All of the responses show an initial offset on the order of 0e.1 quanta sec due to the dynamic response of the loop. In fact, in the ideal case, the offsets are identical. The steady, state velocity drifts of 10a and l0b correspond to approximately 0.3 quanta, less than the worst case ± 0.5 quanta. The drift rate unaffected by the scale factor error. In the ideal case, the navigation output has negligible -drift (as compared to -the torque command)while 0.075ipoanta drift -was induced with the -20% scale factor error. We can infer that in the later case a factor of four increase in the system resolution was realized and with calibration significantly more resolution is achievable.