Quadrature sampling phase detection

The demodulation of phase encoded high frequency carrier signals is a problem encountered in many applications and areas of research. Numerous electronic methods for demodulation of the carrier, including digital counting techniques and quadrature mixing demodulation have become standard.l Alternatively, the waveforms can be digitized and the phase extracted computationally.Q As the carrier frequency increases however, this approach can place severe demands on digitizer speed and memory. This is especially true when more than a single carrier is to be recorded. We here report a simple sampling quadrature method for phase demodulation that is optimum in the sense that the digitizer needs only sample at twice the information bandwidth rather than at twice the carrier frequency. The technique uses interleaved in-phase and quadrature clock pulses at twice the carrier frequency f=o/2~ (or a subharmonic off) as an external clock for digitization of the phase modulated carrier. This clock is derived from an unmodulated reference frequency signal using a tracking phaselocked loop (PLL) and binary counter. As shown in Sec. II, the resultant digitized signals are proportional to the cosine and sine of the phase modulation q(t) which can then be extracted by the computer. The phase noise is limited by the digitizer discretization error and noise on the carrier. It is insensitive to changes in carrier frequency (determined by the tracking range of the PLL) and can be compensated for clock jitter. As well, the carrier amplitude, weakly timevarying dc offsets, and even harmonic contamination can be recovered and/or removed. Many channels can be recorded using the same external clock, obviating the need for electronic phase comparators. This is a great advantage in multichannel systems where many phase detectors must be fabricated, calibrated, and maintained through the life of the experiment.“T5 In high frequency appplications, it may not be possible or desirable to use a reference clock at 4f. Indeed, the digitization rate needs to be sufficient only to capture the temporal variation in the phase modulation cp (t). As shown in Sec. III, it is straightforward, using binary counters, to produce a clock composed of a subharmonic off and its quadrature partner from the PLL output. A possible drawback of this sampling method is that it may be necessary to digitize a signal of known fixed frequency in order to calibrate the timebase. This is because the clock rate depends on the instantaneous carrier frequency which can be tiine dependent. This is the case for the scanning far-infrared laser interferometer/polarimeter application discussed in Sec. IV. This instrument is being developed for measurement of plasma density in the Hl heliac (helical axis stellarator) at the Australian National University.” Though the sampling rate need be sufficient only to accommodate the information bandwidth the digitizer must be able to track and hold accurately signals at the carrier frequency rate. The feasibility of the method is thus limited by present digitizer technology to the range f 5 10 MHz for 12 bit dynamic range.