Magnetic Pickups and Proximity Probes

It's got something to do with it being the only way of getting a flow signal into the electronics, Uncle, only I don't know how it does it.

Not 100% sure, but I think that in the same way indicators and trip instrumentation are often kept independant as a means of fail safe. Both perform the same function, however one takes the place of mechanical overspeed protection and the other is indicative of rotor speed. This was the case on the GE turbines I used to work on. Often the speed control software, requires multiple inputs to calculate deviation, an excess of which will cause trips to initiate. The speed pickups generally are positioned directly above a toothed flywheel at the point where the load couples to the turbine shaft and is expressed as a frequency, where as the keyphaser has an on/off pulse that co-encides with the frequency present. It is these two signal entities that are compared and deviation deduced to monitor speed control systems health.

A keyphasor gives a pulse at a fixed position in the rotation of a shaft. As its name it enables the phase of shaft or casing position changes to be related in phase to a specific angle in the rotation. This pulse also gives a clean signal for speed signal generation.

Magnetic pickups also give in principal a timing signal. However, in practise, they and the gear type teeth with which they work give a sinusoidal signal (at many times the shaft speed) from which one cannot tell the actual angle in a shaft turn [e.g. 50 teeth around a shaft, all identical, give no definite position signal]. Amplitude is proportional to shaft speed, in principal.

In contrast, a keyphasor works at a single notch or projection around machine circumference and its output signal, dependent on clearance, does not change its output amplitude if the shaft turns very slowly