Measurement uncertainties arising from unpowered shipment of dcvoltage references

We have measured the hysteresis in output voltage arising from loss of power for three Fluke 732B dc voltage references. Continuously powering the internal temperature-controlled oven provides the best accuracy with such Zener-diode-based dc voltage references. Batteries are used to provide power continuity when shipping between laboratories. We have shown that for these particular instruments, an apparent hysteresis induced by humidity is likely to dominate any hysteresis arising from loss of battery power when transit between laboratories takes three days or more. Our conclusion is that as much, if not more, care needs to be taken when dealing with humidity than with loss of battery power when shipping between laboratories. When the highest accuracy of dc voltage comparison is required, it becomes necessary to apply voltage corrections using the measured values of humidity coefficients and the relative humidity occurring during transit