Digital balanced detection for fast optical computerized tomography

Analogue Balanced Photo-detection has found extensive usage in high-sensitivity small signal applications e.g. coherent heterodyne detection. It is particularly effective for laser intensity noise removal. Nevertheless, the high cost of the commercially available analogue systems makes them unsuitable for multi-channel applications, such as fast tomography. In this paper a flexible, scalable, inexpensive and compact solution for multi channel digital balanced detection is presented. The proposed system has two components: an analogue front-end, comprising a differential photodiode amplifier for minimizing the external interference noise, and a digital balanced noise remover. The latter component initially calculates a balancing factor (BF) from the average power ratio of the signal and reference photocurrents, measured with the object removed from the signal path. Three digital balancing algorithms (DBAx) are considered for subsequent processing. In DBA 1, BF is directly used in real-time ratiometric calculations. In DBA2, the BF is adjusted in real time by monitoring the window-averaged power of the received photocurrents. In DBAS, first the baseline is removed using differentiation and then ratiometric detection is performed. Using the digital alternative only one measurement of the reference beam is necessary for single-source, multi-channel detection systems. The data from multiple channels are processed in parallel by pipelined hardware, configured as a state machine. The proposed system leads to a fast optical computerized tomography system using digital balanced detection.