A detailed analysis is given of the procedures involved in balancing second order axial gradiometers using superconducting vanes. Building on this analysis a very simple practical procedure emerges that provides a means of quantifying the initial imbalance and correcting it in an iterative manner. The method is illustrated by comparing experimental data with numerical calculations. The effects of test field inhomogeneities and eddy current sources on the balancing procedure and the degree of noise rejection obtained are discussed.
