In a traditional superheterodyne radio receiver, most oscillator drift and instability occurs in the first frequency converter stage, because it is tunable and operating at a high frequency. In theory, if one can eliminate this drift, the receiver will be stable.
Principles of operation
The Wadley Loop works by:
- combining the first oscillator with the received signal in a frequency mixer to translate it to an intermediate frequency that is above the receiver's tuning range,
- mixing the same oscillator with a comb of harmonics from a crystal oscillator,
- selecting one of the results of (2) with a band-pass filter, and
- mixing this with the IF signal from (1).
Since the high-IF of part 1 drifts in the same direction, and the same amount, as the "synthetic oscillator" of part 3, when we mix them in part 4 the drift terms cancel out and the result is a crystal-stable signal at a second intermediate frequency.
But the drift makes it impossible to use high-IF selectivity to reject undesired signals. Instead, the high IF is designed with a bandpass characteristic. Also, since the first oscillator is cancelled out, it cannot be used to tune a particular signal. Instead, it selects an entire band of signals - which one depends on which harmonic was chosen in part 3 above. The size of the band is equal to the spacing of the crystal harmonics. A conventionally tuned "back end" selects the desired signal from the band of signals presented at the second IF.
Let's say we want to pick up signals from 0 to 30 MHz. We'll divide this into 30 1 MHz bands, and translate them to a band at 44-45 MHz. To convert 0-1 MHz, the first oscillator must be 45 MHz, to convert 1-2 MHz it must be 46 MHz, and so on. Meanwhile, we also mix the first oscillator with harmonics from a 1 MHz crystal and put the result through a 42 MHz filter. Only one harmonic gets through. When the first oscillator is 45 MHz, it is the third harmonic, because 45 - 3 = 42. At 46 MHz, it's the fourth harmonic, and so on. The oscillator does not have to be exactly 45, 46, and so on, only close enough to get through the 42 MHz bandpass filter. Let's say it is 45.1 . Then we get 42.1 from the filter, and 45.1 - 42.1 is still 3. When we mix the high IF with the 42 MHz, we get a band of signals from 3 MHz to 2 MHz, from which we select the desired signal, perhaps with a conventional superheterodyne back-end converting 3-2 MHz to 455 kHz and finally demodulating the signal back to audio. The overall receiver drift consists of the crystal's drift plus the 3 MHz back-end, so when we're listening to a 30 MHz signal, this receiver is about ten times as stable as one using a high-frequency tunable VFO.
To a new user, the feel of the first oscillator tuning control is counterintuitive. Although the knob moves in a continuous, analog fashion, its effect on receiver operation is discrete, that is, the tuning advances in 1 MHz jumps.
An optical implementation of a Wadley Loop has recently been proposed. This allows a compact relatively unstable laser to be used as a local oscillator, with the system stability being obtained from a master 'comb source' (usually a pulsed laser, such as a mode-locked laser), possibly common to many receivers within an exchange.