Email alert
RSS2012 Vol. 24, No. 04
Recommend Articles
Display Method:
2012,
24.
2012,
24: 757-764.
doi: 10.3788/HPLPB20122404.0757
Abstract:
The compact Marx generators, which can operate at a certain repetition frequency with small size, light weight, and high energy efficiency, are widely used in narrowband, wideband and ultra-wideband high power microwave (HPM) sources. This type of HPM source based on compact Marx generators is a worldwide research focus in recent years, and is an important trend of development. The developments of this type of HPM source are described systemically in this paper. The output parameters and structural characteristics are reviewed, and the trends of development are discussed. This work provides reference and evidence for us to master the status of the HPM source based on compact Marx generators correctly and to explore its technical routes scientifically.
The compact Marx generators, which can operate at a certain repetition frequency with small size, light weight, and high energy efficiency, are widely used in narrowband, wideband and ultra-wideband high power microwave (HPM) sources. This type of HPM source based on compact Marx generators is a worldwide research focus in recent years, and is an important trend of development. The developments of this type of HPM source are described systemically in this paper. The output parameters and structural characteristics are reviewed, and the trends of development are discussed. This work provides reference and evidence for us to master the status of the HPM source based on compact Marx generators correctly and to explore its technical routes scientifically.
2012,
24: 765-770.
doi: 10.3788/HPLPB20122404.0765
Abstract:
With the development of linear transformer driver, the number of gas switches in the driver increases rapidly, and thus multi-output synchronization trigger has become one of the key technical obstacles. The paper summarizes the recent researches about electical pulse trigger, laser trigger and trigger based on photo-conductive semiconductor switches. Characteristics and trigger requirements of multi-channel multi-gas switches are reported. Developments in key technologies of trigger system are analyzed. The crucial technique problems that should be investigated in the future and some research approaches are presented.
With the development of linear transformer driver, the number of gas switches in the driver increases rapidly, and thus multi-output synchronization trigger has become one of the key technical obstacles. The paper summarizes the recent researches about electical pulse trigger, laser trigger and trigger based on photo-conductive semiconductor switches. Characteristics and trigger requirements of multi-channel multi-gas switches are reported. Developments in key technologies of trigger system are analyzed. The crucial technique problems that should be investigated in the future and some research approaches are presented.
2012,
24: 771-775.
Abstract:
The paper summarizes the recent developments of the pulsed power switch(PPS), and the status of the pulsed power thyristors around the globe, and presents the products and their characteristics for the PPS in China. The design and application of the pulse power thyristor module, and the test platform of the semiconductor PPS are elaborated. The key technologies used in the engineering applications of the pulse power thyristor and its module, such as synchronized drive, power-obtaining based on induction, and voltage-sharing in series, are presented, along with the results of the engineering applications of pulse power thyristor modules. Moreover, the development direction of the semiconductor PPS and its test platform are elaborated.
The paper summarizes the recent developments of the pulsed power switch(PPS), and the status of the pulsed power thyristors around the globe, and presents the products and their characteristics for the PPS in China. The design and application of the pulse power thyristor module, and the test platform of the semiconductor PPS are elaborated. The key technologies used in the engineering applications of the pulse power thyristor and its module, such as synchronized drive, power-obtaining based on induction, and voltage-sharing in series, are presented, along with the results of the engineering applications of pulse power thyristor modules. Moreover, the development direction of the semiconductor PPS and its test platform are elaborated.
2012,
24: 776-779.
doi: 10.3788/HPLPB20122404.0776
Abstract:
An efficient total-variation-diminishing radiation magnetohydrodynamics code FOI-1a has been developed. Calculations with FOI have been performed with in the parameters range of Yang accelerator and PTS facility. Numerical results demonstrate that for light array the radiation at peak is dominated by shock heating, while for massive array by shock heating together with adiabatic compression. The ratio of shock heating to adiabatic compression decreases with increasing of array mass. Precursor plasma produced by ablation weakens shock heating and hence has a negative effect on peak radiation. However, precursor plasma has another stabilization effect on magnetic-Rayleigh-Taylor instability, so there should be an optimum ablation rate that maximizes the peak radiation.
An efficient total-variation-diminishing radiation magnetohydrodynamics code FOI-1a has been developed. Calculations with FOI have been performed with in the parameters range of Yang accelerator and PTS facility. Numerical results demonstrate that for light array the radiation at peak is dominated by shock heating, while for massive array by shock heating together with adiabatic compression. The ratio of shock heating to adiabatic compression decreases with increasing of array mass. Precursor plasma produced by ablation weakens shock heating and hence has a negative effect on peak radiation. However, precursor plasma has another stabilization effect on magnetic-Rayleigh-Taylor instability, so there should be an optimum ablation rate that maximizes the peak radiation.
2012,
24: 780-784.
doi: 10.3788/HPLPB20122404.0780
Abstract:
A measurement system for the high voltage pulse power supply on HL-2A has been developed. It adopts the voltage divider based on high-frequency inductance-free glaze film resistor along with the fiber isolation technology based on voltage-frequency conversion. Moreover, a DSP chip is used in the fiber transmission, to realize A/D conversion, simplifying the circuit, enhancing the data transmission speed and anti-jamming ability of the system. The system has been applied to long-pulse(5 s) discharge experiments on HL-2A. The experimental results show that, the system realizes the potential isolation for power supply, and the measured voltage waveforms reflect the supply output faithfully, providing real-time voltage data for the feedback and protection system of the supply.
A measurement system for the high voltage pulse power supply on HL-2A has been developed. It adopts the voltage divider based on high-frequency inductance-free glaze film resistor along with the fiber isolation technology based on voltage-frequency conversion. Moreover, a DSP chip is used in the fiber transmission, to realize A/D conversion, simplifying the circuit, enhancing the data transmission speed and anti-jamming ability of the system. The system has been applied to long-pulse(5 s) discharge experiments on HL-2A. The experimental results show that, the system realizes the potential isolation for power supply, and the measured voltage waveforms reflect the supply output faithfully, providing real-time voltage data for the feedback and protection system of the supply.
2012,
24: 785-788.
doi: 10.3788/HPLPB20122404.0785
Abstract:
The paper proposes an MLS(multi-layer stacking) method to measure the high-energy X-ray energy spectrum．The MLS method has some unique merits: scattering is more easier to control during the measurement; the uniformity of the light field can be more effectively guaranteed; and measurement from different points of view can be realized. The principle of the MLS method is introduced as well as precautions during measurement. For a given X-ray energy spectrum, measuring devices using two different measuring media are designed with Monte Carlo method, and the scattering influence brought about by the devices can be controlled within 5%.
The paper proposes an MLS(multi-layer stacking) method to measure the high-energy X-ray energy spectrum．The MLS method has some unique merits: scattering is more easier to control during the measurement; the uniformity of the light field can be more effectively guaranteed; and measurement from different points of view can be realized. The principle of the MLS method is introduced as well as precautions during measurement. For a given X-ray energy spectrum, measuring devices using two different measuring media are designed with Monte Carlo method, and the scattering influence brought about by the devices can be controlled within 5%.
2012,
24: 789-792.
doi: 10.3788/HPLPB20122404.0789
Abstract:
The technical path combining pulse forming with inductive adding is one of the important candidates to realize pulsed power miniaturization. In this paper, the cell structure of the linear transformer driver(LTD), which is based on pulse forming network(PFN), with impedance 2.5 and pulse width 200ns is introduced. The outputs of two such LTDs are inductively added to a load through the cathode pole, which forms a double-cell module based on the LTDS. The experiment results show that the high-voltage pulse waveform from the double-cell module at charging voltage of 40 kV and repetition frequency of 20 Hz/40 Hz accords with that from PFN discharged directly, and its jitter is about 2 ns. The coherence of the output pulse waveforms is quite good at repetition mode.
The technical path combining pulse forming with inductive adding is one of the important candidates to realize pulsed power miniaturization. In this paper, the cell structure of the linear transformer driver(LTD), which is based on pulse forming network(PFN), with impedance 2.5 and pulse width 200ns is introduced. The outputs of two such LTDs are inductively added to a load through the cathode pole, which forms a double-cell module based on the LTDS. The experiment results show that the high-voltage pulse waveform from the double-cell module at charging voltage of 40 kV and repetition frequency of 20 Hz/40 Hz accords with that from PFN discharged directly, and its jitter is about 2 ns. The coherence of the output pulse waveforms is quite good at repetition mode.
2012,
24: 793-796.
doi: 10.3788/HPLPB20122404.0793
Abstract:
An isolated voltage divider was designed to measure the muzzle voltage for electromagnetic(EM) rail launcher, by which electrical isolation between probe and measured circuit was realized. Parameters affecting the divider frequency response were analyzed, and the range of frequency response was 14 Hz-18 MHz. The designed isolated voltage divider was calibrated by a P6015A voltage probe. The two output signals are in good agreement with each other. The rise time of the pulse is about 1 s, corresponding to the upper frequency of 350 kHz, which indicates that the frequency response of the divider satisfies the requirements of the EM launcher. The resistor divider and the isolated voltage divider were used to measure the muzzle voltage during EM launch. The measured waveforms are consistent with each other. Therefore, the isolated voltage divider meets the measuring requirements of the launcher muzzle voltage.
An isolated voltage divider was designed to measure the muzzle voltage for electromagnetic(EM) rail launcher, by which electrical isolation between probe and measured circuit was realized. Parameters affecting the divider frequency response were analyzed, and the range of frequency response was 14 Hz-18 MHz. The designed isolated voltage divider was calibrated by a P6015A voltage probe. The two output signals are in good agreement with each other. The rise time of the pulse is about 1 s, corresponding to the upper frequency of 350 kHz, which indicates that the frequency response of the divider satisfies the requirements of the EM launcher. The resistor divider and the isolated voltage divider were used to measure the muzzle voltage during EM launch. The measured waveforms are consistent with each other. Therefore, the isolated voltage divider meets the measuring requirements of the launcher muzzle voltage.
2012,
24: 797-800.
doi: 10.3788/HPLPB20122404.0797
Abstract:
This paper analyzes the insulation resistance characteristics of metallized film capacitor under high electric field, and shows that the insulation resistance is the comprehensive result of the dielectric leakage and self-healing under high electric field. An insulation resistance testing method is presented for the high metallized film capacitors under high electric field. The test results indicate that, for operating electric field ranging from 300 to 400 V/m, the insulation resistance decreases rapidly with the electric field increasing, but because of the self-healing, the decrease gradually slows down. The insulation resistance follows similar rules within the operating temperature range of 20 to 50 ℃.In addition, the dielectric leakage under high operating temperature and electric field reduces the capacitor voltage significantly, affecting the energy storage efficiency of the capacitors and pulsed power systems.
This paper analyzes the insulation resistance characteristics of metallized film capacitor under high electric field, and shows that the insulation resistance is the comprehensive result of the dielectric leakage and self-healing under high electric field. An insulation resistance testing method is presented for the high metallized film capacitors under high electric field. The test results indicate that, for operating electric field ranging from 300 to 400 V/m, the insulation resistance decreases rapidly with the electric field increasing, but because of the self-healing, the decrease gradually slows down. The insulation resistance follows similar rules within the operating temperature range of 20 to 50 ℃.In addition, the dielectric leakage under high operating temperature and electric field reduces the capacitor voltage significantly, affecting the energy storage efficiency of the capacitors and pulsed power systems.
2012,
24: 801-805.
doi: 10.3788/HPLPB20122404.0801
Abstract:
With a long pulse low impedance electron-beam accelerator, the insulating properties of polytetrafluoroethylene (PTFE)， high molecular weight polyethylene (PE) and polycarbonate (PC) are measured in vacuum by the excitation of ～100 ns high-voltage pulses. The effect of grooved structure on enhancing the insulating strength is studied. Experimental results show that flashover delay time can be promisingly increased with proper grooved structure. PC is the most competent dielectric among the three tested materials, its flashover delay time can be increased by 90% with the grooved structure. The analysis by a scanning electron microscope (SEM) also shows that the surface properties of the materials, such as the roughness and homogeneity, are important in initiating a surface flashover.
With a long pulse low impedance electron-beam accelerator, the insulating properties of polytetrafluoroethylene (PTFE)， high molecular weight polyethylene (PE) and polycarbonate (PC) are measured in vacuum by the excitation of ～100 ns high-voltage pulses. The effect of grooved structure on enhancing the insulating strength is studied. Experimental results show that flashover delay time can be promisingly increased with proper grooved structure. PC is the most competent dielectric among the three tested materials, its flashover delay time can be increased by 90% with the grooved structure. The analysis by a scanning electron microscope (SEM) also shows that the surface properties of the materials, such as the roughness and homogeneity, are important in initiating a surface flashover.
2012,
24: 806-808.
doi: 10.3788/HPLPB20122404.0806
Abstract:
Insulator surface flashover restricts the pulsed power systems development toward high-voltage, high-current direction. This article summarizes the optimal insulator geometry under a few different applications, and analyzes the surface charge of different insulator angles. The results indicate that surface charge plays an important role in the flashover performance of insulator, and it is necessary to carry out experimental studies to find out the optimal insulator geometry according to different applications.
Insulator surface flashover restricts the pulsed power systems development toward high-voltage, high-current direction. This article summarizes the optimal insulator geometry under a few different applications, and analyzes the surface charge of different insulator angles. The results indicate that surface charge plays an important role in the flashover performance of insulator, and it is necessary to carry out experimental studies to find out the optimal insulator geometry according to different applications.
2012,
24: 809-812.
doi: 10.3788/HPLPB20122404.0809
Abstract:
A layer-wound mylar dielectric parallel-plates Blumlein line has been developed, which adopts S-type layer winding and is 50 cm18 cm12 cm in size. Its characteristics of high voltage insulation and output pulse were studied. The results simulated using equivalent circuit by PSpice software, consist with the results of experiment. It is found that the increase in switch inductance reduces the rise time of output pulses and distorts the output pulse shape, introducing reflections as well. The parasitic impedance causes some power loss in the line as the output pulse amplitude is lower than expected. The Blumlein line can deliver a maximum output of 50 kV to a matched load with duration of 180 ns and rise time better than 50 ns.
A layer-wound mylar dielectric parallel-plates Blumlein line has been developed, which adopts S-type layer winding and is 50 cm18 cm12 cm in size. Its characteristics of high voltage insulation and output pulse were studied. The results simulated using equivalent circuit by PSpice software, consist with the results of experiment. It is found that the increase in switch inductance reduces the rise time of output pulses and distorts the output pulse shape, introducing reflections as well. The parasitic impedance causes some power loss in the line as the output pulse amplitude is lower than expected. The Blumlein line can deliver a maximum output of 50 kV to a matched load with duration of 180 ns and rise time better than 50 ns.
2012,
24: 813-817.
doi: 10.3788/HPLPB20122404.0813
Abstract:
This paper analyzes the mechanism of radio frequency(RF) field generation using nonlinear transmission lines. According to soliton theory, relevant parameters affecting the field generation are investigated systematically using circuit simulation method. It is found that an increase in the input voltage, and a decrease in the inductance and nominal capacitance of the nonlinear transmission line can increase the frequency of the generated RF field, and an increase in the nonlinear rate of the capacitance can increase the peak voltage and higher frequency components. A nonlinear transmission line, which can generate an RF field with 0.8 GW power and 19.42 MHz frequency, is then designed.
This paper analyzes the mechanism of radio frequency(RF) field generation using nonlinear transmission lines. According to soliton theory, relevant parameters affecting the field generation are investigated systematically using circuit simulation method. It is found that an increase in the input voltage, and a decrease in the inductance and nominal capacitance of the nonlinear transmission line can increase the frequency of the generated RF field, and an increase in the nonlinear rate of the capacitance can increase the peak voltage and higher frequency components. A nonlinear transmission line, which can generate an RF field with 0.8 GW power and 19.42 MHz frequency, is then designed.
2012,
24: 818-822.
doi: 10.3788/HPLPB20122404.0818
Abstract:
This paper evaluates the power transmission efficiency of long magnetically insulated transmission line(MITL) and the effects caused by MITLs structure changes. The steady state flow theory and unsteady state flow theory of the MITL have been introduced. The particle-in-cell simulation and corresponding validation experiments have been conducted. The formula for Mendels arbitrary momentum model after laminar flow approximation are verified by the accurate measurement of the cathode current, anode current and interpolar voltage in the same position. Experiment results of the inductance-supported MITL show that by using the rigid spring of appropriate inductance to fix the internal and external tubes, the requirements of the transmission efficiency can be satisfied.
This paper evaluates the power transmission efficiency of long magnetically insulated transmission line(MITL) and the effects caused by MITLs structure changes. The steady state flow theory and unsteady state flow theory of the MITL have been introduced. The particle-in-cell simulation and corresponding validation experiments have been conducted. The formula for Mendels arbitrary momentum model after laminar flow approximation are verified by the accurate measurement of the cathode current, anode current and interpolar voltage in the same position. Experiment results of the inductance-supported MITL show that by using the rigid spring of appropriate inductance to fix the internal and external tubes, the requirements of the transmission efficiency can be satisfied.
2012,
24: 823-826.
doi: 10.3788/HPLPB20122404.0823
Abstract:
A triggered vacuum switch (TVS) with multi-rod electrode is proposed. The switch adopts three pairs of rod electrodes to enhance the arcing area of main electrode, leading to the promotion of current carrying ability. The arc development of the switch is discussed under different high currents. The parallel arcing and arc transfer process in the multiple channels are discussed according to the change of arc voltage. The arc transfer process helps to keep the vacuum arc in diffusion mode. In experiments, the TVS sample carried a peak current of 225 kA and realized charge transfer of 45 C in one pulse.
A triggered vacuum switch (TVS) with multi-rod electrode is proposed. The switch adopts three pairs of rod electrodes to enhance the arcing area of main electrode, leading to the promotion of current carrying ability. The arc development of the switch is discussed under different high currents. The parallel arcing and arc transfer process in the multiple channels are discussed according to the change of arc voltage. The arc transfer process helps to keep the vacuum arc in diffusion mode. In experiments, the TVS sample carried a peak current of 225 kA and realized charge transfer of 45 C in one pulse.
2012,
24: 827-830.
doi: 10.3788/HPLPB20122404.0827
Abstract:
A double-coil load of the miniature square-pulsed magnetic field generator was proposed, and an evenly-distributed magnetic field was obtained in the load area. Six-stage pulse-forming network(PFN) with 2 single-stage characteristic impedance was used as the primary energy storage and pulse-forming element, and it could output a square drive current pulse on the impedance-matched load. Furthermore, a field-distorted gas spark switch was developed as the main switch, and it could effectively reduce the action-delay time and its dispersion of the generator. Experimental results show that the peak value of the magnetic field at the center of the load area is 0.04 T, the flat-top time is about 3 s, the top-droop ratio is less than 5%, the risetime (10%-90% peak value of magnetic induction) is less than 0.5 s, and the action-delay-time dispersion of the generator is less than 10 ns (standard deviation).
A double-coil load of the miniature square-pulsed magnetic field generator was proposed, and an evenly-distributed magnetic field was obtained in the load area. Six-stage pulse-forming network(PFN) with 2 single-stage characteristic impedance was used as the primary energy storage and pulse-forming element, and it could output a square drive current pulse on the impedance-matched load. Furthermore, a field-distorted gas spark switch was developed as the main switch, and it could effectively reduce the action-delay time and its dispersion of the generator. Experimental results show that the peak value of the magnetic field at the center of the load area is 0.04 T, the flat-top time is about 3 s, the top-droop ratio is less than 5%, the risetime (10%-90% peak value of magnetic induction) is less than 0.5 s, and the action-delay-time dispersion of the generator is less than 10 ns (standard deviation).
2012,
24: 831-834.
doi: 10.3788/HPLPB20122404.0831
Abstract:
The spark channel generated by focused laser in laser triggered nitrogen gas switch was studied, by using high speed framing camera at the exposure time of 100 ns. Experiments on laser triggered gas switch were carried out with different laser energies and different focuses at the laser wavelength of 266 nm. The delay and jitter of the switch decrease as the laser energy increases. The relationship between the spark channel length of the focused laser and the jitter of the switch was analyzed. When the spark channel length of the focused laser reaches 40% of the total spark gap length, the jitter of the gas switch is at sub-nanosecond level.
The spark channel generated by focused laser in laser triggered nitrogen gas switch was studied, by using high speed framing camera at the exposure time of 100 ns. Experiments on laser triggered gas switch were carried out with different laser energies and different focuses at the laser wavelength of 266 nm. The delay and jitter of the switch decrease as the laser energy increases. The relationship between the spark channel length of the focused laser and the jitter of the switch was analyzed. When the spark channel length of the focused laser reaches 40% of the total spark gap length, the jitter of the gas switch is at sub-nanosecond level.
2012,
24: 835-838.
doi: 10.3788/HPLPB20122404.0835
Abstract:
A gas spark gap for LTD (linear transformer driver) switch is studied. The gap has a configuration of an annular gap and a corona needle in parallel. The V-I characteristics of corona discharge in the gap are tested. The effects of different factors including gas types, gas pressures, corona polarity and configuration of the electrode on the V-I characteristics are investigated. The experimental results show that negative corona in electronegative gas has good stability. An effective way to meet the needs of the switch is adjusting the length of the corona needle or the configuration of the electrode.
A gas spark gap for LTD (linear transformer driver) switch is studied. The gap has a configuration of an annular gap and a corona needle in parallel. The V-I characteristics of corona discharge in the gap are tested. The effects of different factors including gas types, gas pressures, corona polarity and configuration of the electrode on the V-I characteristics are investigated. The experimental results show that negative corona in electronegative gas has good stability. An effective way to meet the needs of the switch is adjusting the length of the corona needle or the configuration of the electrode.
2012,
24: 839-842.
doi: 10.3788/HPLPB20122404.0839
Abstract:
The paper presents the design of laser triggering system for Z-pinch primary test stand(PTS) and the experimental results of single test module. The synchronization of 24 lines of pulsed currents was realized by 24 laser triggered switches, which were triggered by 12 Q-Switched Nd:YAG fourth harmonic laser 266 nm. Each laser output was divided into two beams, and then focused for the trigger of adjacent two switches. The experimental results show that the jitter of the laser triggering system is less than 3 ns, and the that of the switches is less than 5 ns. The experiment conditis are laser energy 100 mJ, pulse width 7 ns, and Marx charged voltage 75 kV. The isolation gas chamber has good insulation and seal. The laser optical system is stable and easy for adjustment. The successful test demonstrates that the design of beam splitting is feasible in PTS.
The paper presents the design of laser triggering system for Z-pinch primary test stand(PTS) and the experimental results of single test module. The synchronization of 24 lines of pulsed currents was realized by 24 laser triggered switches, which were triggered by 12 Q-Switched Nd:YAG fourth harmonic laser 266 nm. Each laser output was divided into two beams, and then focused for the trigger of adjacent two switches. The experimental results show that the jitter of the laser triggering system is less than 3 ns, and the that of the switches is less than 5 ns. The experiment conditis are laser energy 100 mJ, pulse width 7 ns, and Marx charged voltage 75 kV. The isolation gas chamber has good insulation and seal. The laser optical system is stable and easy for adjustment. The successful test demonstrates that the design of beam splitting is feasible in PTS.
2012,
24: 843-846.
doi: 10.3788/HPLPB20122404.0843
Abstract:
A repetitive Tesla pulsed power generator with 1 MV, 40 ns output was used to study the resistances of gas spark switch in nitrogen under high gas pressures and large gap. With the switch regarded as a RLC circuit, the resistances were computed under different gas pressures, by using the measured voltages of the pulse forming line and conical transmission line. The results show that the resistances reduce quickly during the formation of discharging channels, and eventually reach ohm level. Experimental and theoretical results are compared to check if Barannik, Rompe-Weizel, Demenik, Toeple, and Vlastos formulas are applicable. The Rompe-Weizel formula is modified by changing a different constant.
A repetitive Tesla pulsed power generator with 1 MV, 40 ns output was used to study the resistances of gas spark switch in nitrogen under high gas pressures and large gap. With the switch regarded as a RLC circuit, the resistances were computed under different gas pressures, by using the measured voltages of the pulse forming line and conical transmission line. The results show that the resistances reduce quickly during the formation of discharging channels, and eventually reach ohm level. Experimental and theoretical results are compared to check if Barannik, Rompe-Weizel, Demenik, Toeple, and Vlastos formulas are applicable. The Rompe-Weizel formula is modified by changing a different constant.
2012,
24: 847-850.
doi: 10.3788/HPLPB20122404.0847
Abstract:
Based on the conduction mechanism of triggered vacuum switch, an experimental scheme has been designed to study the switching-on process of long-gap triggered vacuum switch. The switching-on process and the factors affecting the conduction characteristics of long-gap vacuum triggered switch were discussed according to photos gained from high speed camera. The factors affecting the spark resistance of triggered vacuum switch were analyzed. The experimental results confirm that the spark resistance and delay time can be effectively reduced by shortening the rise time and enhancing the peak of trigger current. The conclusion of experiment verifies theoretical predictions.
Based on the conduction mechanism of triggered vacuum switch, an experimental scheme has been designed to study the switching-on process of long-gap triggered vacuum switch. The switching-on process and the factors affecting the conduction characteristics of long-gap vacuum triggered switch were discussed according to photos gained from high speed camera. The factors affecting the spark resistance of triggered vacuum switch were analyzed. The experimental results confirm that the spark resistance and delay time can be effectively reduced by shortening the rise time and enhancing the peak of trigger current. The conclusion of experiment verifies theoretical predictions.
2012,
24: 851-854.
doi: 10.3788/HPLPB20122404.0851
Abstract:
A cascade gaps circuit model for Rimfire switch has been developed. The circuit model includes all stray capacitances and spark channel on-state conduction characteristics. It can not only describe the behavior of Rimfire switch, but also allow analysis of the time varying voltage and current at each part of the switch, describing the internal characters of the switch. PSpice was used to implement the cascade gaps circuit model and simulate a 700 kV Rimfire switch. The simulation shows that, the voltage of the whole switch will be higher than 700 kV when the laser triggered section has broken down but all cascade gaps keep dielectric, and the voltage of all gaps attenuates with high frequency oscillation.
A cascade gaps circuit model for Rimfire switch has been developed. The circuit model includes all stray capacitances and spark channel on-state conduction characteristics. It can not only describe the behavior of Rimfire switch, but also allow analysis of the time varying voltage and current at each part of the switch, describing the internal characters of the switch. PSpice was used to implement the cascade gaps circuit model and simulate a 700 kV Rimfire switch. The simulation shows that, the voltage of the whole switch will be higher than 700 kV when the laser triggered section has broken down but all cascade gaps keep dielectric, and the voltage of all gaps attenuates with high frequency oscillation.
2012,
24: 855-858.
doi: 10.3788/HPLPB20122404.0855
Abstract:
A three-dimensional model ofvacuum interrupter was established to study the arc motion in longitudinal magnetic field. The PIC method was used to simulate the electromagnetic field and arc motion for 12 kV vacuum interrupter. Structural parameters of contact gaps, shield diameters and grooving width of the contact cup were changed to study the influence of the electromagnetic field distribution in vacuum interrupter. Variation of arc motion with time and arc distribution on contact surface under different contact distances were simulated. The results show that the proper placement of shield in vacuum interrupter can effectively improve the electric field distribution, and the grooving width of the contact cup can influence the magnetic field distribution. When the grooving width increases, the magnetic field strength increases.
A three-dimensional model ofvacuum interrupter was established to study the arc motion in longitudinal magnetic field. The PIC method was used to simulate the electromagnetic field and arc motion for 12 kV vacuum interrupter. Structural parameters of contact gaps, shield diameters and grooving width of the contact cup were changed to study the influence of the electromagnetic field distribution in vacuum interrupter. Variation of arc motion with time and arc distribution on contact surface under different contact distances were simulated. The results show that the proper placement of shield in vacuum interrupter can effectively improve the electric field distribution, and the grooving width of the contact cup can influence the magnetic field distribution. When the grooving width increases, the magnetic field strength increases.
2012,
24: 859-862.
doi: 10.3788/HPLPB20122404.0859
Abstract:
On the basis of one-order magnetic pulse compression (MPC) circuit, this paper puts forward a method for measuring the dynamic magnetic properies of magnetic cores in actual working conditions, such as hysteresis loop and saturated permeability, to guide the design of magnetic switches in the MPC experiments. Widely used amorphous magnetic cores and nanocrystalline magnetic cores are chosen, and their dynamic magnetic parameters are obtained. The magnetic switch using domestic amorphous magnetic core, outputs pulses of 73 ns risetime, 28.3 kV peak voltage, and 503 ns full width at half maximum. The magnetic switch using Japanese nanocrystalline magnetic core outputs pulses of 30 ns risetime, 28.4 kV peak voltage, and 193 ns full width at half maximum.
On the basis of one-order magnetic pulse compression (MPC) circuit, this paper puts forward a method for measuring the dynamic magnetic properies of magnetic cores in actual working conditions, such as hysteresis loop and saturated permeability, to guide the design of magnetic switches in the MPC experiments. Widely used amorphous magnetic cores and nanocrystalline magnetic cores are chosen, and their dynamic magnetic parameters are obtained. The magnetic switch using domestic amorphous magnetic core, outputs pulses of 73 ns risetime, 28.3 kV peak voltage, and 503 ns full width at half maximum. The magnetic switch using Japanese nanocrystalline magnetic core outputs pulses of 30 ns risetime, 28.4 kV peak voltage, and 193 ns full width at half maximum.
2012,
24: 863-867.
doi: 10.3788/HPLPB20122404.0863
Abstract:
In order to improve the performance of solid-state accelerator, the characteristics of magnetic switch used as the main switch of the accelerator have been investigated. The volume of magnetic core, the loss, and saturated inductance of the magnetic switch have been derived. The results show that the spacing factor of the magnetic switch reaches the peak when the height of the magnetic materials is 0.05 m for selected magnetic cores. The saturated inductance of the windings changes slowly when the average magnetic path length of the core is greater than 1 m. The physical process of saturation in the cores was analyzed by using saturation-wave theory. The rise-time factor of the output pulse was derived. The thickness, resistivity and magnetic path length difference of the magnetic core are shown to be key parameters affecting the rise-time factor.
In order to improve the performance of solid-state accelerator, the characteristics of magnetic switch used as the main switch of the accelerator have been investigated. The volume of magnetic core, the loss, and saturated inductance of the magnetic switch have been derived. The results show that the spacing factor of the magnetic switch reaches the peak when the height of the magnetic materials is 0.05 m for selected magnetic cores. The saturated inductance of the windings changes slowly when the average magnetic path length of the core is greater than 1 m. The physical process of saturation in the cores was analyzed by using saturation-wave theory. The rise-time factor of the output pulse was derived. The thickness, resistivity and magnetic path length difference of the magnetic core are shown to be key parameters affecting the rise-time factor.
2012,
24: 868-870.
doi: 10.3788/HPLPB20122404.0868
Abstract:
A three-electrode solid field distortion switch is designed based on over-voltage self-breakdown of semiconductor arrester. It includes two semiconductor arrester arrays and a trigger electrode. When it is triggered, self-breakdown of arrester arrays happens consecutively to turn on the switch. The voltage of load is tested by experiment with specific trigger pulse. It is preliminarily proved that the design of the solid semiconductor arrester field distortion three-electrode discharge switch is feasible. Meantime, measures for decreasing transient inductance, and increasing current, operating voltage and repetitive rate are discussed.
A three-electrode solid field distortion switch is designed based on over-voltage self-breakdown of semiconductor arrester. It includes two semiconductor arrester arrays and a trigger electrode. When it is triggered, self-breakdown of arrester arrays happens consecutively to turn on the switch. The voltage of load is tested by experiment with specific trigger pulse. It is preliminarily proved that the design of the solid semiconductor arrester field distortion three-electrode discharge switch is feasible. Meantime, measures for decreasing transient inductance, and increasing current, operating voltage and repetitive rate are discussed.
2012,
24: 871-875.
doi: 10.3788/HPLPB20122404.0871
Abstract:
A conceptional design of synchronized trigger system for fast linear transformer driver is introduced, which is based on single transmission line pulse forming technology. The trigger system consists of Marx generators, pulse forming lines, main switches, transmission lines and output cables. With the equivalent circuit model, the relationship between Marx generator and pulse forming line is studied. As multiple pulse forming lines are driven in parallel, the number of trigger pulses and the efficiency of energy increase, but the amplitude of trigger pulses decreases. The impedance and number of output cables affect the amplitude as well. With increasing the serial number of cable, the amplitude decreases gradually. As the impedance of transmission line increases, the attenuation rate of the amplitude increases consequently. The trigger signals can be also led out at the same time. When single pulse forming line is driven, 60 trigger signals can be obtained with the amplitude of 293 kV and rise time of 11 ns. When five pulse forming lines are driven in parallel, the amount of trigger signals can reach 300, the amplitude of the pulses is 151 kV and the rise time is 11 ns.
A conceptional design of synchronized trigger system for fast linear transformer driver is introduced, which is based on single transmission line pulse forming technology. The trigger system consists of Marx generators, pulse forming lines, main switches, transmission lines and output cables. With the equivalent circuit model, the relationship between Marx generator and pulse forming line is studied. As multiple pulse forming lines are driven in parallel, the number of trigger pulses and the efficiency of energy increase, but the amplitude of trigger pulses decreases. The impedance and number of output cables affect the amplitude as well. With increasing the serial number of cable, the amplitude decreases gradually. As the impedance of transmission line increases, the attenuation rate of the amplitude increases consequently. The trigger signals can be also led out at the same time. When single pulse forming line is driven, 60 trigger signals can be obtained with the amplitude of 293 kV and rise time of 11 ns. When five pulse forming lines are driven in parallel, the amount of trigger signals can reach 300, the amplitude of the pulses is 151 kV and the rise time is 11 ns.
2012,
24: 876-880.
doi: 10.3788/HPLPB20122404.0876
Abstract:
The structure of the semiconductor pulsed power switch reversely switched dynistor(RSD) was optimized by introducing a buffer layer. The numerical model of the device was established and the simulation analysis was carried out. The results show that the buffer layer can stop the electrical field and make the base region thinner. Compared with the conventional structure, the buffer layer structure improves the blocking voltage and reduces the turn-on voltage. According to the special operating mode of the RSD, the schemes for measuring key parameters such as the turn-on voltage and turn-off time for the RSD were proposed and realized by experiment. The high current turn-on experiments were carried out, in which the peak current of 173 kA passed through the RSD assembly of 7.6 cm diameter successfully. The main discharge voltage was 12 kV and the transferred electric charge amount was 32 C.
The structure of the semiconductor pulsed power switch reversely switched dynistor(RSD) was optimized by introducing a buffer layer. The numerical model of the device was established and the simulation analysis was carried out. The results show that the buffer layer can stop the electrical field and make the base region thinner. Compared with the conventional structure, the buffer layer structure improves the blocking voltage and reduces the turn-on voltage. According to the special operating mode of the RSD, the schemes for measuring key parameters such as the turn-on voltage and turn-off time for the RSD were proposed and realized by experiment. The high current turn-on experiments were carried out, in which the peak current of 173 kA passed through the RSD assembly of 7.6 cm diameter successfully. The main discharge voltage was 12 kV and the transferred electric charge amount was 32 C.
2012,
24: 881-884.
doi: 10.3788/HPLPB20122404.0881
Abstract:
A three-electrode field-distortion gas spark switch used for linear transformer drivers is developed. The switch is insulated with SF6. The size and inductance of the switch is small, and the inductance is experimentally estimated at about 67 nH. When the switch operates at 80 kV charge voltage and 70% working ratio, the delay and jitter are 40.0 ns and 2.8 ns respectively. The influences of the electrode material on the self-breakdown and trigger characteristics are experimentally studied. The results indicate that the switch of CuW electrode has a smaller deviation of self-breakdown voltage, smaller delay and jitter, lower prefire probability and less electrode erosion than that of copper electrode. Compared with copper material, CuW material may be more suitable for three-electrode field-distortion gas switch.
A three-electrode field-distortion gas spark switch used for linear transformer drivers is developed. The switch is insulated with SF6. The size and inductance of the switch is small, and the inductance is experimentally estimated at about 67 nH. When the switch operates at 80 kV charge voltage and 70% working ratio, the delay and jitter are 40.0 ns and 2.8 ns respectively. The influences of the electrode material on the self-breakdown and trigger characteristics are experimentally studied. The results indicate that the switch of CuW electrode has a smaller deviation of self-breakdown voltage, smaller delay and jitter, lower prefire probability and less electrode erosion than that of copper electrode. Compared with copper material, CuW material may be more suitable for three-electrode field-distortion gas switch.
2012,
24: 885-888.
doi: 10.3788/HPLPB20122404.0885
Abstract:
The paper introduces a multiple-switch pulsed power technique based on transmission-line-transformer(TLT). In the circuit, all the switches can synchronize themselves automatically like in a Marx generator. At the output side, a variety of output options are possible, i.e. high voltage, large current or independent pulse generation. A 10-switch, 10-stage TLT system has been successfully developed. At the moment, a repetitive large power generator with 20 switches and a 20-stage TLT is being developed. The peak power is estimated to be 5 GW with a total energy of 140 J per shot. Up to 100 independent simultaneous outputs can be realized with the same peak power.
The paper introduces a multiple-switch pulsed power technique based on transmission-line-transformer(TLT). In the circuit, all the switches can synchronize themselves automatically like in a Marx generator. At the output side, a variety of output options are possible, i.e. high voltage, large current or independent pulse generation. A 10-switch, 10-stage TLT system has been successfully developed. At the moment, a repetitive large power generator with 20 switches and a 20-stage TLT is being developed. The peak power is estimated to be 5 GW with a total energy of 140 J per shot. Up to 100 independent simultaneous outputs can be realized with the same peak power.
2012,
24: 889-892.
doi: 10.3788/HPLPB20122404.0889
Abstract:
The paper presents an all-solid-state high-repetition-rate pulse generator with adjustable output amplitude based on magnetic pulse compression (MPC) technique. The pulse compression network makes use of commercially available IGBTs switching a capacitor bank into a metglas transformer together with a voltage doubling circuit. The capacitor bank is charged to 500 V by a resonant LC charger, and also switched by a commercial diode. The output of the pulse generator is controlled by the gate voltage of the IGBTs. Pulses with a width of 70 ns can be generated with repetition rates up to 5 kHz. The amplitude can be controlled from 4 kV to 40 kV into a 500 load. Equivalent circuits for the final operation stage of the compressor accounting for pre-pulse in magnetic switches are presented and analyzed, and the pre-pulse generation process of the MPC system is discussed. Simulation results show that, increasing the unsaturated inductance of the magnetic switch and reducing the load resistance enhance the pre-pulse peak. Thus to diminish the pre-pulse, a better ferrite core with higher permeability should be considered.
The paper presents an all-solid-state high-repetition-rate pulse generator with adjustable output amplitude based on magnetic pulse compression (MPC) technique. The pulse compression network makes use of commercially available IGBTs switching a capacitor bank into a metglas transformer together with a voltage doubling circuit. The capacitor bank is charged to 500 V by a resonant LC charger, and also switched by a commercial diode. The output of the pulse generator is controlled by the gate voltage of the IGBTs. Pulses with a width of 70 ns can be generated with repetition rates up to 5 kHz. The amplitude can be controlled from 4 kV to 40 kV into a 500 load. Equivalent circuits for the final operation stage of the compressor accounting for pre-pulse in magnetic switches are presented and analyzed, and the pre-pulse generation process of the MPC system is discussed. Simulation results show that, increasing the unsaturated inductance of the magnetic switch and reducing the load resistance enhance the pre-pulse peak. Thus to diminish the pre-pulse, a better ferrite core with higher permeability should be considered.
2012,
24: 893-897.
doi: 10.3788/HPLPB20122404.0893
Abstract:
A module is designed for six-channel, three-pulsed power system, which is based on MHz repeating rate burst high-voltage generator with Blumlein pulse forming lines connected in parallel and multi-stage laser trigger system. Parameter designs of Blumlein device, confluent/blocking network, and trigger system are discussed. The analysis and simulation results show that the module meets the needs of multi-pulse linear induction accelerator. The output voltage with an amplitude of 300 kV, time jitter less than 2.3 ns （standard deviation）, and repetition rate of MHz level is achieved.
A module is designed for six-channel, three-pulsed power system, which is based on MHz repeating rate burst high-voltage generator with Blumlein pulse forming lines connected in parallel and multi-stage laser trigger system. Parameter designs of Blumlein device, confluent/blocking network, and trigger system are discussed. The analysis and simulation results show that the module meets the needs of multi-pulse linear induction accelerator. The output voltage with an amplitude of 300 kV, time jitter less than 2.3 ns （standard deviation）, and repetition rate of MHz level is achieved.
2012,
24: 898-902.
doi: 10.3788/HPLPB20122404.0898
Abstract:
The application of nanosecond discharge plasma in many fields depends greatly on a compact repetitive high voltage nanosecond pulse generator. In this paper, a compact high voltage nanosecond pulse generator is presented, which is constructed with all-solid-state components. The pulse generator consists of DC module, insulated-gate bipolar transistors and its drivers, saturable pulse transformer and magnetic switch and so on. Simulation analysis and experimental investigation show that, the pulse generator can output pulsed voltage of 40 kV with duration about 100 ns and rise-time of 50 ns. Its highest repetitive frequency can be up to 5 kHz.
The application of nanosecond discharge plasma in many fields depends greatly on a compact repetitive high voltage nanosecond pulse generator. In this paper, a compact high voltage nanosecond pulse generator is presented, which is constructed with all-solid-state components. The pulse generator consists of DC module, insulated-gate bipolar transistors and its drivers, saturable pulse transformer and magnetic switch and so on. Simulation analysis and experimental investigation show that, the pulse generator can output pulsed voltage of 40 kV with duration about 100 ns and rise-time of 50 ns. Its highest repetitive frequency can be up to 5 kHz.
2012,
24: 903-906.
doi: 10.3788/HPLPB20122404.0903
Abstract:
A 4 MV pre-triggered Marx generator charged with series resistor has been designed. For better output property, higher charging efficiency and lower voltage difference between stages, the study conducted the analytic solution of its constant current charging(CCC) process. The results show that the voltage differences between different stages of capacitors are in direct proportion to charging resistance, capacitance and charging speed, but are independent of time; the charging efficiency improves with time, but is independent of charging speed. This solution is universally applicable. The simulation of CCC for the Marx generator was done with PSpice, and the results agree with the analytic solution. Compared to constant voltage charging(CVC) with 10 k charging resistors, 400 nF capacitors and 10 kV/s charging speed, the charging efficiency of CCC is 90%, double that of CVC and the charging time is only one third.
A 4 MV pre-triggered Marx generator charged with series resistor has been designed. For better output property, higher charging efficiency and lower voltage difference between stages, the study conducted the analytic solution of its constant current charging(CCC) process. The results show that the voltage differences between different stages of capacitors are in direct proportion to charging resistance, capacitance and charging speed, but are independent of time; the charging efficiency improves with time, but is independent of charging speed. This solution is universally applicable. The simulation of CCC for the Marx generator was done with PSpice, and the results agree with the analytic solution. Compared to constant voltage charging(CVC) with 10 k charging resistors, 400 nF capacitors and 10 kV/s charging speed, the charging efficiency of CCC is 90%, double that of CVC and the charging time is only one third.
2012,
24: 907-911.
doi: 10.3788/HPLPB20122404.0907
Abstract:
A 10-staged low impedance Marx generator is designed, whose volume is about 0.12 m3and output power and pulse width are more than 30 GW and 200 ns respectively. The generator is charged synchronously with positive and negative voltages. Its triggered-switch is three-electrode field distortion switch and other switches are over-voltage self-breakdown switches. After optimization design, the heights of the self-breakdown switches and triggered switch are 33, and 42 mm, and the inductances are 13.1 and 15.2 nH, respectively. The positive and negative electrodes of the capacitor are pressed within the switchs electrode. The insulator within the electrodes of the capacitor is 100 layers of polypropylene, whose total thickness is about 2 mm, and it can withstand the voltage over 100 kV. Additionally, Maxwell 3D simulation shows that, the inductance of the electrode connecting plate is 4.16 nH in this connection manner. Pspice circuit simulation and experiment show that, when the capacitor is charged at 100 kV, the output voltage is more than 600 kV and the peak current is about 50 kA on the 12 load.
A 10-staged low impedance Marx generator is designed, whose volume is about 0.12 m3and output power and pulse width are more than 30 GW and 200 ns respectively. The generator is charged synchronously with positive and negative voltages. Its triggered-switch is three-electrode field distortion switch and other switches are over-voltage self-breakdown switches. After optimization design, the heights of the self-breakdown switches and triggered switch are 33, and 42 mm, and the inductances are 13.1 and 15.2 nH, respectively. The positive and negative electrodes of the capacitor are pressed within the switchs electrode. The insulator within the electrodes of the capacitor is 100 layers of polypropylene, whose total thickness is about 2 mm, and it can withstand the voltage over 100 kV. Additionally, Maxwell 3D simulation shows that, the inductance of the electrode connecting plate is 4.16 nH in this connection manner. Pspice circuit simulation and experiment show that, when the capacitor is charged at 100 kV, the output voltage is more than 600 kV and the peak current is about 50 kA on the 12 load.
2012,
24: 912-916.
doi: 10.3788/HPLPB20122404.0912
Abstract:
The T2LXD1 series armatures are designed and the elastic-plastic characteristics of 6061-T6 aluminum alloy is analyzed. The linear hardening theory, the contact theory, and the ANSYS software are utilized to build the elastic-plastic contact model of the T2LXD1 series armatures. The effect of tail structure on the contact condition of the armature-rail interface is studied, and the interfacial contact parameters such as contact force, contact area and contact pressure are simulated and obtained. A small electromagnetic drive system with short rail is used to execute the sliding experiments of T2LXD1 series armatures, and the interfacial wearing conditions are analyzed, the sliding performance curves are compared. The experimental results show that the long tail can not increase the effective contact area, and longer tail results in lower sliding performance of armature; for the flexible armature and the short rail, to properly increase the pretightening contact force of the interface can improve the sliding performance of armature.
The T2LXD1 series armatures are designed and the elastic-plastic characteristics of 6061-T6 aluminum alloy is analyzed. The linear hardening theory, the contact theory, and the ANSYS software are utilized to build the elastic-plastic contact model of the T2LXD1 series armatures. The effect of tail structure on the contact condition of the armature-rail interface is studied, and the interfacial contact parameters such as contact force, contact area and contact pressure are simulated and obtained. A small electromagnetic drive system with short rail is used to execute the sliding experiments of T2LXD1 series armatures, and the interfacial wearing conditions are analyzed, the sliding performance curves are compared. The experimental results show that the long tail can not increase the effective contact area, and longer tail results in lower sliding performance of armature; for the flexible armature and the short rail, to properly increase the pretightening contact force of the interface can improve the sliding performance of armature.
2012,
24: 917-920.
doi: 10.3788/HPLPB20122404.0917
Abstract:
The paper presents a 500 kV all-solid-state Marx generator adopting Z type circuit. In this Marx generator, twenty-eight isolated gate bipolar transistors(IGBTs) serial modules were utilized as controllable switches. The maximum working voltage of each IGBTs module was designed as high as 22 kV, and the repetition rate at full load condition was designed as high as 200 Hz. The ladder type pulse-forming-networks (PFNs) consisting of metallized film capacitors and wound inductors were designed as energy storage and pulse-forming components. Experimental results indicate that the Marx generator can run with 500 kV output pulses and 50 Hz repetition rate in burst-mode successfully. The peak output pulse voltage of this generator is approximately the same as that of the highest power solid-state Marx generator ever reported, while the peak current is almost twice, reaching 1 000 A. The maximum pulsed power of the generator is as high as 500 MW. As the capacitors were utilized as energy storage components, the pulse width can be adjusted from 3 s to 10 s continuously.
The paper presents a 500 kV all-solid-state Marx generator adopting Z type circuit. In this Marx generator, twenty-eight isolated gate bipolar transistors(IGBTs) serial modules were utilized as controllable switches. The maximum working voltage of each IGBTs module was designed as high as 22 kV, and the repetition rate at full load condition was designed as high as 200 Hz. The ladder type pulse-forming-networks (PFNs) consisting of metallized film capacitors and wound inductors were designed as energy storage and pulse-forming components. Experimental results indicate that the Marx generator can run with 500 kV output pulses and 50 Hz repetition rate in burst-mode successfully. The peak output pulse voltage of this generator is approximately the same as that of the highest power solid-state Marx generator ever reported, while the peak current is almost twice, reaching 1 000 A. The maximum pulsed power of the generator is as high as 500 MW. As the capacitors were utilized as energy storage components, the pulse width can be adjusted from 3 s to 10 s continuously.
2012,
24: 921-924.
doi: 10.3788/HPLPB20122404.0921
Abstract:
A capacitor-discharge seed source based on PIC has been designed. In order to charge quickly and effectively, the source takes advantage of the single-chip pulse width modulation(PWM) to produce a linear voltage which can control the DC-DC converter. And as a result, the charging process is constant current charging. Theoretical and experimental results show that, the seed sourcd charges fast, and reduces the loss during charging. Its energy storage is 250 J, and its volume is smaller than 3 000 cm3.The seed source, which can produce more than 2 kA initial current on 100 H load in calculation, meets the test requirements.
A capacitor-discharge seed source based on PIC has been designed. In order to charge quickly and effectively, the source takes advantage of the single-chip pulse width modulation(PWM) to produce a linear voltage which can control the DC-DC converter. And as a result, the charging process is constant current charging. Theoretical and experimental results show that, the seed sourcd charges fast, and reduces the loss during charging. Its energy storage is 250 J, and its volume is smaller than 3 000 cm3.The seed source, which can produce more than 2 kA initial current on 100 H load in calculation, meets the test requirements.
2012,
24: 925-928.
doi: 10.3788/HPLPB20122404.0925
Abstract:
According to the function and technical requirements of the sampling resistor in high power constant-current source, the design principle, structure and choice criterion of the material of the resistor were systematically summarized based on theoretical analysis and experimental study.The water-cooling and self-cooling sampling resistors of high power constant-current source were compared. The experimental results show that, compared with water-cooling sampling resistor, the self-cooling sampling resistor improves the stability of the constant-current source three times, reduces the ripple coefficient from 2.7% to 2.3%, and increases the life time more than 10 times.
According to the function and technical requirements of the sampling resistor in high power constant-current source, the design principle, structure and choice criterion of the material of the resistor were systematically summarized based on theoretical analysis and experimental study.The water-cooling and self-cooling sampling resistors of high power constant-current source were compared. The experimental results show that, compared with water-cooling sampling resistor, the self-cooling sampling resistor improves the stability of the constant-current source three times, reduces the ripple coefficient from 2.7% to 2.3%, and increases the life time more than 10 times.
2012,
24: 929-932.
doi: 10.3788/HPLPB20122404.0929
Abstract:
To study the characteristics of gaseous discharge in spark gap, a high-voltage pulsed power supply has been designed, whose output voltage amplitude varies between 30～100 kV and repetition-rate varies from 1 Hz to 5 kHz. Adopting the principle of resonant charging, the power supply transferrs energy from the primary source of 10 kV DC to the secondary energy storage capacitor that is charged at least to 18 kV. The hydrogen thyratron conducts under the action of optical trigger signal. Then the current flows through the pulse transformer, and voltage is raises from 18 kV to the peak of 100 kV. The output pulse is a negative voltage one with an pulse width of at least 200ns. and fall timeless than 90 ns. The device can continuously run for no less than 1 min without additional cooling system.
To study the characteristics of gaseous discharge in spark gap, a high-voltage pulsed power supply has been designed, whose output voltage amplitude varies between 30～100 kV and repetition-rate varies from 1 Hz to 5 kHz. Adopting the principle of resonant charging, the power supply transferrs energy from the primary source of 10 kV DC to the secondary energy storage capacitor that is charged at least to 18 kV. The hydrogen thyratron conducts under the action of optical trigger signal. Then the current flows through the pulse transformer, and voltage is raises from 18 kV to the peak of 100 kV. The output pulse is a negative voltage one with an pulse width of at least 200ns. and fall timeless than 90 ns. The device can continuously run for no less than 1 min without additional cooling system.
2012,
24: 933-937.
doi: 10.3788/HPLPB20122404.0933
Abstract:
The conceptional design of a 20 MA/300 ns direct-driven Z-pinch pulsed power driver based on Marx generators is presented, whose Marx generators are composed of 18 series stages with 1.0 F/40 nH/100 kV metal shell capacitors and 100 kV/200 kA gas switches. It has 40 modules in parallel, each consisting of six Marx generators driving the water transmission line. The circuit scheme of the pulsed power driver is set up through PSpice, and the influences of Marx erection jitter on load current are simulated. The results show that when the Marx erection jitter increases from 10 to 60 ns, the peak current varies from 21.0 to 19.8 MA, the rise time of load current is from 282 to 289 ns, while the full width at half magnitude of load current is constant. The effects of the Marx erection jitter on load current declines with the number of Marx generators in parallel increasing. The load current is of maximum when the water line impedance is 4.2 .
The conceptional design of a 20 MA/300 ns direct-driven Z-pinch pulsed power driver based on Marx generators is presented, whose Marx generators are composed of 18 series stages with 1.0 F/40 nH/100 kV metal shell capacitors and 100 kV/200 kA gas switches. It has 40 modules in parallel, each consisting of six Marx generators driving the water transmission line. The circuit scheme of the pulsed power driver is set up through PSpice, and the influences of Marx erection jitter on load current are simulated. The results show that when the Marx erection jitter increases from 10 to 60 ns, the peak current varies from 21.0 to 19.8 MA, the rise time of load current is from 282 to 289 ns, while the full width at half magnitude of load current is constant. The effects of the Marx erection jitter on load current declines with the number of Marx generators in parallel increasing. The load current is of maximum when the water line impedance is 4.2 .
2012,
24: 938-942.
doi: 10.3788/HPLPB20122404.0938
Abstract:
The paper presents a resonant triggering circuit for driving high power reverse switched-on dynistor(RSD). The method for choosing the circuit parameters by optimization is proposed. Owing to the deformation and tail current of the current pulse found in preliminary experiments, two improvements are adopted in the circuit. First, a saturable magnetic ring is utilized to smooth the triggering current pulse. Secondly, a resistance-capacitor circuit is designed to absorb the redundant triggering energy. The experimental results show that the improved triggering circuit can offer an excellent triggering and conducting characteristic for RSD devices.
The paper presents a resonant triggering circuit for driving high power reverse switched-on dynistor(RSD). The method for choosing the circuit parameters by optimization is proposed. Owing to the deformation and tail current of the current pulse found in preliminary experiments, two improvements are adopted in the circuit. First, a saturable magnetic ring is utilized to smooth the triggering current pulse. Secondly, a resistance-capacitor circuit is designed to absorb the redundant triggering energy. The experimental results show that the improved triggering circuit can offer an excellent triggering and conducting characteristic for RSD devices.
2012,
24: 943-948.
doi: 10.3788/HPLPB20122404.0943
Abstract:
To meet the miniaturization requirement, a high-voltage capacitor charging power supply(CCPS) has been developed, which is capable of charging the load capacitor at 20 kJ/s at 10 kV output voltage (40 kW peak power output at 10 kV rated output voltage). This power supply adopts series resonant, soft-switching technology and thermal design technology and its switching frequency is 50 kHz. The effect of distributed parameters on its operating characteristic was analyzed, and solutions were introduced according to different distributed parameters. The thermal design technology of high power density CCPS at interval working condition was experimentally studied. On the basis of the interval operating and thermal characteristics of sub-units in the CCPS, cooling measures were designed and experimental results of temperature rise were provided. The performance test of the CCPS has been carried out, and it demonstrates that, with above designs the power density of the CCPS reaches 2.15 MW/m3.
To meet the miniaturization requirement, a high-voltage capacitor charging power supply(CCPS) has been developed, which is capable of charging the load capacitor at 20 kJ/s at 10 kV output voltage (40 kW peak power output at 10 kV rated output voltage). This power supply adopts series resonant, soft-switching technology and thermal design technology and its switching frequency is 50 kHz. The effect of distributed parameters on its operating characteristic was analyzed, and solutions were introduced according to different distributed parameters. The thermal design technology of high power density CCPS at interval working condition was experimentally studied. On the basis of the interval operating and thermal characteristics of sub-units in the CCPS, cooling measures were designed and experimental results of temperature rise were provided. The performance test of the CCPS has been carried out, and it demonstrates that, with above designs the power density of the CCPS reaches 2.15 MW/m3.
2012,
24: 949-952.
doi: 10.3788/HPLPB20122404.0949
Abstract:
The paper presents the experimental layout and characteristics of the beam parameters measurement system for linear induction accelerator, and analyzes the change of electronic components of the system under high energy electron irradiation. Then measures for component protection are investigated. Under strong electromagnetic interference, high speed signals can be transmitted by using fiber to maintain the narrow pulse with small delay and jitter; using compact embedded computing can improve the systems immunity to electromagnetic interference.
The paper presents the experimental layout and characteristics of the beam parameters measurement system for linear induction accelerator, and analyzes the change of electronic components of the system under high energy electron irradiation. Then measures for component protection are investigated. Under strong electromagnetic interference, high speed signals can be transmitted by using fiber to maintain the narrow pulse with small delay and jitter; using compact embedded computing can improve the systems immunity to electromagnetic interference.
2012,
24: 953-956.
doi: 10.3788/HPLPB20122404.0953
Abstract:
By using the existing beam transport and detection system, experiments have been done to detect the change in the single-pulse linear induction accelerator (LIA) beam transport when the pulse symmetrical feed-in mode changes into the unilateral one. The results show that changing the pulse feed-in mode will result in a visible offset of the electron beam center in the acceleration segment. Thus the pulse unilateral feed-in mode will markedly disturb the beam transport of an LIA. According to the experiments, some feasible improvements have been advanced about the pulse feed-in mode for the double-pulse LIA based on cable delay.
By using the existing beam transport and detection system, experiments have been done to detect the change in the single-pulse linear induction accelerator (LIA) beam transport when the pulse symmetrical feed-in mode changes into the unilateral one. The results show that changing the pulse feed-in mode will result in a visible offset of the electron beam center in the acceleration segment. Thus the pulse unilateral feed-in mode will markedly disturb the beam transport of an LIA. According to the experiments, some feasible improvements have been advanced about the pulse feed-in mode for the double-pulse LIA based on cable delay.
2012,
24: 957-962.
doi: 10.3788/HPLPB20122404.0957
Abstract:
Experiments on intense pulsed emission properties of velvet and carbon nanotube(CNT) cathodes have been conducted on the 2 MeV linear induction accelerator injector. The results show that both velvet and CNT cathodes can emit intense electron beams under 1.61 MV pulsed high voltage. The intense electron beams are 84 and 108 A/cm2 in current density, and the turn-on times are 21 and 40 ns, respectively. The amount of outgassing is estimated to be 0.29 and 0.91 PaL, and the number of outgassed molecules per electron is 64 and 225 for velvet and CNT cathodes. Moreover, the outgassing mass spectra of the two cathodes are similar to each other.
Experiments on intense pulsed emission properties of velvet and carbon nanotube(CNT) cathodes have been conducted on the 2 MeV linear induction accelerator injector. The results show that both velvet and CNT cathodes can emit intense electron beams under 1.61 MV pulsed high voltage. The intense electron beams are 84 and 108 A/cm2 in current density, and the turn-on times are 21 and 40 ns, respectively. The amount of outgassing is estimated to be 0.29 and 0.91 PaL, and the number of outgassed molecules per electron is 64 and 225 for velvet and CNT cathodes. Moreover, the outgassing mass spectra of the two cathodes are similar to each other.
2012,
24: 963-967.
doi: 10.3788/HPLPB20122404.0963
Abstract:
A nanosecond time-resolved imaging platform for diode plasmas diagnostics has been constructed based on the pulsed electron beam accelerator and high speed framing camera(HSFC). The accelerator can provide an electrical pulse with voltages of 200-500 kV, rise-time (from 10% to 90% amplitude) of 25 ns and duration of 110 ns. The diode currents up to kA level can be extracted. The trigger signal for camera was picked up by a water-resistor voltage divider after the main switch of the accelerator, which could avoid the disadvantageous influence of the time jitter caused by the breakdown of the gas gaps. Then the sampled negative electrical pulse was converted into a transistor-transistor logic(TTL) signal (5 V) with rise time of about 1.5 ns and time jitter less than 1 ns via a processor. And this signal was taken as the synchronization time base. According to the working characteristics of the camera, the synchronization scheme relying mainly on electrical pulse delay method supplemented by light signal delay method was determined to make sure that the camera can work synchronously with the light production and transportation from the diode plasmas within the time scale of nanosecond. Moreover, shielding and filtering methods were used to restrain the interference on the measurement system from the accelerator. Finally, time resolved 2-D framing images of the diode plasmas were acquired.
A nanosecond time-resolved imaging platform for diode plasmas diagnostics has been constructed based on the pulsed electron beam accelerator and high speed framing camera(HSFC). The accelerator can provide an electrical pulse with voltages of 200-500 kV, rise-time (from 10% to 90% amplitude) of 25 ns and duration of 110 ns. The diode currents up to kA level can be extracted. The trigger signal for camera was picked up by a water-resistor voltage divider after the main switch of the accelerator, which could avoid the disadvantageous influence of the time jitter caused by the breakdown of the gas gaps. Then the sampled negative electrical pulse was converted into a transistor-transistor logic(TTL) signal (5 V) with rise time of about 1.5 ns and time jitter less than 1 ns via a processor. And this signal was taken as the synchronization time base. According to the working characteristics of the camera, the synchronization scheme relying mainly on electrical pulse delay method supplemented by light signal delay method was determined to make sure that the camera can work synchronously with the light production and transportation from the diode plasmas within the time scale of nanosecond. Moreover, shielding and filtering methods were used to restrain the interference on the measurement system from the accelerator. Finally, time resolved 2-D framing images of the diode plasmas were acquired.
2012,
24: 968-970.
doi: 10.3788/HPLPB20122404.0968
Abstract:
The paper investigates the generation and transport by multijet cathode of annular intense relativistic electron beam which has been widely used in high power microwave generation. The influence of multijet cathode on the operation of an X-band relativistic backward wave oscillator(RBWO) has been experimentally studied, by using the high-current accelerator SINUS881. The results indicate that, due to its transverse motion, the electron beam will tend to an annularly uniform beam with the increase of cathode jet number, and the output power and pulse width of the RBWO are similar to those of the RBWO with annular cathode.
The paper investigates the generation and transport by multijet cathode of annular intense relativistic electron beam which has been widely used in high power microwave generation. The influence of multijet cathode on the operation of an X-band relativistic backward wave oscillator(RBWO) has been experimentally studied, by using the high-current accelerator SINUS881. The results indicate that, due to its transverse motion, the electron beam will tend to an annularly uniform beam with the increase of cathode jet number, and the output power and pulse width of the RBWO are similar to those of the RBWO with annular cathode.
2012,
24: 971-974.
doi: 10.3788/HPLPB20122404.0971
Abstract:
Based on a novel compact P-band coaxial relativistic backward wave oscillator(RBWO) with only three periods slow wave structure (SWS), an L-band coaxial RBWO is proposed and investigated. Particle simulation results show that, with the 591 kV, 8.2 kA electron beam guided by the 0.8 T solenoidal field, the microwave is generated with the power of 1.50 GW operating at about 1.64 GHz. The interaction efficiency is about 31% and the optimized size of SWS is only f96 mm207 mm. Furthermore, the electrodynamics characteristics of the proposed device, expressly the longitudinal resonant modes and quality factors, are analyzed. Finally, a principle of design for low-band coaxial RBWO is presented with the results of P-band coaxial RBWO and L-band coaxial RBWO, i.e., the length of SWS is about equal to the wavelength, the period length of SWS, and the depths of outer and inner conductor ripples are respectively about 5/13, 1/10, and 1/30 times the wavelength, the radius of electron beam is about 0.7 times the average radius of outer conductor ripple, the longitudinal resonant mode is 0.8 mode and the Q-factor is about 16.
Based on a novel compact P-band coaxial relativistic backward wave oscillator(RBWO) with only three periods slow wave structure (SWS), an L-band coaxial RBWO is proposed and investigated. Particle simulation results show that, with the 591 kV, 8.2 kA electron beam guided by the 0.8 T solenoidal field, the microwave is generated with the power of 1.50 GW operating at about 1.64 GHz. The interaction efficiency is about 31% and the optimized size of SWS is only f96 mm207 mm. Furthermore, the electrodynamics characteristics of the proposed device, expressly the longitudinal resonant modes and quality factors, are analyzed. Finally, a principle of design for low-band coaxial RBWO is presented with the results of P-band coaxial RBWO and L-band coaxial RBWO, i.e., the length of SWS is about equal to the wavelength, the period length of SWS, and the depths of outer and inner conductor ripples are respectively about 5/13, 1/10, and 1/30 times the wavelength, the radius of electron beam is about 0.7 times the average radius of outer conductor ripple, the longitudinal resonant mode is 0.8 mode and the Q-factor is about 16.
2012,
24: 975-979.
doi: 10.3788/HPLPB20122404.0975
Abstract:
Based on the experiment platform with ultra-wide spectrum pulse source of GW level, the thesis investigates the influence on the delay time and jitter of laser-triggered switch of some relative parameters, such as laser pulse energy, laser wavelength, working voltage, and focal length of lens. The experimental results show that, the delay time of the switch decreases gradually as the laser energy or the working voltage increases. When 266 nm laser is adopted, smaller jitter can be achieved than using 1 064 nm laser. On condition of high pressure and small switch gap, the use of lens with smaller focal length shortens the delay time. The jitter is less than 0.3 ns while the self-breakdown voltage is 220 kV, the working ratio is larger than 80% and the energy of the 266 nm laser pulse is 7 mJ. In this condition, the synchronization experiment of two GW-level pulse sources is carried out and the synchronization difference is below 0.3 ns at the probability of 85%.
Based on the experiment platform with ultra-wide spectrum pulse source of GW level, the thesis investigates the influence on the delay time and jitter of laser-triggered switch of some relative parameters, such as laser pulse energy, laser wavelength, working voltage, and focal length of lens. The experimental results show that, the delay time of the switch decreases gradually as the laser energy or the working voltage increases. When 266 nm laser is adopted, smaller jitter can be achieved than using 1 064 nm laser. On condition of high pressure and small switch gap, the use of lens with smaller focal length shortens the delay time. The jitter is less than 0.3 ns while the self-breakdown voltage is 220 kV, the working ratio is larger than 80% and the energy of the 266 nm laser pulse is 7 mJ. In this condition, the synchronization experiment of two GW-level pulse sources is carried out and the synchronization difference is below 0.3 ns at the probability of 85%.
2012,
24: 980-984.
doi: 10.3788/HPLPB20122404.0980
Abstract:
This paper studies the dispersion characteristics and the Pierce interaction impedance of Ku-band folded-waveguide(FW) slow-wave structure(SWS) Shielded by Photonic crystals(PhC). The structure is simulated with the code CST MWS, and the selectivity of PhC to electromagnetic waves in the SWS is discussed. The results reveal that the dispersion of the FW SWS with PhC is smaller than that without PhC, while its Pierce interaction impedance is slightly higher.
This paper studies the dispersion characteristics and the Pierce interaction impedance of Ku-band folded-waveguide(FW) slow-wave structure(SWS) Shielded by Photonic crystals(PhC). The structure is simulated with the code CST MWS, and the selectivity of PhC to electromagnetic waves in the SWS is discussed. The results reveal that the dispersion of the FW SWS with PhC is smaller than that without PhC, while its Pierce interaction impedance is slightly higher.
2012,
24: 985-988.
doi: 10.3788/HPLPB20122404.0985
Abstract:
The article gives the preliminary experimental results of a foilless low-impedance transition-radiation oscillator. In the experiments, copper dielectric and polymer velvet are respectively adopted as explosive emission cathodes and stainless steel and graphite are respectively adopted as collectors of exiting electrons. The experimental results from various states are compared and some possible causes affecting device operation are analyzed. As indicated by the experimental results, the combination of velvet cathode and graphite collector is more advantageous to microwave generation. Dissipating the energy of an intense relativistic electron beam and choosing a proper manner to collect electrons are important to realizing the stable operation of the oscillator.
The article gives the preliminary experimental results of a foilless low-impedance transition-radiation oscillator. In the experiments, copper dielectric and polymer velvet are respectively adopted as explosive emission cathodes and stainless steel and graphite are respectively adopted as collectors of exiting electrons. The experimental results from various states are compared and some possible causes affecting device operation are analyzed. As indicated by the experimental results, the combination of velvet cathode and graphite collector is more advantageous to microwave generation. Dissipating the energy of an intense relativistic electron beam and choosing a proper manner to collect electrons are important to realizing the stable operation of the oscillator.
2012,
24: 989-992.
doi: 10.3788/HPLPB20122404.0989
Abstract:
On the Basis of Khanh Nguyens multiple-beam FWG-TWT structure, the paper investigates the structure of 140 GHz multi-beam FWG-TWT, which can suppress the self-excited oscillation effectively. A lot of simulations have been done about the two-beam FWG-TWT. In comparison with one-beam model, the two-beam model can achieve higher power gain while the length is reduced and the self-excited oscillation is inhibited effectively.
On the Basis of Khanh Nguyens multiple-beam FWG-TWT structure, the paper investigates the structure of 140 GHz multi-beam FWG-TWT, which can suppress the self-excited oscillation effectively. A lot of simulations have been done about the two-beam FWG-TWT. In comparison with one-beam model, the two-beam model can achieve higher power gain while the length is reduced and the self-excited oscillation is inhibited effectively.
2012,
24: 993-997.
doi: 10.3788/HPLPB20122404.0993
Abstract:
A THz-device, Clinotron, has been designed by theoretical analysis and numerical simulation. The parameters of its structrue and performance are obtained by solving the dispersion relation of the Clinotron. The conditions for stable sheet beams transport in periodic permanent magnet fields(PPM) are derived in consideration of space-charge field and cyclotron motion. The Clinotron is simulated and optimized by a three-dimensional PIC code. It gains an averaged output power of 40 W in a period and an efficiency of 4.68% at 218 GHz, when the input voltage is 4.5 kV, input current is 190 mA and the peck magnetic field of focusing PPM is 0.25 T.
A THz-device, Clinotron, has been designed by theoretical analysis and numerical simulation. The parameters of its structrue and performance are obtained by solving the dispersion relation of the Clinotron. The conditions for stable sheet beams transport in periodic permanent magnet fields(PPM) are derived in consideration of space-charge field and cyclotron motion. The Clinotron is simulated and optimized by a three-dimensional PIC code. It gains an averaged output power of 40 W in a period and an efficiency of 4.68% at 218 GHz, when the input voltage is 4.5 kV, input current is 190 mA and the peck magnetic field of focusing PPM is 0.25 T.
2012,
24: 998-1002.
doi: 10.3788/HPLPB20122404.0998
Abstract:
The generation and coupling out of high power wideband microwave based on the switched quarter-wavelength coaxial transmission line oscillator has been addressed. The switched oscillator works at 200 MHz, and is comprised of quarter-wavelength long coaxial transmission lines with a ring-shaped multichannel gaseous spark gap switch at one end. The coupler consists of blocking capacitor and distributed inductance. The generation and coupling out of wideband micorwave in the oscillator was simulated by commercial software. The switching characteristics were also studied. The results of simulation and experiment verify that directly-fed, high impedance and tooth socket structure can improve the stability of the multichannel switch. The wideband microwave generation system has been set up and charged by a high power Marx pulse source with output voltage of 500 kV. The center frequency of radiation is 195 MHz. The radiated peak electric field at a distance of 1 m is 150 kV and the full width at half maximum of frequency spectrum is about 30%.
The generation and coupling out of high power wideband microwave based on the switched quarter-wavelength coaxial transmission line oscillator has been addressed. The switched oscillator works at 200 MHz, and is comprised of quarter-wavelength long coaxial transmission lines with a ring-shaped multichannel gaseous spark gap switch at one end. The coupler consists of blocking capacitor and distributed inductance. The generation and coupling out of wideband micorwave in the oscillator was simulated by commercial software. The switching characteristics were also studied. The results of simulation and experiment verify that directly-fed, high impedance and tooth socket structure can improve the stability of the multichannel switch. The wideband microwave generation system has been set up and charged by a high power Marx pulse source with output voltage of 500 kV. The center frequency of radiation is 195 MHz. The radiated peak electric field at a distance of 1 m is 150 kV and the full width at half maximum of frequency spectrum is about 30%.
2012,
24: 1003-1008.
doi: 10.3788/HPLPB20122404.1003
Abstract:
To improve the vacuum environment of an L-band, giga watt-level magnetically insulated line oscillator(MILO), an optimized pumping model was developed for the molecular movement and collision in the high-current vacuum diode chamber and the MILO tube with Monte-Carlo method. According to the three dimensional molecular distributions in MILO, a scheme of distributed pumping was proposed. In this scheme, another pump system which was close to the velvet cathode was introduced and located between the slow wave structure and antenna. Simulation predictions were verified through the experimental test carried out on Torch-01 pulser. Distributed pumping can efficiently reduce the characteristic time of pressure drop to 0.22 times that of single pumping. Moreover, at repetitive operation of MILO, distributed pumping maintains the vacuum degree in MILO better, and for our experimental condition, it can improve the MILOs repetition rate by 5 times.
To improve the vacuum environment of an L-band, giga watt-level magnetically insulated line oscillator(MILO), an optimized pumping model was developed for the molecular movement and collision in the high-current vacuum diode chamber and the MILO tube with Monte-Carlo method. According to the three dimensional molecular distributions in MILO, a scheme of distributed pumping was proposed. In this scheme, another pump system which was close to the velvet cathode was introduced and located between the slow wave structure and antenna. Simulation predictions were verified through the experimental test carried out on Torch-01 pulser. Distributed pumping can efficiently reduce the characteristic time of pressure drop to 0.22 times that of single pumping. Moreover, at repetitive operation of MILO, distributed pumping maintains the vacuum degree in MILO better, and for our experimental condition, it can improve the MILOs repetition rate by 5 times.
