GENERATOR OF HIGH-POWER NANOSECOND PULSES WITH A SWITCH ON GAS MICROGAPS

P. A. VOROB’EV, Yu. F. POTALITSYN

(Presented by the scientific seminar of the Department of High-Voltage Engineering)

A generator has been built with one shaping cable and one transmission cable of the RK-106 type, producing pulses with amplitudes from 3 to 40 kV at a pulse duration of 0.25 μs on a capacitive load of 30 pF. The rise and fall times of the pulse on the load are no more than \(3 \cdot 10^{-9}\) s. The generator is triggered by a pushbutton and by an external trigger pulse at a repetition frequency of 20 Hz. The switching element is a spark gap on gas microgaps, operating without readjustment over the indicated voltage range.

Fig. 1. Block diagram of the generator.

The generator consists of the following main units (Fig. 1):

1. A multisection spark gap filled with nitrogen at a pressure of 5 atm.
2. A shaping cable \(L_1\) and a transmission cable \(L_2\) of the RK-106 type.
3. A triggering device and a generator-control device.
4. A charging device with a voltage of 50 kV.
5. A coaxial capacitive voltage divider for recording pulses on an oscilloscope \((C1\text{-}14)\) SO-1.

The triggering device and the high-voltage control device (Fig. 2), together with the spark gap, are housed in one enclosure. The principal technical characteristics of the triggering device are as follows:

1. Trigger-pulse amplitude — 7 kV.
2. Trigger-pulse duration — \(3 \cdot 10^{-7}\) s.
3. Rise time — \(2 \cdot 10^{-8}\) s.

1. Maximum repetition frequency of the trigger pulse
   — 20 Hz.

2. Input capacitance
   — 25 pF.

The trigger-pulse generator (Fig. 2) is assembled with two tubes: tube $L_2$, 6N8S (a delayed blocking oscillator), and tube $L_1$, TGI1-90/8. A high-voltage pulse of negative voltage is formed at the control electrode of the switch (Fig. 3) when the RK-1 cable and capacitance $C$ are discharged through tube $L_1$. Control of the generator is carried out either from a pushbutton or from the “start” terminal from an external 50-V pulse source.

Fig. 2. Circuit of the trigger device and the generator control circuits.

The switch device is shown in Fig. 3. The ring electrodes 1 are mounted on the hollow insulating base 2 and are separated by gaps 200 µm long each. The voltage across the gaps is distributed by resistors 3. The forming cable 4 is connected to the end electrode at the top, and the formed pulse is taken off through cable 5, which is connected to the end electrode at the bottom. The control electrode is the metal cylinder 6, to which the pulse is fed through cable 7. The entire device is located in a metal housing 8, inside which an insulating cylinder 9 is installed for insulation. The coupling of the switching elements and the trigger circuits during operation of the device is effected by the electric field arising at the moment the trigger pulse is applied to the control elect-

des. The displacement currents arising in the capacitances \(C_1\), \(C_2\), and \(C_3\) (Fig. 4) sharply distort the linear distribution of voltage across the gaps of the switch. The magnitude of the voltage step at the \(k\)-th gap, when conditions are created for breakdown of the gaps in the direction of motion of the pulse being formed, is determined by the expression [1]:

Fig. 3. Switch design: 1 — working electrode; 2 — insulating base; 3 — resistor \((R = 5.1\ \text{M}\Omega)\); 4 — forming cable; 5 — transmission cable; 6 — control electrode; 7 — trigger cable; 8 — shield (chamber); 9 — insulating cylinder.

\[ \Delta U_k = \frac{U_0}{m} + \frac{2j U_0}{m}\cdot \operatorname{sh}\frac{\gamma}{2}\cdot \frac{\operatorname{ch}\left(m-k+\frac{1}{2}\right)\gamma} {\operatorname{sh}(m-j)\gamma} + \frac{4U_{\text{п}}C_1}{C_1+C_3}\operatorname{sh}\frac{\gamma}{2}\cdot \frac{ \operatorname{sh}\left(\frac{m+j+1}{2}-k\right)\gamma }{ \operatorname{ch}\frac{m-j}{2}\gamma }, \tag{1} \]

where \(U_0\) is the operating charging voltage of the forming line \(L_1\);

\(m\) is the total number of gaps;

\(k\) is the number of the gap on which \(\Delta U_k\) is determined;

\(j\) is the number of gaps that have broken down.

\(U_p\) is the amplitude of the triggering pulse of negative polarity.

\[ \gamma=\ln\left[1+\frac{C_1+C_3}{2C_2}\left(1+\sqrt{1+\frac{4C_2}{C_1+C_3}}\right)\right]. \tag{2} \]

In the switch that was built, \(\gamma=1.45\), \(\dfrac{C_1}{C_3}\approx 6\); \(m=30\). With such design parameters, as follows from (1), the predominant part of the triggering-pulse voltage is initially applied to the first gap, and after its breakdown—to the second, and so on. The overvoltage on

Fig. 4. Equivalent electrical circuit of the switch.

the subsequent gaps increases as a result of redistribution of the operating voltage. This helps reduce the triggering delay time and increase stability; and when the amplitude of the triggering pulse is 3–4 times greater than the static breakdown voltage of a single gap, breakdown of the gaps can occur in the absence of charging voltage on them.

In [2] we presented comprehensive studies of this switch. The limiting values of the triggering delay times of the device, for the above-indicated parameters of the triggering pulse and charging voltage, vary respectively from 150 to 20 nsec, while the absolute spread of delay times varies from 50 to 2 nsec. The leading edge of the output pulse over the entire range of operating voltages is 1 nsec.

In developing the generator, the design features of such devices set forth in [3] were taken into account.

The generator that was built is used at the Physics Institute of the Armenian Academy of Sciences of the USSR for experiments in quantum electronics.

REFERENCES

1. P. A. Vorob'ev, “On the theory of a high-power nanosecond switch on gas microgaps,” Izv. vyssh. uchebn. zavedenii—Elektromekhanika, in press.

2. P. A. Vorob'ev, G. A. Mesyats, Yu. F. Potalitsyn, “A new high-power controlled nanosecond switch,” ZhTF, No. 8, 1966.

3. P. A. Vorob'ev, V. V. Kremnev, “Generator of high-voltage nanosecond pulses,” Izd. GOSINTI, No. 18—65—1637/95, 1965.