Phase Switching is a system used to increase the discrimination and sensitivity of an interferometer; where an extra half-wave path difference is switched in, at well defined frequency, between the two interfering signal sources. In-phase signals then become out of phase and vice versa, so that the signal output becomes modulated by the switching frequency, and can be more easily filtered from the internally-generated noise. (The discrimination is highest for sources which are small compared with the interferometer-fringe spacing)
In the block diagram a phase switch is used to introduces either 180° or 360° (equivalent to 0° zero degrees) of phase shift into the right-hand transmission line, at a specific frequency easily demodulated by the receiver detector.
This switching signal generator is a square wave which changes periodically from one state to the other, many times per second, so that at one instant the first interference pattern is obtained, and at the next a second pattern. The same switching signal is applied to a phase-sensitive detector, which acts in synchronism with the phase switch and so subtracts the second pattern from the first. The resulting pattern is shown in the third part of Figure 1.; below, each maximum of the first pattern appears as a positive peak, but each maximum of the second pattern appears as a negative peak.
The instrument known as a phase-switching interferometerwas invented by Sir Martin Ryle in 1951, and is one of the major innovations in radio astronomy for which he was awarded the Nobel prize in 1974. This principle and later versions are now very widely used in radio astronomy.
You might ask, why go to so much trouble? Why not just use the simple adding interferometer and eliminate the phase switch and phase-sensitive detector? Although it is true the same information about a cosmic source is available in either type, the phase-switching instrument can be made much less sensitive to variations inherent in the receivers’ electronics. Also, the phase switch causes the cosmic signal to be modulated at the switching frequency before it is amplified in the receiver. The noise generated by the receiver has the same general character as the cosmic noise, and this switch modulation acts as an identification tag helping to distinguish it from the receiver noise.
Also, the phase-switching interferometer responds less to extended sources, such as the general background radiation of the Milky Way, which might otherwise obscure the fainter radiation from weaker and small-diameter sources. In summary phase switching interferometers or more current versions, the correlator interferometer, are vastly superior to a simple adding interferometer and now universally used in preference to it.
“Phase switching is a clever trick how to make a basically additive interferometer behave like a multiplicative one. A plain additive interferometer has some undesired traits, like it outputs the fringes on a big DC pedestal, and needs the channels to be reasonably amplitude balanced to give good results. The phase switching is a way to avoid these.” – Marko Cebokli
Refer : http://www.hardhack.org.au/book/export/html/71
The early “adding” interferometers operated by adding two voltages, then detecting their sum using a square law detector with the result
This has the disadvantage that the autocorrelated signals are often orders of magnitude stronger than the cross-correlated signals. In the 1950’s, Ryle developed the technique of phase switching, in which the phase of one of the voltages is periodically flipped, so that the signal flips back and forth rapidly between and . By taking the difference, the signal becomes , which is the desired cross-correlation only.
Refer : http://scienceworld.wolfram.com/physics/PhaseSwitching.html
Information based on articles from Sky & Telescope by G.W. Swenson & W. Swenson, Jr.