In the early stages of breakdown of a gas individual avalanches of charged particles are formed. Measurements on avalanche currents with the electrical method can yield new information on discharge processes if the measuring system has a high time resolution (1.3 ns in our case) and if the avalanche is initiated at the cathode by means of a very short (0.6 ns) light pulse. Avalanches in atmospheric air show an aftercurrent caused by delayed electrons which result from detachment. Measurements at various humidities, an extensive analysis and comparison with computer calculations give, for E /p = 35 V /cm Torr, the following result. The negative oxygen ions formed lose their electrons gradually with a timeconstant of 83 ns. Collisions however, cause conversion to a nondetaching type of ion. This conversion process has a characteristic time of 32 ns in dry air and is more rapid when water molecules are present, down to 3 ns at a water vapor pressure of 11.25 Torr. The occurrence of these processes, where water molecules cause a reduction of the number offree electrons, provides a new explanation for the increase of the breakdown voltage of air with humidity. The measurements also yield values of the true (a -17), the ionization minus the attachment coefficient of air. This parameter leads to the number of electrons in the head of the avalanche which, used in the streamer criterion, gives good agreement with experimental breakdown voltages. Older (a -17) values included the effect of detachment and gave no agreement with the streamer criterion.PACS numbers: 52.80.Dy, 51.50. + v 3286

In this article we discuss a 10 kW high-voltage pulse generator for producing corona-induced plasma. The generator mainly consists of a three-step resonant charging circuit switched by thyristors, a transmission line transformer, and a triggered spark-gap switch. Voltage pulses of 30-100 kV with a rise time of about 20 ns, a pulse duration of 50-250 ns, pulse repetition rate of 1-900 pulses per second, energy of up to 12 J/pulse, and an average power of up to 10 kW have been achieved with a total energy transfer efficiency of about 80%-90%. At each frequency, the deviation of the energy per pulse is around 1.0%. Moreover, the generator has been tested for more than 100 h for both industrial demonstrations and laboratory investigations at an average output power of 1-10 kW.

Two pilot pulsed power sources were developed for fundamental investigations and industrial demonstrations of treatment of conducting liquids. The developed heavy-duty power sources have an output voltage of 100 kV (rise time 10 ns, pulse duration 150 ns, pulse repetition rate maximum 1000 pps). A pulse energy of 0.5-3 J/pulse and an average pulse power of 1.5 kW have been achieved with an efficiency of about 80%. In addition, adequate electromagnetic compatibility is achieved between the high-voltage pulse sources and the surrounding equipment. Various applications, such as the use of pulsed electric fields (PEF's) or pulsed corona discharges for inactivation of microorganisms in liquids or air, have been tested in the laboratory. For PEF treatment, homogeneous electric fields in the liquid of up to 70 kV/cm at a pulse repetition rate of 10-400 pps could be achieved. The inactivation is found to be 85 kJ/L per log reduction for Pseudomonas fluorescens and 500 kJ/L per log reduction for spores of Bacillus cereus. Corona directly applied to the liquid is found to be more efficient than PEF. With direct corona we achieve 25 kJ/L per log reduction for both Gram positive and Gram negative bacteria. For air disinfection using our corona pulse source, the measured efficiencies are excellent: 2 J/L per log reduction. Index Terms-Cold pasteurization, inactivation of microorganisms, membrane breakdown, pulsed corona, pulsed electric fields (PEF's), pulsed power.

The electrical method is often used to study avalanches in gases. In this paper a substantial improvement of the time resolution of the measuring system is reported. To release the primary electrons within a very short time a transversely excited atmospheric pressure N 2 laser with a pulse width of 0.6 ns is used. Secondly the high-frequency response of the entire circuit is analyzed. As one of the results of this analysis a subdivided cathode is used consisting of a central measuring electrode surrounded by a grounded electrode. The diameter of the circular measuring electrode determines a maximum gap distance for reliable measurements, which can be calculated from a theory of Shockley and Ramo. As an example of the high-frequency response of our experimental setup avalanches in nitrogen are shown and analyzed. Avalanche measurements in air clearly show a dependence on the humidity, presumably caused by rapid detachment and conversion processes.

No abstract