APPLICATION OF A PHOTOMULTIPLIER TO THE STUDY OF LONG SPARKS

Student A. V. Shkilev

Supervisor: Professor, Doctor I. S. Stekolnikov

(Presented by the scientific seminar on dielectrics)

Knowledge of the formidable natural phenomenon of lightning has long occupied the minds of people. The first scientific experiments were carried out by Lomonosov, Richman, and Franklin. The study of the physics of lightning began especially intensively in the 1920s–1930s of our century. In a short time a large body of experimental material was accumulated. It was established that the mechanism of discharge in long gaps is analogous to the phenomenon of lightning. In both cases the discharge consists of two principal stages: 1) the leader stage and 2) the stage of the main channel. Therefore, the more accessible experimental study of spark discharge under laboratory conditions has made it possible to come close to solving many questions connected with elucidating the mechanism of lightning and with lightning protection.

The principal methods [1, 2] for studying a long spark in the laboratory are oscillographic recording of discharge currents together with optical methods (direct photography, Boys photography, Kerr photography). However, optical methods have a number of shortcomings that significantly affect the processing of results.

Science and technology at the present time possess high-quality electro-optical instruments—photomultipliers (PMTs) [4], whose characteristics make it possible to use them for studying light processes that change at high speed, as occurs in spark discharge. Therefore, in the Laboratory of High-Voltage Gas Discharge (LVGR) of the Energy Institute of the Academy of Sciences of the USSR, a number of investigations have been carried out on the application of PMTs to the study of long sparks.

Below two fundamentally different cases of the use of a photomultiplier in studying spark discharge will be shown: 1) a PMT as an auxiliary device for synchronizing an impulse-voltage generator with an oscillograph when measuring current at high potential;

2) a PMT as a recording device for studying the character of the discharge glow.

PMT for Synchronizing the PIG with an Oscillograph When Measuring the Leader Current at High Potential

At the LVGR, a method was developed for recording the leader current “from below” (point a, Fig. 1) and “from above” (point b, Fig. 1) [1].

For oscillographing the leader current at point “b,” a suspended cabin was mounted in which the oscillograph was placed. The oscillograph was powered with a dc voltage \(U = 15\) kV from a battery of accumulators by means of a converter with an output voltage of 110 V at a frequency of 400 Hz and a rectifying device. The ignition voltage pulse from the resistance \(R_3 = (10 \div 100)\) kΩ, connected ahead of the damping resistance \(R_{\mathrm{T}}\), synchronizes the switching-on of the oscillograph (opening of the beam and start of the time sweep) with the phenomenon under investigation (Fig. 1). A substantial drawback of such direct synchronization is the delay in switching on the oscillograph relative to the phenomenon. Figure 2 shows an oscillogram of the leader current obtained with the described synchronization method.

Fig. 1. Diagram for recording the current of the leader of a long spark.

Fig. 2. Oscillogram of the leader current when measured at high potential in the case of direct synchronization.

As can be seen, the beginning of the phenomenon is absent from the oscillogram. In practice the delay time (the time from the beginning of discharge development to the appearance of the beam on the screen of the cathode-ray tube) reaches about 8 μsec. With small damping resistances \((R_t = 1.6 \div 5.5\ \text{k}\Omega)\), the leader development time is of the same order as the delay time. It is therefore necessary to use another synchronization circuit, in which the oscillograph would be switched on before the last spark gap of the impulse-voltage generator breaks down. At the LVEG of ENIN, a circuit for synchronizing the impulse-voltage generator with the oscillograph during measurements at high potential by means of a photomultiplier was developed and implemented. Its block diagram is shown in Fig. 3.

Fig. 3. Block diagram for synchronizing the impulse-voltage generator and the oscillograph by means of light pulses using a photomultiplier.

On the outer wall of the suspended cabin a spark gap is mounted (breakdown voltage \(U_{\text{br}} = 5\ \text{kV}\)); when it breaks down, a light pulse is emitted. The latter is received by a photomultiplier located at “ground,” and is then converted into a voltage pulse with a steep front and an amplitude of 25 kV, which fires the impulse-voltage generator. The basic circuit of the photomultiplier unit and of the converter device is shown in Fig. 4.

Fig. 4. Electrical circuit of the photomultiplier unit and converter device of the synchronization circuit.

The circuit for switching on the oscillograph synchronously with the separation of the light pulse by the spark on the wall of the booth is shown in Fig. 5.

When the experimenter closes switch \(K_1\), breakdown of the gap \(U_1\) occurs. The moment at which the electronic oscillograph is switched on is regulated by a delay element.

Fig. 5. Starting circuit of the oscillograph and the GIS

Using the described circuit, oscillograms of the leader current were obtained, one of which is shown in Fig. 6.

Fig. 6. Oscillogram of the leader current measured at high potential (synchronization by means of the photocell)

Experimental work with the circuit for synchronizing the GIS with the oscillograph by means of a photocell showed the following advantages:

1. The circuit makes it possible to obtain practically any delay time of the GIS (phenomenon) relative to the moment at which the oscillograph is switched on.

2. The circuit is simple, stable, and reliable in operation.

3. The circuit makes it possible to obtain the complete oscillogram of the leader current (compare Fig. 2 and Fig. 6) both for prolonged leader processes (with a duration on the order of \(300\ \mu\text{sec}\)) and for short ones (with a duration on the order of several tens of \(\mu\text{sec}\)).

A Photomultiplier for Studying the Character of the Glow of a Long Spark over Time

The characteristics of the photomultiplier make it possible to obtain quantitative relationships with a sufficient degree of accuracy at high rates of change in glow intensity, as occurs in the leader and channel stages of discharge development.

The photomultiplier converts pulses of light energy into current pulses. To record the glow of the leader and of the main discharge channel over time, a high-speed high-voltage oscilloscope is required. In order to carry out a qualitative interpretation of the oscillograms, a voltage pulse with a maximum amplitude of not less than 150 V must be applied to the deflection plates of the oscilloscope. Across the load resistance of the photomultiplier, whose value is determined by the output capacitance of the photomultiplier \((C_{\text{out}}\simeq 10\ \text{pF})\) and by the steepness of the decay of the pulse under investigation, a pulse with an amplitude of the order of 1 V is produced. Therefore this pulse must be amplified without distortion by approximately a factor of 200.

The block diagram for recording the character of the glow of a long spark over time, for the leader and channel stages, is shown in Fig. 7.

Fig. 7. Block diagram for recording the character of the glow of a long spark over time.

The light pulses of the discharge are converted by the photomultiplier into electrical pulses, amplified, and applied to the deflection plates of the oscilloscope. The oscilloscope is switched on strictly synchronously at the required instant of time. This is accomplished by selecting the corresponding delay time.

When studying the character of the glow of a spark in a high-voltage hall, it is necessary to reduce the light fluxes reaching the photocathode of the photomultiplier from other radiation sources (scattered daylight, electric lighting). For this purpose the method of an “optical pit,” used at the High-Voltage Generator Laboratory of ENIN when photographing the discharge, is employed, as well as a tube installed in front of the photomultiplier. The arrangement for recording the character of the glow of a long spark is shown in Fig. 8.

The study of the leader and channel stages with the aid of a photomultiplier will be carried out in two ways:

1) the photomultiplier records the light flux from the entire discharge gap (Fig. 9);

2) the photomultiplier records the light flux only from a section of the spark gap (Fig. 10).

When recording the light flux of the spark by the second method, a narrow slit is used to “cut out” a small section of the discharge channel.

Fig. 8. Circuit for recording the character of the glow of a long spark

Fig. 9. First method of recording the glow of a long spark

Fig. 10. Second method of recording the glow of a long spark

slit. The light from the spark can reach the photomultiplier photocathode only through the slit. By this method it is possible to determine the velocities of development of the leader and of the main channel by comparing oscillograms of the spark glow in time for different sections of the discharge gap.

When the photomultiplier is supplied with a dc voltage of \((1000 \div 1800)\) V, the anode current of the photomultiplier has a small value (approximately 1 mA), and therefore at the output of the photomultiplier the voltage is of the order of 1 V. The literature describes cases of using a photomultiplier to study physical processes with pulsed high-voltage supply (5 kV). In this case the possible anode current of the photomultiplier increased to 200 mA. Current pulses in the anode circuit of the photomultiplier of magnitude \((100 \div 200)\) mA produce, across a load resistance equal to 1000 ohms, voltage pulses with an amplitude of \((100 \div 200)\) V.

Thus, such voltage pulses can be recorded directly by an electronic oscillograph without an amplifier at higher sweep speeds. It is therefore expedient to use a photomultiplier with pulsed supply. In this case, for recording a phenomenon with a high sweep speed, it is also convenient to use an oscillograph with pulsed supply. The recording circuit is shown in Fig. 11.

Fig. 11. Circuit for connecting the photomultiplier and oscillograph with pulsed supply.

In the initial state, two identical artificial lines are charged to a voltage \(U_0 = 2\) kV. When the thyratron T11—0.1/1.3 is “tripped” by a synchronizing pulse, a pulse with amplitude \(kU_0\) is produced on the load resistance of the pulse transformer, equal to the reduced double surge impedance of the lines \(2k^2 R_B\) (\(k\) is the transformation coefficient).

The winding of the pulse transformer is made bifilar, and the filament voltage from a low-voltage transformer is supplied through two conductors to its “m” and “n” terminals.

The advantage of such a circuit is its compactness. In addition, the photomultiplier gives a pulse of the phenomenon only at the moment of the sweep, i.e., the phenomenon is always located on the screen, and furthermore the entire system is put into operation by a single synchronizing pulse, which can be applied at any required instant of time.

References

1. I. S. Stekolnikov. Study of the nature of the long spark. Izvestiya of the Academy of Sciences of the USSR, Department of Technical Sciences, No. 11, 1952.

2. I. S. Stekolnikov. Study of the nature of the long spark. Izvestiya of the Academy of Sciences of the USSR, Department of Technical Sciences, No. 2, 1953.

3. I. S. Stekolnikov. Pulse Oscillography and Its Application. Moscow–Leningrad, Publishing House of the Academy of Sciences of the USSR, 1949.

4. M. O. Chechik, S. M. Feinstein, and T. N. Lifshits. Electron Multipliers, 1954.

5. Ya. S. Itskhoki. Pulse Technology. Moscow, Sovetskoe Radio, 1949.