scispace - formally typeset
Search or ask a question

Showing papers on "Rise time published in 1989"


Journal ArticleDOI
TL;DR: In this article, a Fourier-series calculation is presented that predicts the occurrence of Brillouin precursors when the incident pulses have sufficiently short rise times, and an analysis of the numerical error resulting from truncation of the Fourier series is given.
Abstract: Numerical research is reported on the propagation of short microwave pulses into living, biological materials. These materials are dispersive, and data on the dielectric constant and conductivity for these materials follow a Debye model. A Fourier-series calculation is presented that predicts the occurrence of Brillouin precursors when the incident pulses have sufficiently short rise times. These transients are attenuated with increasing propagation distance but are attenuated more slowly than the carrier frequency of the pulse, which is attenuated exponentially with distance. An analysis of the numerical error resulting from truncation of the Fourier series is given. Upper-bound estimates of truncation error show good series convergence.

83 citations


Journal ArticleDOI
TL;DR: In this paper, a modified Mach-Zehnder interferometer was used to measure the ultrashort time strain response of a sample using the dispersion of piezoelectric and electrostrictive coefficients of several commonly used ferroelectrics.
Abstract: In this paper we report on a modified Mach‐Zehnder interferometer which is capable of measuring the ultrashort time strain response of a sample Using this instrument, the dispersion of piezoelectric and electrostrictive coefficients of several commonly used ferroelectrics was measured The strain switching behavior of lead lanthanum zirconate titanate (PLZT) 9/65/35 driven by a square pulse electric field was studied The results show that the switching rise time of the strain response for PLZT 9/65/35 can be much shorter than 5 μs at a pulse field strength near 25 kV/cm and that the fall time of the strain response is considerably shorter than the rise time in the experimental field strength range

69 citations


Journal ArticleDOI
TL;DR: The results for this particular gripper indicate that the additional sensory feedback can decrease the force step response rise time by 88% while maintaining a monotonic zero-overshoot response.
Abstract: A novel robot gripper control system is presented that uses PVDF (polyvinylidene fluoride) piezoelectric sensors to damp exerted force actively. By monitoring the current developed by the PVDF sensor, an output proportional to the rate of change of the force exerted by the gripper is obtained. The signals from the PVDF sensor and strain-gauge force sensor are arranged in a proportional and derivative control system for the control of force. The control system was tested on an instrumented Rhino XR-1 manipulator hand. The capabilities of the control system are analyzed and are verified experimentally. The results for this particular gripper indicate that the additional sensory feedback can decrease the force step response rise time by 88% while maintaining a monotonic zero-overshoot response. The inclusion of the rate feedback increases the damping ratio of the dominant poles while maintaining the step response rise time. >

66 citations


Patent
05 Jan 1989
TL;DR: In this article, an interface control circuit comprising a buffer, an inverter, an OR gate, and delay means is proposed for communication handshaking such that a buffer output current will actively flow through line stray capacitance thereby greatly reducing the rise time from a LOW state to a HIGH state.
Abstract: An interface control circuit comprising a buffer, an inverter, an OR gate and delay means. The interface control circuit can be utilized in digital systems for communication handshaking such that a buffer output current will actively flow through line stray capacitance thereby greatly reducing the rise time from a LOW state to a HIGH state, or charges in the line stray capacitance will actively discharge through the buffer means if negative logic mode is used thereby enabling a great reduction in fall time, therefore a much faster and efficient digital system can be obtained through the use of the invention.

38 citations


Journal ArticleDOI
TL;DR: In this paper, the response time of a GaAs/Alx Ga1−x As multiquantum well infrared detector (at a wavelength of λ=6.4 μm) was determined to be less than 300 ps.
Abstract: We have measured the response time of a GaAs/Alx Ga1−x As multiquantum well infrared detector (at a wavelength of λ=6.4 μm). The intrinsic rise time is determined to be less than 300 ps.

30 citations


Journal ArticleDOI
TL;DR: In this article, a new analytical procedure is developed to predict the dynamic response of axially restrained beams to intense pulse loads, where the beams are assumed to have doubly symmetric cross-sections and to obey a rigid-plastic material law.

25 citations


Journal ArticleDOI
TL;DR: In this paper, a VDMOSFET structure with an additional p-region at the surface of the epitaxial layer is proposed to realize low reverse transfer capacitance without significantly degrading resistance.
Abstract: A VDMOSFET structure with an additional p-region at the surface of the epitaxial layer is proposed. This structure realizes low reverse transfer capacitance without significantly degrading resistance. Reduction of the reverse transfer capacitance results in an improvement of the switching characteristics. The measured rise time is 49% and the fall time is 33% of the conventional VDMOSFET. >

25 citations


Proceedings ArticleDOI
13 Mar 1989
TL;DR: In this article, the authors derived the theoretical time and frequency-domain waveforms of common-mode conducted emissions for an arbitrary SMPS duty cycle and showed that the magnitude of commonmode emissions is independent of collector voltage rise time.
Abstract: The theoretical time- and frequency-domain waveforms of common-mode conducted emissions for an arbitrary SMPS duty cycle is derived. It is shown that the magnitude of common-mode emissions is independent of collector voltage rise time. Experimental data is presented for model validation. Common mode emissions are shown to be independent of switching transistor rise time. The model predicts the variation of emissions with duty cycle, using Fourier analysis of the common-mode waveforms with the duty cycle as a parameter. >

21 citations


Patent
Charles S. Walker1
23 Jan 1989
TL;DR: In this paper, a pulse width modulator is provided with circuitry to eliminate high frequency noise spikes in its output, which is effective to slow down the switching time of the MOSFET switching transistor, increase the rise time of voltage across the catch diode and improve the high frequency characteristics of output filter.
Abstract: A pulse width modulator is provided with circuitry to eliminate high frequency noise spikes in its output. This circuitry is effective to slow down the switching time of the MOSFET switching transistor, increase the rise time of the voltage across the catch diode and improve the high frequency characteristics of the output filter. These benefits are achieved by only a few number of components.

13 citations


Patent
29 Dec 1989
TL;DR: In this article, a delay circuit with a complementary insulated gate device is considered, where the transistors Q1 and Q2 of the transmission gate (20) transmit only a litle increase of the input voltage to the inverter (10) during the increase of input voltage.
Abstract: A delay circuit having a complementary insulated gate device comprises an inverter (10) having a series connection of a p type field effect transistor Q3 and an n type field effect transistor Q4 and a transmission gate (20) having a parallel connection of a p type field effect transistor Q1 and an n type field effect transistor Q2 connected to the preceding stage of the inverter (10), and the gates of the transistors Q1 and Q2 are connected together to an output terminal (3). The logical threshold voltage of the inverter (10) is set at a higher value in the range of the input voltage of the inverter (10). There is a peculiar period in which the transistors Q1 and Q2 of the transmission gate (20) transmit only a litle increase of the input voltage to the inverter (10) during the increase of the input voltage. Due to the existence of this peculiar period, this circuit outputs a delayed output signal only when the input voltage increases. In addition, the rise time and fall time of the output signal are short.

12 citations


Journal ArticleDOI
TL;DR: In this paper, the authors presented a discussion on the selection of suitable solid-state active devices for operation at such a low temperature, as well as the results of measurements of static characteristic and noise behaviour performed on GaAs MESFETs.
Abstract: Particle bolometric detectors operate at temperatures ranging from a few to hundreds of mK. When the readout thermistor (in thermal contact with the absorbing crystal) presents a high impedance, a front-end electronics located very close to the detector is required. This imposes on the electronic circuit the condition of operating at very low temperature, in our case 1 K. A discussion on the selection of suitable solid-state active devices for operation at such a low temperature, as well as the results of measurements of static characteristic and noise behaviour performed on GaAs MESFETs are presented. After component selection a complete voltage-sensitive preamplifier was constructed and tested. With a closed loop gain of 10, the rise time is 50 ns and the 1/⨍ noise has an A⨍ coefficient of 1.1 × 10−11 V2. The noise spectral power density at 1 MHz is 2.5 × 10−18 V2/Hz.

Journal ArticleDOI
TL;DR: The SLITS (SF/sub 6/ laser-induced triggering system) as mentioned in this paper is an add-on fast switching capability that makes use of the existing equipment and can be triggered at voltages as low as 10% of the spontaneous breakdown voltage of the gap.
Abstract: A description is given of a nanosecond-risetime, 750-kV spark gap impulse generator that is triggered by an ultraviolet laser pulse (KrF, lambda =249 nm). The SLITS (SF/sub 6/ laser-induced triggering system) is an add-on fast switching capability that makes use of the existing equipment (Marx, AC or DC) and can be triggered at voltages as low as 10% of the spontaneous breakdown voltage of the gap. The output voltage can be varied simply by adjusting the internal delay of the laser or the source voltage, eliminating the awkward adjustment of the sphere gaps required in a conventional pulse generator. Jitter is reduced too less than 10 ns at 80% of the spontaneous breakdown voltage, making the output waveform highly reproducible. >

Journal ArticleDOI
TL;DR: In this article, a two-stage MCP ion detector for steady ion currents of the order of some pA and ion pulses of a pulse width between 10 and 500 ns is described.
Abstract: A two-stage MCP ion detector for steady ion currents of the order of some pA and ion pulses of a pulse width between 10 and 500 ns is described. Detector current gain for Ne+ and H2+ ions at an energy of 0.1 to 4 keV was found to be dependent on ion energy and species. In pulse current mode, linearity is maintained up to an output charge density of 1.9 nC cm-2; the detector rise time is below 3.5 ns. For the first time an MCP ion detector is used at an ambient gas pressure of 10-3 Pa to investigate ions emitted by a transient gas discharge. With an appropriate operating voltage the detector operated without degradation of performance for at least 500 hours. First measurements of ion species, ion-energy distribution function, ion current density and pulse shape of the axial ion pulse of a pseudospark discharge were performed.

Proceedings ArticleDOI
29 Aug 1989
TL;DR: Equations which relate capacitive and inductive parameters to trace geometry, including intervening grounded traces, are presented and examples of crosstalk, as calculated by a SPICE model using derived parameters, are shown.
Abstract: Signals propagating on PC (printed circuit) board traces are subject to overshoot and crosstalk, both of which increase as rise time decreases. Overshoot depends upon the ratio of self-inductance to capacitance to ground; crosstalk depends also upon mutual inductance between traces and trace-to-trace capacitance. Multilayer boards with ground planes close to traces are shown to decrease inductance and coupling capacitance, thereby tending to reduce overshoot and crosstalk. A similar reduction is obtained by placing grounded traces between signal traces. Equations which relate capacitive and inductive parameters to trace geometry, including intervening grounded traces, are presented. Examples of crosstalk, as calculated by a SPICE model using derived parameters, are shown. >

Journal ArticleDOI
TL;DR: In this article, the effect of rise time and pulse width on the transport and interaction of a 1 MeV, 15 kA relativistic electron beam (REB) in argon and hydrogen in the pressure range 0.05-1.0 Torr is reported.
Abstract: A numerical study of the effect of rise time (10–30 nsec) and pulse width (30–100 nsec) on the transport and interaction of a 1 MeV, 15 kA relativistic electron beam (REB) in argon and hydrogen in the pressure range 0.05–1.0 Torr is reported. The computations have been carried out taking into account charge and current neutralization effects, plasma heating by return current, and two‐stream instabilities. The results of the study show that charge transport increases for short rise times and longer pulse widths at the optimum pressure in the range 0.1–0.2 Torr. With a rise time of 30 nsec and a pulse width of 100 nsec, the 1 MeV, 15 kA REB generates a plasma with density 1–10×1021 m−3, an electron temperature of 4–20 eV, and transports with 90% efficiency at the optimum pressure.

Patent
11 Apr 1989
TL;DR: In this paper, a two-step process and apparatus for the electrolytic coloring of anodized aluminum using an electrode and the anodised aluminum immersed in an electrolyte is described.
Abstract: A two-step process and apparatus for the electrolytic coloring of anodized aluminum using an electrode and the anodized aluminum immersed in an electrolyte. First the aluminum is anodized to form an oxidized film on its surface and then a coloring step is performed. In the coloring step a modified sinusoidal voltage waveform is applied to the anodized aluminum and the electrode to color the oxidized film. Suitably gated thyristors in the secondary of a power supply may be used to provide the desired voltage waveform. The voltage waveform is generated with a leading edge rise time in the negative portion substantially longer than in the positive portion. The average negative voltage is greater than the average positive voltage to provide good throwing power, coloring speed, color uniformity and electrode dissolution, with tha capability to subtract color. Excellent throwing power with fair coloring speed is obtained if the average negative voltage substantially equals the average positive voltage.

Patent
William G. Miller1
15 Feb 1989
TL;DR: In this article, a coring circuit in a luminance processor has a characteristic which starts at the O input/O output point and extends linearly in both a positive and negative direction, implemented by a linear amplifier from which the input luminance signal is blocked until it has reached a predetermined threshold amplitude.
Abstract: The coring circuit in a luminance processor has a characteristic which starts at the O input/O output point and extends linearly in both a positive and negative direction. The characteristic is implemented by a linear amplifier from which the input luminance signal is blocked until it has reached a predetermined threshold amplitude. Low amplitude, high frequency noise is therefore removed without causing overpeaking at high transition levels. The coating stage is preceded in the signal processor by a circuit which furnishes a peaking signal whose rise and fall time are independent of the amplitude of the transition. This circuit is implemented by feeding back the peaked luminance signal through a first filter which has a cut-off at 4 MHz and therefore furnishes a signal responsive both to the change in amplitude and the rate of change of that amplitude, and a second filter whose output is proportional only to the change in amplitude. The output of the second filter is subtracted from that of the first, so that the resultant gain control signal varies only as a function of the transition time. This gain control signal is then applied to a gain control amplifier in the forward loop. The feedback operates to keep the transition times constant.

Patent
24 Apr 1989
TL;DR: In this paper, the authors proposed to shorten the rise time until the stable state of plasma after start by previously applying a low-frequency power to a counter electrode through a specific method on the start of plasma, applying a highfrequency power and generating plasma.
Abstract: PURPOSE:To shorten the rise time until the state reaches the stable state of plasma after start by previously applying a low-frequency power to a counter electrode through a specific method on the start of plasma, applying a high-frequency power and generating plasma. CONSTITUTION:A start control means 19 controlling the start of plasma is provided. Low-frequency voltage is applied previously to counter t electrodes 16a, 16b through a low-frequency power 16 and an applied-voltage control circuit 91, and low-frequency power is supplied. The high-frequency voltage of a high- frequency power 15 is applied after switch-closing time difference T2[sec] through a power-closing delay circuit 92, counter electrodes 15a, 15b are supplied with high-frequency power through an automatic matching appliance 17, and plasma is generated. The rise time of the low-frequency power applied to the counter electrodes 16a, 16b is set at a value shorter than the rise time of the high-frequency power by a rise-time setting circuit 93, high-frequency power fed to the counter electrodes 15a, 15b is controlled through the automatic matching appliance 17 in the high-frequency power 15, and plasma is generated.

Journal ArticleDOI
15 Feb 1989
TL;DR: In this article, a high-resolution class-B video output driver IC for analog or digital video display systems is described, which uses a single 5V supply, combined with high-voltage discrete output power transistors to drive either a color or a monochrome CRT (cathode ray tube).
Abstract: The authors describe a high-resolution class-B video output driver IC for analog or digital video display systems. The IC, which uses a single 5-V supply, is combined with high-voltage discrete output power transistors to drive either a color or a monochrome CRT (cathode ray tube). The system yields greater bandwidth while consuming one third the power required by conventional class-A CRT drivers. The circuit was fabricated in a 2- mu m oxide-isolated bipolar process with peak f/sub T/ approximately=8 GHz. Using discrete output transistors (f/sub T/=1 GHz, C/sub CB/=2pF), board testing on a CRT with a nominal 10-pF load yielded approximately 15-ns output rise and fall times with 45-V swing and 60-mA slew current limit. >

Journal ArticleDOI
TL;DR: In this paper, the authors conducted short-duration voltage surge tests on commercially available power line filters and two types of transient voltage suppressors to determine their effectiveness for protection against nuclear electromagnetic pulse (EMP) induced power-line transients.
Abstract: Steep-front short-duration voltage surge tests were conducted on commercially available power-line filters and two types of transient voltage suppressors to determine their effectiveness for protection against nuclear electromagnetic pulse (EMP) induced power-line transients. The tested power line filters were single-phase industrial-type units in the 120/208 V and 277/480 V class rated at 100 or 200 A. It was found that the tested power-line filters provide from 38 to 54 dB of attenuation for steep front surges (20 to 40 ns rise time and 150 to 400 ns fall time to half-value waves for 2 to 20 kV peak voltage), and transient voltage suppressors add an additional 2 to 8 dB of attenuation. It was also found that while the tested power line filters provide good protection against EMP induced surges, degradation of the filter characteristics may occur and delayed failures of filter components are possible. Transient voltage suppressors reduce the EMP induced surge by 40% and thus provide some protection for the filter. >

Journal ArticleDOI
TL;DR: The results indicated that the mechanical response is indeed time dependent and the assumption of rigid body segments in these experiments appeared reasonably justified when the rise time of the applied force was greater than approximately 300 ms, however, for input signals that had a shorter rise time the trunk did not behave rigidly.

Proceedings ArticleDOI
15 May 1989
TL;DR: A 4.0-GB/s NMOS laser driver has been designed and fabricated in 0.75-μm technology with a typical Leff=0.65 μm providing an adjustable modulation current pulse of 20-60 mA into an equivalent 25-Ω load.
Abstract: A 4.0-GB/s NMOS laser driver has been designed and fabricated in 0.75-mm technology with a typical L eff=0.65 mm. The IC provides an adjustable modulation current pulse of 20-60 mA into an equivalent 25-O load. The rise and fall times (10 to 90%) are approximately 70 ps and 120 ps, respectively with a V DD=3.3 Vp10%

Proceedings ArticleDOI
14 Aug 1989
TL;DR: In this paper, a model for the propagation delay and rise (fall) times of an inverter is discussed, based on the approximate analysis of the differential equations that govern the behavior of the inverter.
Abstract: A model for the propagation delay and rise (fall) times of an inverter is discussed. The model is based on the approximate analysis of the differential equations that govern the behavior of the inverter. The slopes of the input waveform along with the currents of the pMOS (nMOS) transistor in the falling (rising) outputs are included in the model. Simulations are made using the model for different values of transconductances of transistors, loads, and rise (fall) times. These simulations are compared to the results of SPICE simulations. >

Journal ArticleDOI
TL;DR: In this article, the authors describe a method of generating both line and point sources of acoustic emission (AE) signals by a powerful, transient Joule heating of a thin, conductive film deposited on an insulating plate or a small diameter metal wire buried inside it.
Abstract: This paper describes a method of generating both line and point sources of acoustic emission (AE) signals by a powerful, transient Joule heating of a thin, conductive film deposited on an insulating plate or a small diameter metal wire buried inside it. Point‐and line‐source AE signals are generated by heating a short or a long section of thin film, respectively. The versatility of this method is demonstated by the diversity of AE signals that can be produced by simply changing the rise time, magnitude, and duration of the current pulse that provides the heating. An analysis shows that both the line and the point AE sources are of dipolar type for the case of the heated thin film. The line source resulting from an imbedded long, thin wire appears to be a two‐dimensional center of dilatation. Short‐duration current pulses generate a step source whose rise time is approximately equal to that of the input electrical power pulse while long‐current pulses produce a linear ramp source that resembles the tempera...

Patent
29 Sep 1989
TL;DR: In this article, the authors proposed a method to prevent the change of a rise time of an optical pulse output by a method wherein a voltage difference between base voltages given to a first and second transistors of a pulse current supply circuit respectively is made to change corresponding to the temperature change of the laser diode.
Abstract: PURPOSE:To prevent the change of a rise time of an optical pulse output by a method wherein a voltage difference between base voltages given to a first and second transistors of a pulse current supply circuit respectively is made to change corresponding to the temperature change of a laser diode. CONSTITUTION:A voltage between a base and an emitter of each of transistors(Trs) 22 and 23 and Trs used in the circuits 12 and 12' and a voltage between the ends of each of diodes 19 and 19' have a negative temperature dependency and the emitter area of a Tr 16' of the circuit 12' is made sufficiently larger than that of Tr 23 and a current almost equal to that of a constant current source 21 is made to flow through the Tr 16', where by voltage difference between voltages given to the bases of Trs 13 and 13' from base voltage generating circuits 12 and 12' of a pulse current supply circuit is made to have a positive temperature dependency. Therefore, a current pulse Ip supplied to a laser diode 1 from a circuit 6 is made to have a positive temperature dependency, so that even if the laser diode 1 changes in temperature, a light pulse whose rise time is kept constant can be projected from the laser diode 1.

Journal ArticleDOI
TL;DR: In this paper, the transients in photorefractive two-wave mixing are investigated under the condition when the electron generation time is comparable with the dielectric relaxation time and it is shown that for high enough illumination, oscillatory transient behaviour is absent and rise time is proportional to the square root of the input intensity.

Patent
21 Feb 1989
TL;DR: In this paper, the authors present an approach for conditioning high voltage pulses from particle accelerators in order to shorten the rise times of the pulses, using flashover switches in the cathode stalk of the transmission line hold off conduction for a determinable period of time.
Abstract: Apparatus for conditioning high voltage pulses from particle accelerators in order to shorten the rise times of the pulses. Flashover switches in the cathode stalk of the transmission line hold off conduction for a determinable period of time, reflecting the early portion of the pulses. Diodes upstream of the switches divert energy into the magnetic and electrostatic storage of the capacitance and inductance inherent to the transmission line until the switches close.

Journal ArticleDOI
TL;DR: In this article, a charge sensitive preamplifier with low power consumption was developed in two versions: the first using a JFET SNJ903L has a matching capacitance of 310 pF whereas the second one employing a SNJ3600L has 1100 pF.
Abstract: A charge-sensitive preamplifier showing very low power consumption was developed in two versions: the first using a JFET SNJ903L has a matching capacitance of 310 pF whereas the second one employing a SNJ3600L has 1100 pF. Both versions have a charge sensitivity of 8 μV/e−. The rise time is lower than 200 ns for both low and high capacitance versions at an input capacitance of 1000 and 3000 pF respectively. The low capacitance version has an equivalent noise charge of 220 e− rms at 0 pF for a CR-(RC)2 shaping at 10 μs shaping time. The noise at 1 μs is 510 e− rms at 0 pF and the noise slope is 1.59 e−/pF. The high capacitance version exhibits 400 e− rms at 0 pF and 10 μs shaping time. At 1 μs the equivalent noise charge is 1100 e− rms and the noise slope is 0.91 e−/pF. The second stage uses operational amplifiers of 16 MHz gain-bandwidth product and is prepared to make a differentiation with pole-zero cancellation if required. The final stage provides a differential output with 50 ω output impedance and can drive a 100 ω twisted pair cable loaded with its characteristic impedance. The total power dissipation is 190 mW for the SNJ903L version whereas the SNJ3600L version dissipates 265 mW. Both versions use ±6 V. The circuit is mounted using thick film technology onto a 20×50 mm2 ceramic substrate. The project is complemented with the design of a 3000 V, 24 pF/m coaxial cable for detector biasing and signal transport with minimum contribution to the ENC.

01 Jan 1989
TL;DR: In this article, the effect of parameters spread in superconducting cavities operating in pulsed mode was studied when several cavities are fed by one klystron and the influence of a spread in external Qs (from coupler tolerances or on purpose for having different fields from cavity to cavity).
Abstract: The Lorentz forces (also called “radiation pressure”) and microphonics, by shifting the cavity frequency, are the main bunch-to-bunch energy spread sources. With superconducting cavities operating in pulsed mode, the Lorentz Forces problem arises from the wall deformation response time [l] . The cavity frequency goes on to shift after the field rise time, whereas the beam is passing through the cavity. After a brieve review of the two methods [2,3] coping with the Lorentz forces detuning when one cavity only is fed by one klystron, the effect of parameters spreads is studied when several cavities are fed by one klystron. External feedback loops to minimize the residual amplitude and phase errors are then added and the loop gains are determined. The influence of a spread in external Qs (from coupler tolerances or on purpose for having different fields from cavity to cavity) is analysed and the extra power needed to stabilize the total accelerating voltage is given after an optimization of the beam injection time. Finally, microphonics effects, which can increase dramatically the field errors, are considered and a remedy, allowing to alleviate the problem, is proposed [41.

Proceedings ArticleDOI
M.M. Kekez1, J. LoVetri1, A.S. Podgorski1, J.G. Dunn1, G. Gibson1 
11 Jun 1989
TL;DR: In this paper, the authors describe a low energy Marx generator system capable of producing a fast (1 ns rise time) /spl Ichemo/ 600 kV pulse with a peak power approaching 10 GW.
Abstract: This paper describes work on the design and development of a low energy Marx generator system capable of producing a fast (1 ns rise time) /spl Ichemo/ 600 kV pulse. The system consists of a Marx generator, peaking circuit and a short length coaxial line with a passive load. The generator delivers the power to 10-150 ohm loads with a peak power approaching 10 GW. The unique features are simplicity of design, ease of operation, and above all compactness which permits easy portability and in-situ application. The system can be used: a) to produce an intense relativistic electron/ion beam in the virtual cathode reflex triode for high power microwave generation; b) for flash X-ray radiography to record and study dynamic events where interposed material, debris, or flame exclude the use of high speed cameras; c) to provide 10/sup 6/ to 10/sup 7/ D-D (2.45 MeV) or D-T (14.1 MeV) neutrons in several nanoseconds; d) to study the dielectric properties of matter; e) for NEMP studies; and f) to preionize a powerful (electron beam controlled) eximer laser system. Experimental results are presented along with computer simulations using lumped-parameter approximations.