Resistance-Based Scaling of Noise Temperatures From 1 kHz to 1 MHz

A conventional technique for scaling temperatures in Johnson noise thermometry is via resistance ratios. We describe measurements from 1 kHz to 1 MHz using this approach via correlation methods in the frequency domain. We show that the effects of the mismatch in time constants of the input networks may be empirically treated to extract the noise power and temperature ratios in the low-frequency limit with uncertainties < 60 muK/K