By The tasks of a communications receiver to demodulate the transmitted signal begin with selecting the signal within a specific bandwidth at a desired frequency, commonly known as a particular channel. In another article, we discuss specifications for a radio receiver such as dynamic range, noise floor and sensitivity. Today we discuss an architecture used in earlier generations of radios.
To avoid interference from the neighboring channels, the most straightforward approach is to filter out the spectral contents outside this channel and amplify the desired signal in one or more RF amplification stages. This was one of the earliest techniques employed for building radio receivers known as Tuned Radio Frequency (TRF) systems and consists of an adjustable, or tunable, Bandpass Filter (BPF) around the desired frequency. From a signal processing viewpoint, the main idea is to move the filter to the signal.
For example, the signal at frequency $F_{C,2}$ is selected through tuning the bandpass filter to frequency $F_{C,2}$ that removes all other spectral contents, as illustrated in the figure below. If an adjacent channel were to be chosen, the tuning control mechanism consisting of circuit elements like variable capacitors and inductors just altered the center frequency of this filter. The major problem with a TRF is designing tunable bandpass filter with constant bandwidth and sufficient frequency selectivity over the entire tunable range.
Suffering from poor selectivity and low sensitivity, TRF receivers quickly became obsolete within the early years of radio and gave way to superheterodyne and direct conversion receivers. In today’s age of fast digital electronics, there would have been no need to even mention them as a category for Rx architectures. Surprisingly, however, they do have a reincarnation in the realm of digital signal processing where it is straightforward to move the filter to the channel through appropriate conversion of a lowpass filter to a bandpass filter.
