A thermistor bridge datalogger: an amplifier of gain and input-referred noise density…

PHYSICS · Multi-part · Fermi University Physics · Electronics, Instrumentation & Communication · Feedback, operational amplifiers and analogue signal processing · Digital electronics, embedded systems and data acquisition

Electronics for Measurement, Ch Sampling, quantisation and the acquisition chain

Problem

A thermistor bridge datalogger: an amplifier of gain 200200 and input-referred noise density 10.0 nV/Hz10.0\,\mathrm{nV}/\sqrt{\mathrm{Hz}} feeds an anti-alias filter of noise-equivalent bandwidth 20.0 Hz20.0\,\mathrm{Hz} (matched to the 00 to 20.0 Hz20.0\,\mathrm{Hz} signal band), then a 14-bit converter on a 3.30 V3.30\,\mathrm{V} reference started by a timer with 30.0 ns30.0\,\mathrm{ns} rms of aperture jitter, evaluated at the top of the band. The reference drifts 25.025.0 parts per million per kelvin, and the logger's enclosure is expected to swing 15.0 K15.0\,\mathrm{K} over a day. ((a)) Build the random-error budget at the converter input (quantisation, amplifier, jitter), and find the total rms error, the signal-to-noise ratio for a full-scale sine and the effective number of bits. ((b)) Find the systematic error the reference drift produces at full scale over the stated temperature swing, and compare it with the random total from part (a). A colleague proposes replacing the 14-bit converter with an 18-bit one, keeping everything else unchanged, to improve the logger's accuracy. Decide whether this would help, and name a cheaper fix.

Browse the problem library