The power supply interference rejection of four commercial quasi-digital sensors (ADXL202, SMT160-30, TMP04, MAX6676) has been experimentally analyzed by using a test method that allows us to easily select the amplitude and frequency of the interfering signal. Slow and fast changes of the sensor supply voltage yield bias errors and increase the uncertainty of the output signal parameter (time interval, mark-space ratio) that carries the information about the measured physical quantity. The effects of the interference on the output information are unique when the interfering frequency equals a multiple of the output signal frequency. For the sensors analyzed, a 100 mV sine-wave interference increased the uncertainty from 14 to 140 times that obtained when the supply voltage was clean. These experimental results cannot be directly predicted from the power supply rejection ratio value and the performance figures in manufacturer's data sheets.