Glitch Compensation for a Digital-to-Analogue Converter

The resolution of precision mechatronic systems is fundamentally limited by the noise and distortion performance of digital-to-analogue converters (DACs). The sources of noise and distortion include quantisation error, non-linearity, thermal and semiconductor noise, and glitches. In precision systems it is desirable to approach the theoretical performance limit in order to maximise dynamic range and bandwidth. Currently, effective methods exist to mitigate DAC error sources, except for glitches, which is the focus of this paper. A glitch compensation technique is introduced based on an open-loop observer that utilises knowledge about glitches while quantisation error mitigation remains intact. The compensation technique attempts to minimise the glitch effects present on the analogue output via additional control effort, combined with Delta Sigma-modulation which is a method for mitigating quantisation error. The method is applied to a DAC model including the glitch effects and show significant improvements over earlier efforts.