S.M. Turnbull

The operation of repetitive high-pressure spark gap switches

Two parallel experiments have been undertaken to investigate the mechanisms which govern the recovery of spark gap switches in repetitive applications. Measurement of the neutral gas density is accomplished using a laser Schlieren technique which enables spatial and temporal cooling of the gas to be observed. The rate of rise of voltage recovery is recorded using a double-pulse modulator system which determines the breakdown voltage of the gas at specific time intervals after breakdown by the first pulse. A continuous sweep voltage is applied to the switch in order to establish the influence of ions on the recovery characteristics. Results are presented for SF6/bin and air which show that voltage recovery is not solely governed by recovery of neutral gas density. There is a significant influence from the residual ion population created by the previous discharge. This influence is dominant for several hundred milliseconds after breakdown. The application of a suitable sweep voltage effectively minimizes the ion population, which results in significantly improved voltage recovery characteristics.

The influence of polarity on trigatron switching performance

This paper reports on a study into the effect of trigger voltage polarity and main gap voltage polarity on the switching performance of a pulse charged trigatron. The four polarity combinations possible, two homopolarity (trigger voltage and main gap voltage of the same polarity) and two heteropolarity (trigger voltage and main gap voltage of opposite polarity), were studied and compared in terms of their effect upon the switching range, delay time to breakdown, jitter, and voltage collapse time. It was found that the two heteropolarity configurations were superior to the homopolarity configurations in terms of the above switching characteristics, with the positive trigger/negative main gap heteropolarity configuration performing the best. The results are discussed, and an explanation of the influence of the polarity configuration on switching is suggested in terms of the discharge initiation mechanisms and subsequent development.

The design and operation of a compact high-voltage, high pulse repetition frequency trigger generator

A high-voltage trigger generator has been developed to trigger an in-line plasma closing transfer switch. The trigger generator is capable of producing 70 kV voltage pulses, having a rise time and duration of 10 and 60 ns respectively. It uses a stacked Blumlein cable generator fired using a triggered corona stabilized (TCS) switch and is fully controllable in frequency from single shot up to 10 kHz in a burst mode. The device is very compact, portable and is operated via a fibre optic link from a solid state pulse generator. The trigger generator can also be used as a fast high-voltage, high peak power and high-frequency driving device.

Enhanced spark gap switch recovery using nonlinear V/p curves

A method of improving the voltage recovery time of high pressure gas switches is described and demonstrated. By changing the electrode geometry of a normally uniform field switch, it is possible to suitably alter the breakdown voltage/pressure (V/p) characteristic. The desired V/p characteristic for improved voltage recovery is one which exhibits two different rates of rise of breakdown voltage with pressure, such that at high enough pressures, the breakdown voltage becomes almost independent of pressure. Full voltage recovery may then be achieved during the fast dV/dp section of the curve while only partial pressure recovery is achieved. This effect is demonstrated in a switch filled with SF/sub 6/. An improvement in the voltage recovery time of greater than 10 ms to /spl sim/1 ms was observed when the operating pressure was increased from 3-6 bar. >

A quantitative laser schlieren method for measurement of neutral gas density in high-pressure gas switches

A quantitative laser schlieren technique has been developed to measure the changes in neutral gas density inside a spark gap switch after breakdown. The measurement system incorporates an array of narrow laser beams which are used to probe the inter-electrode gas volume at discrete distances from the spark axis. Changes in the refractive index gradient within the switch, caused by variations in neutral gas density, result in beam deflection, which is measured using a neutral density wedge/photodiode combination. The resulting schlieren profiles yield both spatially and temporally resolved refractive index data, which enable calculation of the gas density. The use of four averaged trials to obtain the final neutral gas density profiles minimizes errors caused by the turbulent nature of the cooling gas. A typical set of results is presented for SF6 at a pressure of 0.5 bar.

An investigation of trigatron breakdown by two different mechanisms

An investigation has been carried out into the mechanisms controlling high voltage trigatron operation. This paper describes the experimental work identifying the two different sequences of triggered breakdown. The operating conditions required for each of these breakdown mechanisms to take place are described and discussed. One mechanism (mode 1) is when the trigger pulse is applied to the trigger pin resulting in breakdown across the gas between the pin and earthed electrode. The resulting intense UV radiation is believed to produce free electrons through either photo-ionisation or photo-detachment in the gas which in turn leads to the initiation of an electron avalanche followed by breakdown of the main gap. The second mechanism (mode 2) is when the trigger pulse initiates breakdown between the pin and HV electrode first. After the main gap voltage has collapsed to the trigger pin, this produces a high degree of over-voltage between the pin and the adjacent earthed electrode and consequently, the voltage collapse time is very short. This mode of operation is desirable and can be ensured by careful design of the trigger/earthed electrode configuration. The jitter performance of the trigatron when operating in each of the above modes has been measured at operating voltages of 200 kV for mode 1 and 500 kV for mode 2. The results from this study indicate that the smaller jitter is obtained when the trigatron is operating mode 2, as expected.

Triggered switch performance in SF/sub 6/, air, and an SF/sub 6//air mixture

This paper reports on the work undertaken to determine the performance and operational characteristics of a triggered corona stabilized (TCS) high pulse repetition frequency (PRF) switch, developed as part of a high PRF trigger generator. The voltage/pressure (V/p) and trigger range characteristics of the switch were measured for SF/sub 6/, air, and a 25%/75% SF/sub 6//air mixture up to a pressure of 4 bar (1 bar=100 kPa), with different main and trigger gap spacings. Other parameters such as the magnitude of the trigger pulse and the effect of certain circuit elements on the trigger range characteristics and performance of the TCS switch were also investigated. The results presented are evaluated in order to identify a suitable operating region for the switch with each of the gases used during this study.

Methods of improving the pulse repetition frequency of high pressure gas switches

This paper describes several methods which have been developed to improve the pulse repetition frequency (PRF) of high pressure spark gap switches. Under pulse charged conditions, the voltage recovery process of the spark gap has been shown to be restricted by the residual ion population. This may be minimised by applying a suitable bias voltage across the gap. It is also possible to manipulate the V-p characteristic of a spark gap to improve the rate of rise of recovery voltage by reducing the recovery voltage dependence upon pressure. The combination of these effects has been shown to reduce the voltage recovery time of pulsed charged spark gaps from several hundred ms to several ms. Under DC conditions, it is possible to employ a corona discharge, found in a highly nonuniform field, to stabilise and control the voltage breakdown process. The use of corona stabilisation has enabled the operation of a spark gap at a PRF of more than 5 kHz, without employing gas flow techniques.

Increasing the prf of high pressure spark gap switches

This paper reports on recent investigations into the mechanisms which govern the recovery of spark gap switches for repetitive applications. The study involved the use of two parallel experiments , the first to monitor neutral gas density after switching and the second to measure the rate of rise of voltage recovery. Measurement of the neutral gas density is accomplished using a quantitative laser schlieren technique which enables both spatial and temporal cooling of the gas to be observed. The rate of rise of recovery voltage is recorded using a novel double pulse modulator system which determines the dielectric strength of the gas at specific time intervals after breakdown by the first pulse. Results are presented for SF, and air which show that voltage recovery is not solely governed by the recovery of the neutral gas density. There is a significant influence from the residual ion population created by the previous discharge, which is dominant for several hundred milliseconds after breakdown. The applicati-on of a suitable sweep voltage has been shown to effectively minimise this ion population resulting in significantly improved voltage recovery characteristics. For both gases, the recovery time can be reduced by almost two orders of magnitude.

A PFN Marx generator based on high-voltage transmission lines

A novel, high-voltage Marx generator based on distributed transmission lines is described. The generator is capable of producing a nominally rectangular output voltage pulse of up to 200 kV into an open-circuit load. The voltage-pulse duration and risetime are 100 and 30 ns respectively. The generator is 50 cm in diameter and 50 cm in height, and is comprised of five voltage-multiplying stages with each stage made from 25 m of 50 , 40 kV dc co-axial transmission-line cable. The transmission-line cable for each stage is configured as two parallel pulse-forming lines with a resultant output impedance per stage of 25 . For five stages the output impedance is therefore 125 . The generator construction is described and the performance is evaluated in terms of the risetime, erection time, jitter in the erection time and output-voltage efficiency.