Influence of the test set-up in HV divider calibrations

The authors describe an investigation which showed that any standard test set-up that does not go to extremes will not influence the results of a calibration of approved measuring systems by comparison with a reference measuring system, regardless of the type of divider being calibrated. The peak value standard deviation of all 900 impulses used in the nine calibrations with different common connection lengths and different angles between the divider connections was 0.16% for the damped capacitive divider C and 0.13% for the resistive divider O. On the other hand, care should be taken that interference is kept to a minimum and that proper interference tests are made before a calibration to ensure this