Performance Variance of Low Noise Resonant Capacitance Bridges While Replacing their Ungapped MnZn Ferrite Cores

TL;DR

This study evaluates the performance consistency of low noise resonant capacitance bridges using ungapped MnZn ferrite cores across different temperatures, demonstrating minimal variance and reliable noise characteristics for high-precision measurements.

Contribution

It provides detailed measurements of noise performance and variance of ferrite cores at room and cryogenic temperatures, highlighting their suitability for precision AC resonant capacitance bridges.

Findings

Variance of <0.3% at 293 K among cores

Noise minima around 0.3 aF/rtHz at 85 kHz and 0.28 aF/rtHz scaled to 100 kHz at 293 K

Consistent noise performance within 9-6% of previous measurements across temperatures

Abstract

Precision AC resonant capacitance bridges, with a planar printed circuit board transformer using an ungapped MnZn ferrite core, have shown excellent noise performance in high-precision measurements. As many applications use an ensemble of bridges, consistency of performance is critical to the functionality of the systems. If part of the same manufacturing batch and scaled to the same frequency, the noise performance of the ferrite cores at 293 K has a weighted mean variance of <0.3%, with all the cores within a +/- 1% band. At 140 K the weighted mean variance is <0.6% and the 90% inclusion band is +/- 1.5%. Ten cores of SIFERRIT Material N41, from TDK Ferrites Accessories, were tested at room temperature, 293 K, and in a liquid nitrogen dewar at 140 K. Fitted to a parabolic function, to + 3dB on both frequency sides, the weighted mean of the noise minima at 293 K was 0.3027 aF/rtHz, with a variance of 0.0012 aF/rtHz spanning a frequency range of 85.0 +/- 2.7 kHz. Scaled to 100 kHz, the weighted mean and variance were 0.2788 aF/rtHz and 0.0008 aF/rtHz. Corresponding noise values at 140 K were 0.1815 aF/rtHz with a variance of 0.0016 aF/rtHz for the range of 152.7 +/- 6.3 kHz, and 0.1835 aF/rtHz with a variance of 0.0011 aF/rtHz when scaled to 150 kHz. Scaled for resonant frequency and temperature these results are consistent with a previous measurement of another ungapped core (same supplier and type, different manufacturing batch) to within 9% and 6% at 293 K and 140 K respectively.