (AM) electrical signal. Such a signal is received, usually at radio frequency, with information impressed in one of several forms. The carrier signal may be modulated by the information signal as double-sideband (DSB) suppressed-carrier (DSSC or DSBSC), double-sideband transmitted-carrier (DSTC or DSBTC), single-sideband suppressed-carrier (SSBSC or SSB), vestigial sideband (VSB), or quadrature-amplitude modulated (QAM).
The field of amplitude-modulation detector requirements splits by application, complexity, and cost into the two categories of synchronous and asynchronous detection. Analog implementation of nonlinear asynchronous detection, which is typically carried out with a diode circuit, is favored for consumer applications, AM-broadcast radio receivers, minimum-cost products, and less critical performance requirements. Synchronous detectors, in which the received signal is multiplied by a replica of the carrier signal, are implemented directly according to their mathematics and block diagrams, and the same general detector satisfies the detection requirements of all SSB, DSB, and VSB signals. Although synchronous detectors may operate in the analog domain by using integrated circuits, more commonly digital circuits are used because of cost, performance, reliability, and power advantages.
In synchronous detection there are two conceptual approaches: to reverse the modulation process (which is rather difficult), or to remodulate the signal from the passband (at or near the transmitter's carrier frequency) to the baseband (centered at dc or zero frequency). The remodulation approach is routine. Unfortunately, a nearly exact replica of the transmitter's carrier signal is needed at the receiver in order to synchronously demodulate a transmitted signal. Synchronous means that the carrier-signal reference in the receiver has the same frequency and phase as the carrier signal in the transmitter. There are three means available to obtain a carrier-signal reference. First, the carrier signal may actually be available via a second channel. There are no difficulties with this method because a perfect replica is in hand. Second, the carrier signal may be transmitted with the modulated signal. It then must be recovered by a circuit known as a phase-lock loop with potential phase and frequency errors. Third, the carrier signal may be synthesized by a local oscillator at the receiver, with great potential for errors. Unless otherwise stated, it will be assumed that a perfect replica of the carrier signal is available. See Oscillator, Phase-locked loops
Asynchronous detection applies to DSTC signals whose modulation index is less than 1, and is quite simple. First the received signal is full-wave rectified, then the result is low-pass filtered to eliminate the carrier frequency and its products, and finally the average value is removed (by dc blocking, that is, ac coupling). This result is identical to that which is obtained by synchronous detection with a reference that has been amplitude distorted to a square wave. The cheaper but less efficient half-wave rectifier can also be used, in which case the demodulation process is called envelope detection.
A diode detector circuit in a radio receiver has three stages (see illustration). The first stage is the signal source, which consists of a pair of tuned circuits that represent the last intermediate-frequency (i.f.) transformer which couples the signal energy out of the i.f.-amplifier stage into the detector. The second stage is a diode rectifier, which may be either full wave or half wave. Finally, the signal is passed through the third stage, a filter, to smooth high-frequency noise artifacts and remove the average value. See Diode, Intermediate-frequency amplifier, Rectifier
The waveform shaping at the input point of the filter (see illustration) is determined by a capacitor C1 in parallel with an equivalent resistance of R1, the latter in series with a parallel combination of resistors, R2 and R4. The filter also has a capacitor C3 between R2 and R4, and a parallel combination of a resistor R3 and a capacitor C2 in series with R2. The reactances of both capacitors C2 and C3 are quite small at the information frequency, so the capacitors can be viewed as short circuits. Meanwhile, the C3-R4 combination serves as a dc-blocking circuit to eliminate the constant or dc component at the output point of the filter. In order to bias the output point properly for the next amplifier stage, R4 is replaced by a biasing resistor network in a real filter; R4 is the single-resistor equivalent.
The strength of the signal arriving at the detector is proportional to the mean value of the signal at the input point of the filter and is sensed as the automatic-gain-control (AGC) voltage. Changes in this voltage level are used to adjust the amplification before the detector so that the signal strength at the detector input can remain relatively constant, although the signal strength at the receiver's antenna may fluctuate. Since capacitor C2 shunts all signal energy and any surviving carrier energy to ground, the AGC voltage is roughly the average value at the input point of the filter scaled by R3/(R1 + R2 + R3) because R1 is much smaller than R4.
Additional filtering can be provided as necessary to reduce the noise to an acceptable level. This amplified and filtered signal is finally delivered to an output device, such as a loudspeaker. The ragged waveform of the filter output contrasts with the smooth waveform of the information signal. The raggedness vanishes as the ratio of the i.f. frequency to the information frequency increases. While a synthetic example with a low ratio can be used to clearly show the effects within the demodulator, the amplitude of this raggedness noise decreases in almost direct proportion to the increase in the frequency ratios. The raggedness of the output signal from the filter after half-wave rectification is much greater than that of the full-wave-rectifier case.
The actual worst-case ratio for standard-broadcast amplitude-modulation radio is 46.5:1. In this case the raggedness on the filtered outputs is reduced to a fuzz that can be seen in graphs of the waveforms but is well outside the frequency range of audio circuits, loudspeakers, and human hearing. See Amplitude modulator