ДИНАМИКА ПОТЕНЦИАЛА И ТОКА ПРИ ЭЛЕКТРИЧЕСКОМ ПРОБОЕ АТОМАРНОГО ГАЗА В ЭКРАНИРОВАННОЙ ТРУБКЕ

POTENTIAL AND CURRENT DYNAMICS AT THE ELECTRIC BREAKDOWN OF ATOMIC GAS IN A SHIELDED TUBE WITH THE GROUNDED ANODE

1. Introduction.

The ability to propagation is one of the basic properties of the non-stationary electric discharge. However in different discharges it is shown in a various degree and differently. So some modes of the impulse capillary erosive discharge, considered in connection with experimental modelling of the ball lightning (BL) [1], differ fast development from an interelectrode gap (~ 3-25 millimeters) thin (some millimeters) and long (some tens centimeters) plasma jet having complex current system [2] . In [3] we denoted analogy of this process to movement of the breakdown wave’s leader though in these two cases set of the transfer phenomena has obvious distinctions. For clearing one side of the propagation mechanism and discharge structure formation we shall consider some special cases of electric breakdown: prebreakdown condition and breakdown at slow increase of negative voltage relative to the period of a breakdown wave movement, the pulse breakdown and, probably, breakdown via high-frequency and microwave pulse, the gliding discharge. In spite of the fact that original experiments had been carried out one and a half decades ago, the present expanded reproduction, in our opinion, has not lost the barest necessity. In principle, all considered sides of the breakdown phenomenon are mentioned in [4], we shall take concrete cases and receive concrete dependences of parameters at the measurements.

2. Experiment procedure.

For the experimental description of breakdown as complex electric process the measurement of potentials and currents in the circuit, one element with distributed and variable in time parameters of which is plasma, is necessary. Therefore we measured dependences on the time of the cathode potential U₀ (t) and at 56 points l_k along the plasma volume U_pl (l_k, t), total current onto the ground I (t), current onto the grounded anode I_a (t) and the displacement current I_b (t) onto grounded shield.

Reception of these parameters measurements being least dependent from each other can be achieved only under small enough diameter of plasma volume relative to distance between electrodes, small distance between a probe and plasma, and also in the presence of a screen placed close enough to plasma and isolator between it and plasma. (Then the displacement current increases and substitutes for the conduction current onto the shield, but the measured currents are interpreted unequivocally and have classical sense.) These requirements are satisfied in a sufficient degree in the case of a thin long dielectric tube with gas of the lowered pressure placed in a metal pipe, diameter of which D_p is much less than a interelectrode distance L but is much more than diameter of discharge tube D_t and, accordingly, than diameter of cylindrical plasma volume d_pl , i.e.:

L>>D_p>> D_t ~ d_pl , (1)

and distance between the end of a probe R_z and plasma d_pl /2 is about radius of an aperture of the tube d_t, step of a probe Dl is about the distance from the probe end to axis of a tube:

R_z – d_pl /2 ~ d_t /2 (2)

Dl = (l_k – l_k-1 ) ~ R_z (3)

In view of the big resistance of a breakdown plasma under slow increase of a voltage and the big role of the displacement current the placing of a probe inside of a tube results in significant distortions of all researched process. In this situation the most simple decision represents the application of a capacitive probe which is placed near to a tube but small enough in size. The best form of a probe - a ball on the most thin wire but because of a low mechanical durability of such design we used a wire probe as a relocatable one. When the wire diameter is ~ d_t /10 and condition (2) its capacity onto plasma C_z-pl is about the one tenth share of a picofarad, when capacity of the probe onto shield is C_z-sc = 15 nF the signal taking off from C_z-sc as a lower arm of the capacitor divider, has magnitude of the order of a millivolt while magnitude of U_pl is of the order of a kilovolt. Even without the voltage follower the time constant of this divider is t₀ = C_z-sc ^. R_osc = 15 ms (resistance of the oscillograph input is R_osc = 1MW). For a deviation of a signal from true value of potential less than on ~ of 5 %, the linear increase of cathode potential should result in breakdown during 1,5 ms. Such increase will be slow when period of the breakdown wave movement is about ~ 10^-5 s and less. More big accuracy is hardly expedient as there are also other sources of a mistake. Besides ones mentioned above and connected to a limit of the resolution on axial coordinate there is also a mistake connected with the neglect of radial distribution. So for example, the positive streamer has very small diameter of the beginning part. In the given case the negative streamer is considered. For the control of a probe influence the signal U₀ (t) was recorded on oscillograms U_pl (l_k, t) at all l_k with the help of a double-beam or double-channel oscillograph.

3. Experimental setup.

In view of the specified requirements the installation for creation and researches of breakdown consisted of the shielded discharge tube, shaper of an increasing negative voltage pulse, squarer, high-frequency and microwave pulses, measuring resistors, capacitive probes and recording oscillographs with a videocamera. The oscillograms displaying dependences U₀ (t), U_pl (l_k, t), I(t), I_a (t) и I_b (t) were fixed by a videocamera, were entered in a computer, and axial distribution of potential for the characteristic time moments was determined from them. Videoshooting of a luminescence of discharge's parts including a stroboscopic mode allowed to observe features of its radial and axial distribution, their dynamics.

Fig.1. Discharge tube GSh-2 and its electrodes. Anode (at the left) was used as the cathode.

As the discharge tube the noise generator GSh-2 filled with a neon at pressure ~ 8 Torr ( Fig. 1) was used. Length of the glass tube was 360 mm, the small diameter - 7 mm, the big diameter - 18 mm, the distance between electrodes L = 310 mm, the distance between the anode and the expanded part of the tube - 290 mm. The anode is located at 2 cm from the ending of a tube of small diameter and represents a metal beaker in diameter and length 4 mm opened to the second electrode and fixed axially at a wire lead in diameter 1,1 mm. The heated cathode as a disk in diameter 6 mm was located orthogonally to axes of the tube and inserted into a thin-walled tube in diameter 7 mm (in the experiment the heat was not used).

Fig. 2. Shielding pipe with the relocatable

and control capacitive probes.

The discharge tube settled down axially in a brass silvered pipe 34 x 40 mm in the length 410 mm (fig. 2) and was fixed on the end of big diameter by a dielectric ring and on the other end - a metal disk. The disk had the central aperture with a spring clip for wire lead of anode GSh and was fixed within a pipe without galvanic contact to it by a dielectric ring or by a ring plunger creating a reliable microwave short circuit of the electrode to shielding pipe. At the microwave breakdown the pipe with internal diameter 20 mm was used, and the discharge tube was a part of the central conductor of a coaxial line between a pulse magnetron and a terminating load. In special cases other schemata of inclusion of a tube in a screen were used also. Shielding pipes had a longitudinal slit, and the shielding body of coaxial leads of probes had grooves for fixing in a slit with an possisility of moving along it (fig. 3) and contained feedthrough condenser of the lower arm of capacitive divider C_z-sc .

Fig. 3. A capacitive probe

The circuit of the increasing negative voltage pulse shaper is represented at fig. 4. The pulse from generator G5-63 triggered off the trace of a double-channel oscillograph and with an adjustable delay the squarer which gave a pulse in duration 1 ms and voltage up to 25 kV. This pulse carried out a fast charge of the condenser C₁ = 100 pF through the diode D₁ (2 x Д1006) and resistor R₁ = 1 kW with the time costant t₁ = 0,1 ms. For return of the circuit to the initial state before the following start the discharge of the condenser C₁ through resistor R₂ = 100 MW with the time constant t₂ = 10 ms was realized. Condenser C₁ charged condenser C₂ = 15 pF through resistor R₃ = 43,22 MW with the time costant t₃ = 0,65 ms. The triggering regime was monopulse or periodic with the least frequency at which the process was periodic — 10 Hz and higher.

Рис. 4. The circuit of the increasing negative voltage pulse shaper.

The voltage from condenser C₂ (the shaper's output) was given onto the cathode of the discharge tube. Its anode and shielding pipe were grounded through measuring resistors R₄ = R₃ = 2,645 kW accordingly the signals from which were given onto double-channel (double-beam) oscillograph C1-83 (C1-74) for measurement I_b (t) и I_a (t) (fig. 5). For measurement of total current I (t) = I_a (t) + I_b (t) the anode was joined to shield, and amplification of the oscillograph was increased twice. For measurement of potentials the inputs of the oscillograph were switched to outputs of the probes, and shielding pipe was grounded. For the undistorted display of small currents of a prebreakdown stage the voltage limiters 0,6 V of diodes were included in parallel to the inputs of the oscillograph.

Fig. 5. Circuit of measured signals.

Calibration of the relocatable probe.

For calibration the relocatable probe was established against middle of the cathode (anode GSH), and by means of moving of the second probe relative a cathodic conductor of the circuit (fine tuning C_z-c) we achieved identical signals on outputs of both probes at known amplitude of a pulse of microsecond duration. As the charge, formed in plasma surrounding the cathode at development of prebreakdown process , changes a signal on the input of a probe, then for calibration the mode of monopulse triggering was used when in rare cases a breakdown did not occur, and the prebreakdown condition developed poorly (fig. 6).

Fig. 6. The oscillogram at calibration of a relocatable probe.

5. Some results of the measurements.

Voltage-current characteristic of GSh-2 without a shielding pipe with linear and logarithmic scale of a current is represented at fig. 7.

Fig.7. Voltage-current characteristic of the discharge tube (the anode of GSh used as the cathode).

At fig. 8. oscillograms of probes signals U₀ (t) and U_pl (l₀,t) are represented. The relocatable probe is placed against a cut of the cathode.

Fig. 8. The probe signal at a cut of the cathode l₀ = 0 at frequency of the triggering 10 Hz.

At fig. 9 the oscillogram of the probe signal at the point of a tube axis with the maximal negative residual potential before the triggering is presented l₉ = 45 mm.

Fig.9. The probe signal at the point of a tube axis with the maximal negative residual potential before the triggering is presented 10 Hz.

At fig. 10 the oscillogram of the total current during breakdown and discharge is presented, and at fig. 11 — the anode current I_a (t) and the displacement current I_b (t) onto the screen.

Fig. 10. Total current of the breakdown and discharge consisting

of the anode current I_a and the displacement current I_b .

Fig. 11. "Splitting" of the total current of breakdown and discharge

to the anode current I_a and the displacement current I_b .

At fig. 12 the oscillogram of the anode current I_a (t) and the displacement current I_b (t) containing the prebreakdown period is given. In parallel to inputs of an oscillograph the voltage limiters on 0,6 V are connected up.

Fig. 12. The oscillogram of the anode current I_a (t) and the displacement current I_b (t) containing the prebreakdown period.

At fig. 13 the assemblage of 56 oscillograms of a plasma potential during the breakdown and discharge at the triggering frequency 10 Hz is presented. One division across — 40 ms, the upper horizontal line — «0,8 kV», lower — «- 0,8 kV», middle — «0 kV», and numbers indicated l_k — the distance from a probe to the cathode cut in millimeters.

Fig. 13 The assemblage of 56 oscillograms of a plasma potential during the breakdown and discharge at the triggering frequency 10 Hz. One division across — 40 ms, the upper horizontal line — «0,8 kV», lower — «- 0,8 kV», middle — «0 kV», and numbers indicated l_k — the distance from a probe to the cathode cut in millimeters. The special points of dependences the sense of which is explained in the text further are marked at the insertion placed on the right and below.

Let's pay attention to some special points at oscillograms, marked at the insertion placed on the right and below at fig. 13.

U_A — the plasma potential before the front of the streamer,

U_b1 — potential behind the front of the streamer,

U_B — potential of the streamer before the back stroke wave,

U_b2 — potential of the streamer after the back stroke (the beginning of the discharge onto the anode),

U_C (l) — potential of plasma after the discharge,

U_b1-A — the potential jump at the streamer front,

U_BA — the potential jump of the streamer.

Our capacitive probe measures only the alternative potential changing sufficiently fast. Therefore zero lines at the oscillograms at various values l do not correspond not only to zero of potential but even to some single value. As the potential of the cathode falls almost up to zero when the discharge is over (fig. 8), then we shall choose U_C (l) as the reference level.

At fig. 14 the remarkable assemblage of curves showing a distribution of some of the specified potentials between the cathode and the anode of the discharge tube with the exception of a site of big diameter (290 - 310 mm) and close to it (285 - 290 mm) is presented.

Fig. 14. Distribution of potentials along a tube at the characteristic moments of transient:

U₀ – U_C — residual potential to the triggering moment,

U_A – U₀ — additon of potential during the prebreakdown period,

U_A – U_C — potential before the breakdown,

U_B — U_A — potential jump of the streamer,

U_B – U_C — potential before the back stroke,

U_B2 – U_C — potential after the back stroke.

At increase of the triggering frequency up to 400 Hz the voltage of breakdown fell, duration of the period from a start up to breakdown decreased, duration of the breakdown and the discharge decreased, and the discharge current grew. The zone of a tube with residual potential expanded. The near-cathode contraction also was displaced to the cathode, it amplified and enriched by stratum. Transition through the frequency 400 Hz was accompanied by sharp increase of duration of a pulse of the breakdown-discharge and full disappearance of the contraction. At fig. 15 the images of a breakdown and a discharge in a cathodic half of tube are given at frequency of triggering 50 Hz, and at fig. 16 - integrated images of the near-cathode plasma at the triggering frequency 100 Hz.

Fig. 15. Stroboscopic images of a cathodic half of the tube made with a period 5 ms

at the triggering frequency 50 Hz.

Fig. 16. Integrated images of the near-cathode plasma at the triggering frequency 100 Hz.

6. The brief description of the given case of a breakdown.

Let's consider the transient occuring in a tube. It is essential that due to a periodic mode, it begins not from the background concentration of charges in the gas, determined by a radioactivity background, action of cosmic rays and being random variable, but from the residual concentration which is small to form plasma without breakdown, but it is great enough not only to remove casual influence of the external factor, but also as to change a cross structure of the discharge.

After a microsecond charge of condenser C₁ the current I_R1 appears in spurts going through resistor R₁ and composed of the currents: I_C2 – the charge current of C₂ , I_C2P – the charge current of the supply conductor's capacity outside of a pipe C_2P , I_CA – the charge current of the cathode capacity and the supply conductor's capacity inside of a pipe C_A and very weak cathode current — I_k . Apparently from I_a (t) at fig. 12 the last one does not contain an essential current onto the anode up to the back stroke and is a component of the displacement current I_b (t).

Distribution of residual potential to the triggering moment at frequency 10 Hz is presented at fig. 14 (U₀ (t=0) - U_C). It looks like "barrier" (for electrons) located within a cathodic half of tube. The cathode current creates around a cathode the plasma of very low concentration and with the electron excess. The negative charge (U_A – U₀ (t=0) at fig. 14) fills in "hole" between the cathode and "barrier", and this process goes non-uniformly. At dependences of the plasma potential up to 6 cm from the cathode the jumps to which short impulses of a current correspond (fig. 12) are observed. In time their size grows, and the position moves away from the cathode. It - the short streamers which are not finishing by a breakdown. In the view of the characteristic fall of plasma potential and increase of a current previous to them the completed streamer (resulting to a breakdown) also should be referred to this sequence.

Before the beginning of the concluding streamer the distribution of potential along a tube (U_A - U_C fig. 14) looks like the "slope" stretching out up to the middle of a tube. It is visible from fig. 13 that birth of the streamer occurs in the maximal field, the potential jump (U_B – U_A at fig. 14) becomes appreciable only after 15 mm mark. Here it has obvious feature: though U_b1 –U_A is small relative U_A - U_C but it is great relative U_B – U_b1 . At this site the potential drop on length of the streamer is small. When l is about 95 mm the point b₁ appears in the middle between A and B, and after l ~ 120 mm it can not be fixed at all. In the middle of a tube the maximum of U_B - U_A is, its distribution along a tube repeats the shape of a current impulse of breakdown at fig. 11. At a culmination the streamer current reaches ~ - 1,5 mA. After passage of the middle of a tube the potential jump of the streamer decreases, and near anode only its small part is still. Distribution of potential before a touch of the streamer with the anode is presented at fig. 14 (U_B – U_C ).

Short circuit of plasma with the anode causes a wave of potential redistribution to a condition determined by the plasma conductivity which is formed both direct and a back wave (U_b2 – U_C рис. 14) . This wave, the back stroke, has in the beginning so abrupt front as well as the streamer but goes along plasma, already created by the streamer, faster and fades to the cathode. The current of an excess negative charge taken off from plasma mixs up with the discharge current of all capacities charged earlier and flows down onto the anode (fig. 10 and 11).

When a contact with the anode the streamer current is less by third than the maximal value - 0,5 mA, i.e. resistance of plasma is about 3 MW. During the discharge the electron multiplication begins, and resistance has time to decrease almost by the order. With reduction of the triggering period the initial concentration grows, therefore duration of the breakdown-discharge impulse decreases while current grows.

The streamer transit time of the interelectrode interval in length 0,31 m is about 5 ms, so the average speed of breakdown in the given mode is only ~ 60 km/s. Maximal electric power demand during the breakdown is 1,4 kV ^.1,4 mA = ~ 2 W, and total electric energy ~ 7 mJ.

Features of a transverse structure of the discharge.

The contraction concentrates within the near-cathode area, i.e. where ions collect at the dark discharge [4]. It is most expressed when the cathode potential is supported at the trancritical level almost all time. The received experimental data allow to put forward the assumption of an essential influence of the non-quasi-neutral plasma and its distribution on the characteristics of a breakdown and a subsequent non-stationary discharge concerning features of its transverse structure. The specified effects were observed in atomic gas which is chemically poorly active as against molecular gases and a dust plasma especially on the basis of electrets [5].

(In our opinion simple variants of installation can be used as the experimental part of a school study on the phenomenon of a breakdown)

References.

1. Avramenko R. F. Ball lightning in the laboratory. Moscow:Chimiya, 1994. 256 p. (in Russian)

2. Kirko D.L., Savjolov A.S. The current structure of a capillary discharge’s torch. Proc. of V Russian Seminar “Modern means of plasma diagnostics and application of them for matter and environment control”. Moscow, MIPhI, 27-29 july 2006. P. 106.

3. Emelin S.E., Pirozerski A.L., Semenov V.S., Skvortsov G.E. Propagation characteristics of the dynamic state in a capillary discharge jet. Tech. Phys. Lett. 23 (10), 1997. pp. 758-759

4. Raizerр Yu.P. Physics of the gas discharge. Moscow: Nauka. 1992. — 536 p.

5. Bychkov V.L. On the electric charging of polymeric structures. Preprint MIPhI. Moscow: 1992, 16 p.

And how occur an electric breakdown under the same conditions but in absence of the prebreakdown period, at the "momentary" appearance of a voltage? We shall know about it in New 2007 Year!

IMPULSE BREAKDOWN