ASIC System Design Considerations for Under-glass Fingerprint Sensors An investigation into system performance using capacitive technology

Abstract

Placing a fingerprint sensor under thick glass puts heavy performance requirements on the mixed signal path in the ASIC that constitutes the sensor. This thesis will explore and explain different effects that can be observed when the distance between a finger and a capacitive sensor is increased. We will motivate why these effects are detrimental to performance and try to combat them with proposed system-level design changes. Fast-capture of fingerprint images have been used to characterize contrast over time; the time behavior of contrast could not be used to circumvent detrimental effects of using thick glass. Instead, focus has been placed on reducing noise using extensive multisampling, using a dynamic sample count based on finger movement, with positive results. A reduction in pixel noise power of approximately 50% compared to a currently used sensor design has been achieved within the same sampling time frame.