A transimpedance-amplifier-based subtraction principle for optimum signal resolution in mixed-signal current sensor systems

Generally, there are two strategies to obtain a signal difference of two current sources: analog or digital subtraction. Digital subtraction limits the final resolution of the difference. Analog subtraction yields limitations in gain, range and sensitivity, respectively and may suffer from imperfections of the analog subtraction circuitry (e.g. matching, non-linearity, etc.). In this article, an approach is explained and demonstrated to maximize signal range, sensitivity and final resolution for the difference of two or more (sensor) input signals by utilizing integrating amplifiers with differential outputs. The correlated double-sampling concept is extended therefore. Signal properties and system constraints are explained. The applicability is demonstrated by a 0.6μm-CMOS implementation example for the subtraction of two photo-current input signals within a single transimpedance amplification stage.