Optimizing the resolution of resistance bridges

This paper identifies the influence variables affecting the resolution of thermometry resistance bridges. The theoretical limit to the resolution is determined by the spectral density of all noise sources, including the thermal noise from the resistors and amplifiers, quantization noise, and the equivalent noise bandwidth. Optimization of resolution is equivalent to minimization of measurement time, which has a Heisenberg-uncertainty-like relationship with equivalent noise bandwidth. Bridge design has a major influence on the compromises between the different noise sources, as well as influencing compromises with respect to measurement time. The paper explains the concept of an ideal averaging filter and illustrates the use of the Allan variance to evaluate the dependence of the resolution on measurement time.