The influence of irradiation on the measurement of impulse voltages with sphere-gaps

Abstract

The impulse breakdown of short gaps between spheres is greatly affected by the amount of irradiation present, and recommendations are made in the British Standard “Rules for the Measurement of Voltage with Sphere-Gaps”1 that appropriate precautions should be taken when voltages of less than 50 kV (peak) are being measured. These precautions are generally considered to be unnecessary in the measurement of higher voltages, and, in most laboratories, no deliberate steps are taken to provide irradiation.In a previous publication the author has shown that appreciable errors may be incurred if irradiation is not used with spheres of 6.25 cm diameter in the measurement of voltages up to 100 kV (peak). Further measurements have now been made up to 400 kV (peak) with 1/5 and 1/50 impulse voltage waves, of positive and negative polarities, for spheres of 6.25 cm, 12.5 cm and 25 cm diameter. The present results confirm and extend the previous observations.The irradiation provided by the ultra-violet illumination from the impulse generator spark-gaps has a marked influence on the behaviour of the sphere-gap, in the absence of other forms of irradiation, and consequently the position of the sphere-gap relative to the impulse generator is observed to affect the results. This factor does not appear to have been considered in other laboratories, and it may explain some of the differences between the results obtained by various investigators in the calibration of the sphere-gap with impulse voltages.As it is not always convenient to illuminate the gap with the light from the impulse generator spark-gaps, the irradiation provided by the insertion of 0.5 mg of radium in the high-voltage sphere has been examined and is found to give satisfactory results. Measurements of the impulse breakdown of the unirradiated sphere-gap show that there is a gradual transition from the voltage which causes breakdown for 10% of the applied impulses to that which causes 90% breakdown, whereas a sharply defined breakdown voltage is obtained when the gap is irradiated. The mean breakdown voltage of the unirradiated gap exceeds that of the irradiated gap by an amount which varies with the gap length. For a 3 cm gap between 12.5 cm diameter spheres, breakdown of the unirradiated gap takes place at a mean voltage about 13% in excess of that for the irradiated gap. With larger gaps, for which a difference between the breakdown voltages for positive and negative impulses is recorded when one sphere is earthed, the difference is found to be largely influenced by irradiation. The mean breakdown voltage for an unirradiated gap of 9 cm between 12.5 cm diameter spheres is 20% higher than the irradiated value for a positive impulse, and 1.5% higher for a negative impulse. The negative impulse breakdown voltages agree closely with the corresponding values given in the B.S. 358 Tables, whereas the voltages given in the Tables for the positive impulse breakdown are intermediate between those obtained in the present experiments for the unirradiated and the irradiated gaps.The results show that some revision of the Standard Rules for the use of sphere-gaps is necessary, and, in particular, that more definite recommendations should be made concerning irradiation. Further, it is evident that the influence of irradiation is an important factor to be considered in the breakdown of other forms of discharge gaps in which corona is not observed to precede breakdown. Also, the breakdown of discharge gaps on the wavefront may be expected to be governed largely by the amount of irradiation present, and, though his has been realized by many workers for short gaps, no consideration appears to have been given to irradiation in the measurement of wavefront breakdown for longer gaps.