Günter Lins

Measurement of the tungsten ion concentration after forced extinction of a vacuum arc

The concentrations of singly ionized and neutral tungsten atoms were measured by laser-induced fluorescence after the forced extinction of vacuum arcs between tungsten-copper butt contacts, 28-mm in diam. and 10-mm apart. The 50-Hz current was forced to zero at its maximum of 200 A in 1.3 mu s by application of a reverse voltage. Near current zero, the ion concentration of 4*10/sup 17/ m/sup -3/ is of the same order of magnitude as the atomic tungsten concentration, which is 6*10/sup 17/ m/sup -3/. While the concentration of the neutrals remains virtually constant during 20 mu s after current zero, the ion concentration decays by three orders of magnitude in the same time. The decay-time constant varies from 1.9 mu s close to the postarc cathode to 3.6 mu s near the postarc anode. It is concluded that the dielectric recovery of vacuum gaps after diffuse arcs is mainly controlled by residual charge carriers. >

Neutral copper vapor density and electric recovery after forced extinction of vacuum arcs

Dielectric recovery data were obtained for vacuum arcs between chromium copper butt contacts 30 mm in diameter and 2 mm apart. The 50-Hz arc current was forced to zero at its maximum of 200 A in about 1 mu s. Following current zero, high-voltage pulses of a sufficient amplitude to always cause breakdown were applied to the gap. Gap recovery is characterized by the measured breakdown voltage as a function of time. Dielectric strength of the gap rises sharply within the first few microseconds after current zero, reaching its final value in about 10 mu s. Neutral copper concentration in the center of the gap was measured by laser-induced fluorescence under conditions very similar to those of the recovery measurements. In contrast to the fast gap recovery, the copper vapor concentration does not change substantially during the first 100 mu s from its value of 1.4*10/sup 18/ m/sup -3/ near current zero. It is concluded that the neutral copper vapor concentration does not play a decisive role in gap recovery under these experimental conditions. This is corroborated by the fact that the mean free path for electron-impact ionization of copper atoms exceeds the gap length by four orders of magnitude. >

Influence of electrode separation on ion density in the vacuum arc

The density of singly ionized chromium shortly before and after forced extinction of vacuum arcs between chromium-copper electrodes was measured by laser-induced fluorescence for 2- and 10-mm contact gaps and currents between 200 A and 1 kA. In all cases studied, the ion density was constant before ramping down to the current and decayed exponentially after current zero. The ion density at current zero was found to be lower and to decay faster for a short gap than for a longer one, clearly indicating the effect of the contact separation on the charge carrier density. The variation of the time constant for the ion density decay with contact separation is closely analogous to the influence of contact separation on the recovery time of a switch gap. Furthermore, the recovery of dielectric strength of a chromium-copper gap proceeds on the same timescale as the decay of the density of singly ionized chromium. Both of these findings confirm that the ion density has a strong impact on the recovery of a vacuum gap. >

Collisional transfer and neutral copper vapour density during a diffuse vacuum arc

Measurements of the number density of neutral metal vapour during vacuum arcs by laser-induced fluorescence (LIF) are disturbed by electronic collisions which depopulate the upper fluorescence level. To overcome this difficulty, the sinusoidal arc current was rapidly forced to zero which resulted in a rapid reduction of the collisional interaction. This is shown by measurement of the collisional transfer from the optically pumped 42P3/2 level to the neighbouring 42P1/2 level of copper. Measurements of the laser-induced fluorescence were performed shortly before and after the artificial current zero. Due to the reduction of the collisional interaction, the LIF signal after current zero rises by an order of magnitude above its value before the artificial current zero, reaches a maximum two microseconds after current zero and decays as the vapour disperses. The vapour density inferred from the maximum LIF signal is considered as an approximate value for the density before the artificial current zero. By forcing current zeros at preselected times during a 50 Hz sinusoidal half-cycle with 200 A peak current, a lower limit for the vapour density during arcing was obtained. The copper vapour density rises to a maximum value of 3.6*1018 m-3 near current crest and drops to 5*1016 m-3 at the natural current zero of the half-cycle. Afterwards it decays by about two orders of magnitude within 1 ms.

Metal vapour densities in pseudospark switches with tantalum carbide cathodes

In order to combine high electron emission with low electrode erosion, the electrodes of high-current pseudospark switches should be made of materials with low work function and high melting point, such as the carbides of certain metals. An experimental pseudospark switch was successively fitted with cathodes of pure molybdenum and molybdenum with additions of 10 and 30 per cent of tantalum carbide, respectively. The number densities of molybdenum and tantalum vapours were measured by laser-induced fluorescence, immediately after the switch had been subjected to current pulses of 12 kA in amplitude and 1600 ns duration. For a pure molybdenum cathode the maximum molybdenum vapour concentration amounted to about 1018 m-3. It did not change substantially when 10% of TaC were added to the cathode material, but it increased by a factor of two when the TaC content was raised to 30%. The number density of tantalum vapour was twice as high for 10% of TaC in the electrode material as for 30%. It is concluded that electron emission occurs preferably in the high-melting TaC regions. With an increase in the concentration of TaC in the bulk material the total current is distributed over a greater number of emission sites such that the thermal load on an individual emitter decreases and less material is evaporated.

Measurement of the rotational temperature of oxygen in a high-power inductively coupled plasma

In a high-power (30 kW), inductively coupled argon-oxygen plasma used as a tool to deposit high-temperature superconducting YBa2Cu3O7-x films rotational bands of the Schumann-Runge system B3 Sigma (u-)-X3 Sigma (g-) oxygen were detected in the spectral regions around 340 and 360 nm. After derivation of the line strength factors for the triplet system, rotational temperatures can be determined by fitting calculated relative intensities to those observed experimentally, and the contributions of R and P branches to the lines observed can be identified. The rotational temperature 7 at a distance of 34.5 cm from the exit of the plasma torch where the deposition of superconducting films takes place lies in the range 1850-2050 K depending on the RF power input which was varied in the range 8-27 kW. Near the torch exit, temperatures in the range 2600-3400 K are observed. They depend strongly on gas composition and gas flow rate, in contrast to positions far away from the torch exit. Excitation temperatures near the torch exit which were derived from Ar lines are considerably higher than the rotational temperatures, indicating that the latter are typical of the colder outer parts of the plasma.

Measurement of the temperature of evaporating macroparticles after current zero of vacuum arcs

The concentration of copper and chromium vapor after current zero of vacuum arcs between chromium-copper contacts was measured simultaneously by the atomic-absorption technique. Temperatures between 1900 K and 2200 K were derived from the absorption data, assuming that the ratio of concentrations of copper and chromium vapor is equal to the ratio of vapor pressures at a given temperature. The temperatures obtained in this way were assigned to droplets of molten contact material evaporating in flight. >

The densities of neutral and ionized metal vapour in a recovering pseudospark switch

The number densities of tungsten atoms (W I) and singly charged tungsten ions (W II) in a recovering pseudospark switch were measured by laser-induced fluorescence after current pulses with amplitudes 6-17 kA and 1.6 mu s duration. Atomic and ionic energy levels up to 0.77 eV were taken into account to obtain approximate values for the total number densities. At the end of current flow the density of both neutral species and ions is of the order of 1018 m-3. Within 100 mu s the densities of W I and W II drop by three and five orders of magnitude, respectively. Excitation temperatures of W I exceed 4000 K initially, and have dropped to 2000 K by 100 mu s after the end of current flow. The densities of both W I and W II are found to be too low relative to the densities of neutral hydrogen and electrons, respectively, to have a significant influence on the recovery process.

Measurement of the radial metal vapour distribution in a pseudospark switch

Radial density profiles of molybdenum vapour in the centre of the electrode gap of a pseudospark switch have been obtained using one-dimensional laser-induced fluorescence imaging. Owing to the random nature of generation and motion of electron emission sites, which are the predominant metal vapour sources, the vapour density distribution fluctuates strongly from one discharge to another. However, from the positions of the maxima of the distribution it can be inferred that emission sites occur primarily on the edges of the cathode hole. Although the absolute number density of Mo vapour emanating from cathodes containing tantalum carbide increases with the tantalum carbide concentration, no marked differences in shape and temporal development of the density profiles can be identified.

PSEUDOSPARK SWITCH DEVELOPMENT AT SIEMENS - COMPACT HIGH POWER SWITCHES FOR GAS DISCHARGE LASERS AND

The development of high average power, high peak power pseudospark switches for use in TE gas discharge lasers and applications requiring similar ratings is reported. The results achieved as of today in metaceramic sealed-off experimental tubes of conventional pseudospark geometry are: hold-off voltage 32 kV; peak current 15 kA; pulse duration 300 ns; current rate of rise > 300 Wp; pulse repetition rate > 250 pps; tum-on delay 200 ns; temporal jitter < 10 ns. These are all simultaneous ratings, with limitations mainly imposed by the power supply and the pulse forming network, respectively. An attractive alternative to spark gaps, with their lifetime being limited to the order of 40 kC, also is the pseudospark switch, which incorporates the advantages of both thyratron and spark gap high dI/dt, high reverse current capability, high charge transfer capability, long lifetime, low jitter, cold-cathode operation without having the correspondmg disadvantages. At SIEMENS, a pseudospark switch the CAVATRON has been developed in order to replace a commercial spark gap in the SIEMENS LITHOSTAR kidneystone hsintegrator. Typical operating parameters (simultaneous ratings) are: hold-off voltage 22 kV; current rate of rise > 10 kA/p; peak current 10 kA; peak reverse current 5 kA; pulse duration 3.5 ps; pulse repetition rate 5 5 pps. The total lifetime amounts to 1.5*105 A*s of transfered charge, as compared to 4*1