Figure 1 - uploaded by Werner Hartmann
Content may be subject to copyright.
Source publication
An analysis of commercial semiconductor switches favors the IGBT for pulsed power applications, in particular for high average power, high pulse repetition rate applications, due to its availability based on its widespread use in drive applications. High power IGBT modules rated at 6.5 kV/700 A of two different technologies have been investigated i...
Contexts in source publication
Context 1
... low inductance snubber capacitors directly attached to the collector terminal screws of the IGBT under test. A low inductance stripline made from copper sheet connects the capacitors and the load resistor. The load resistor is made from a couple of parallel, low inductance carbon bulk resistors directly soldered to the copper stripline ends, c. Fig. 1. The capacitors are charged at the collector side of the IGBT by a regulated high voltage power supply to voltages between 500 V and 4 kV. The schematic of the single IGBT setup is shown in figure 2: While in standard applications the gate -emitter voltage rate of rise usually is limited by damping series resistors in order to prevent ...
Context 2
... fast gate drive unit with high current capability is mandatory. The schematic of the gate drive unit is shown in Fig. 2. It is based on a commercial, high current, fast driver IC with a current capability of 30 A. The printed circuit board (PCB) is directly screwed to the IGBT gate and emitter terminals for minimal inductance of the circuit, c. Fig. 1. It allows gate current risetimes as low as a few nanoseconds as shown in Fig. 3. For testing of IGBT modules, the nominal gate driver supply voltage has been set to 15 V, well below the limit of the gate -emitter voltages of 20 V of the devices under test [1], ...
Similar publications
Many magnetic confinement machines use active feedback stabilization with magnetic actuators. We present a novel magnetic actuators system with a response time much faster than previous ones, making it capable of coping with the fast plasma instabilities. The system achieved a response time of 3 μs with maximal current of 500 A in a coil with induc...
Citations
... The most obvious way to control the switching speed of a power switch like a MOSFET or an IGBT, to avoid gatevoltage ringing, limit gate-voltage overshoot, and reduce the generated EMI, is by inserting a gate resistor between the driver output and the gate of the power switch [1]. The gate-driver in this configuration performs in general as a voltage source providing at least two constant levels of voltage, most commonly two voltage values between -7V and +20V. ...
This paper presents a galvanically isolated gate-driver integrated circuit realized as an ASIC chipset providing a flexible control of the switching speed of the driven power switches (i.e., IGBT or MOSFET). The driver chipset provides signal and power transmission over a galvanic isolation, thus being able to drive low-side and high-side power switches in power converters. It provides independent control of turn-on and turn-off switching speed by modulating the gate turn-on and turn-off voltage slopes using burst pulses in the MHz range. This function is combined with regenerative switching, thus reducing the energy losses in the gate-driver circuit of the power switch by more than 50%. The gate-driver ASIC chipset was manufactured in a high-temperature automotive grade 0.35μm mixed-signal CMOS technology, thus allowing switching speeds in the MHz range at voltage amplitudes as high as 18V. The paper shows the novel proposed driving concept with its implemented topology and simulation results. Experimental results validate the proposed gate-driver concept based on the manufactured ASIC chipset combined with a typical IGBT as power switch.
... Semiconductor devices have made dramatic progress in power handling over the last decade. Today's technology and production capabilities make it possible to produce devices with high blocking voltage combined with very high current handling (Das 2011;Hartmann 2013, Lopez 2012Molloy 2011). Depending on the design and the device structure, also very high current rise rates in the range of up to several tenths of kA/µs are possible. ...
... A semi-planar, rotationally symmetric power combiner is described which will be used as a building block for an inductive voltage adder-based high voltage pulse generator [1], [2], [3]. It has been realized using a hard-wired parallel circuit of four power transfer stages which feed into a common radial transmission line. ...
... From there on, it can be directly coupled to a load or a cable, or coupled into the transformer section of an inductive voltage adder stage [2], [3]. The capacitors are screwed directly to the collector contacts of the IGBT module to minimize the circuit inductance [1], and are charged in parallel by a common high voltage power supply. The switches are realized using standard high voltage (6.5 kV) IGBT power modules [1], [4]; simultaneous triggering of the IGBTs discharges all of these capacitors simultaneously into the common load. ...
... The capacitors are screwed directly to the collector contacts of the IGBT module to minimize the circuit inductance [1], and are charged in parallel by a common high voltage power supply. The switches are realized using standard high voltage (6.5 kV) IGBT power modules [1], [4]; simultaneous triggering of the IGBTs discharges all of these capacitors simultaneously into the common load. In order to realize this concept, it is necessary to -produce minimal losses in each of the branches; -have identical switching delays of all switches; -have minimal switch jitter in each brick; -provide identical pulse shapes from each brick; -provide optimal matching of the radial transmission line and the load. ...
A semi-planar, rotationally symmetric power combiner has been realized using a hard-wired parallel circuit of four power transfer stages which feed into a common radial transmission line. For testing purposes, the radial transmission line is terminated with a matched ohmic resistor. The power combiner is designed to produce a double exponential pulse by optimizing the transmission line geometry with the help of electrodynamic modeling using CST Microwave Suite. Each of the four stages contains three parallel capacitors of 200nF each, the geometrical circuit inductance, a switch, and are discharged into the common resistive load. The switch is realized with an industrial type IGBT module, at a DC link voltage of up to 4.5 kV. Lifetime estimations show a permissible peak current of up to 2kA for a single IGBT module, at pulse durations of around 1 μs. Hence, a peak current of over 6 kA can be achieved by paralleling four of these power transfer stages in the semi-planar power combiner structure. First experimental results show that the semi-planar power combiner is a suitable functional unit for pulsed power applications. The circuit was characterized at a DC link voltage of up to 4 kV, a peak current of 1.68kA, and a pulse duration of 1μs per IGBT module. The low switching losses of the IGBT when using a hard gate drive allow using the IGBT at high pulse repetition rates (PRF) up to kHz, at pulse durations around 1μs. The circuit presented is suitable to be used as a modular component of an inductive voltage adder to increase the available voltage and peak power levels, respectively.
... We report on an Inductive Voltage Adder (IVA) development designated for industrial applications like electroporation, environmental applications, etc. The IVA described in this work is a three-stage demonstrator which shows the feasibility of using conventional high power IGBT semiconductor switch modules [2] instead of spark gaps or arrays of low-power IGBTs. The maximum pulse repetition rate in such a configuration is estimated to be around 2 kHz without minimal forced cooling of the semiconductor switches considering the thermal properties of the components. ...
... The IVA demonstrator is designed with three stages to demonstrate the feasibility of the concept at long pulse durations of typically >1 µs FHWM (full width at half maximum). Each stage is designed as a radial transmission line fed by several parallel pulse modules ("bricks") [2] which contain the electrical components, like capacitors, inductors, and switches. Each brick is connected over a matching network to the radial transmission line, which combines the power of all bricks [3] in a common inductive adder stage called "transformer". ...
... As described in Refs. (2] and [3], the experimental IVA presented hereafter consists of a "series" array of three IVA modules each in turn consisting of 4 parallel pulse modules. All pulse modules ("bricks") are identical and consist of a high voltage IGBT switch [2] and a parallel pair of snubber capacitors; the current leads are made from aluminum plate and are equipped with sets of pin contacts for easy plug-and-play assembly. ...
We report on an Inductive Voltage Adder (IVA) [1] development designated for industrial applications like electroporation, environmental applications, etc. The IVA described [2]-[4] is a three-stage demonstrator which shows the feasibility of using conventional high power IGBT semiconductor switch modules instead of spark gaps or arrays of low-power IGBTs. The pulse generator described produces pulses with peak voltages and currents of up to 12kV and 6kA, respectively, at a pulse duration of typically 1μs FWHM (full width at half maximum). The IVA is tested in single pulse mode and at low repetition rates due to load and power supply constrictions. Each IVA stage is designed as a radial transmission line fed by several parallel pulse modules (“bricks”) which contain the electrical components, like capacitors, inductors, and switches. The combined power of an individual stage is added to the preceding stages in a section called “transformer”. The transformer matches the electric and magnetic fields and is realized as a combination of radial and coaxial transmission lines. Each stage of the IVA is matched to the next stage and is connected in series with a coaxial transmission line. The mechanical dimensions of the three-stage IVA demonstrator are: outer diameter 820 mm, outer diameter of the coaxial transmission line 210 mm, height of the IVA 352 mm. The I VA geometry, in particular the most critical parts radial transmission line and transformer section, is simulated by a transient electromagnetic field solver to analyze the reflections and transmission coefficients of the device.
... The concept of an inductive voltage adder (IVA) [4] which is suitable for industrial pulsed power applications like electroporation (food and beverages, produce / fruits, stark containing products like corn and potatoes, water / wastewater, algae, wet bio material as feedstock for energy applications, etc.), pulsed electrostatic precipitators; pulsed x-ray sources for medical, industrial and security applications; and other novel, environmentally friendly processes is described. The IVA is based on the use of high power semiconductor switch modules [1] which have a longer lifetime, higher reliability and need much less maintenance than conventional spark gaps which are used in today's Marx generators. The pulse generator is designed to produce a double exponential waveform at voltage amplitudes of tens of kV to >100 kV, and current amplitudes of typically 10 kA into a matched load, at a pulse duration of typically 1 µs FWHM (full width at half maximum). ...
... The pulse generator is designed to produce a double exponential waveform at voltage amplitudes of tens of kV to >100 kV, and current amplitudes of typically 10 kA into a matched load, at a pulse duration of typically 1 µs FWHM (full width at half maximum). In contrast to conventional Marx generators, which use spark gaps as switching elements and therefore are limited to pulse repetition rates (PRF) of typically <50 Hz, the use of solid state technology [1] enables pulse repetition rates far in excess of 100 Hz. Therefore, applications requiring high average power like electroporation of produce, sludge, etc., can be realized much less cost intensive using solid state switching technology. ...
... These are the maximal operating voltage of the capacitor, the breakdown voltage of the semiconductor, and the electrical field strength of the transmission line system. For single high voltage semiconductor switches, in particular for 6.5 kV high voltage IGBT modules which are commercially available for standard applications, typical operating (DC link) voltages are around 4 to 4.5 kV [1]. ...
The concept of an inductive voltage adder (IVA) which is suitable for industrial pulsed power applications is described. The pulse generator is designed to produce a double exponential waveform at voltage amplitudes of tens of kV to >100 kV, and current amplitudes of typically 10 kA into a matched load, at a pulse duration of typically 1 μs FWHM (full width at half maximum). In the IVA concept, the power is added through vector addition of electromagnetic fields rather than connecting a large number of semiconductor switches in series as in standard (Marx generator) technologies. This work comprises conceptual and simulation work. Fundamental components of the generator like the magnetic coupling (“transformer” section), magnetic isolation and the combining of power have been designed on the basis of 3D electromagnetic (EM) simulation work using CST Microwave Suite. Complementary experimental work based on these simulations comprises the verification of a single pulse module [1], single IVA stages incorporating several parallel pulse modules [2], and a multi-module, multi-stage IVA [3]. The IVA concept and results of these simulations are presented.
This article investigates factors affecting the contributions of battery units to fault currents in grid-connected battery storage systems (BSSs). The work in this article is intended to examine the effects of the state-of-charge (SOC) on battery currents that are drawn due to faults. This article also examines the impacts of charger controller actions on the currents drawn from battery units to faults in grid-connected BSSs. The impacts of the SOC and charger controller on battery currents due to faults are examined for the lead-acid, lithium-ion, and nickel–cadmium battery units. Examination results show that the battery currents due to faults are directly dependent on the SOC. Moreover, these results show that actions of charger controller can support the battery terminal voltage, thus preventing the fast reduction of the SOC. The support of the battery terminal voltage helps in limiting the currents drawn from battery units during faults. The effects of the SOC and charger controller are verified using a 1-MW,
$3\phi$
grid-connected BSS, which has lead-acid battery units. Several faults have been created during charging and discharging operations, and at different values of SOC. Test results confirm the direct dependence of battery currents (due to faults) on the SOC. In addition, obtained results demonstrate the ability of charger controller to limit the currents drawn from battery units due to faults in different parts of a grid-connected BSS.
This paper investigates factors affecting the contributions of battery units to fault currents in grid-connected battery storage systems (BSSs). The work in this paper is intended to examine effects of the state-of-charge (SOC) on battery currents that are drawn due to faults. This paper also examines impacts of charger controller actions on the currents drawn from battery units to faults in grid-connected BSSs. The impacts of the SOC and charger controller on battery currents due to faults, are examined for the Lead-Acid, Lithium-Ion, and Nickle-Cadmium battery units. Examination results show that the battery currents due to faults are directly dependent on the SOC. Moreover, these results show that actions of charger controller can support the battery terminal voltage, thus preventing the fast reduction of the SOC. The support of the battery terminal voltage helps in limiting the currents drawn from battery units during faults. The effects of the SOC and charger controller are verified using a 1 MW, 3ϕ grid-connected BSS, which has Lead-Acid battery units. Several faults have been created during charging and discharging operations, and at different values of SOC. Test results confirm the direct dependence of battery currents (due to faults) on the SOC. In addition, obtained results demonstrate the ability of charger controller to limit the currents drawn from battery units due to faults in different parts of a grid-connected BSS.
In this paper, a novel insulated gate triggered thyristor with the Schottky barrier (SB-IGTT) is proposed for improved the repetitive pulse life and high-di/dt characteristics. Different from the conventional cathode shorted MOS-controlled thyristor (CS-MCT), an SB is specially imbedded to enlarge the effective turn-on area and enhance the electron-hole plasma spread during short duration pulse, which contributes significantly to relaxing the thermal concentration and improving the repetitive pulse life as well as achieves superior di/dt characteristics. The experimental results show that the proposed SB-IGTT continuously undergoes more than 220000 shots at the pulse frequency of 5 Hz, yielding a 10× longer repetitive pulse life than the conventional CS-MCT. Simultaneously, SB-IGTT performs a di/dt up to 120 kA/μs with peak current near 10 kA, increasing di/dt by about 20%. Improved repetitive pulse life and simultaneous superior di/dt characteristics indicate that the proposed SB-IGTT is suitable for repetitive pulse power applications.
In this article, a voltage coupling enhancement (VCE) technique is developed to suppress the transient gate overvoltage of an insulated gate trigger thyristor (IGTT) in ultrahigh di/dt pulse applications. It is found that, due to the gate-cathode voltage (V
<sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">G</sub>
-V
<sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">C</sub>
) coupling associated with the intrinsic gate-cathode capacitor (C
<sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">gc</sub>
) and inevitable common source inductance (LC) of IGTT, the gate voltage (V
<sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">G</sub>
) would oscillate with the cathode voltage V
<sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">C</sub>
(= L
<sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">C</sub>
× di/dt), which produces high gate-cathode voltage (V
<sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">GC</sub>
) oscillation. This easily leads to device failure especially at high di/dt pulse condition. Enhanced V
<sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">G</sub>
- V
<sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">C</sub>
coupling by increasing intrinsic C
<sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">gc</sub>
can contribute to the close following of V
<sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">G</sub>
against V
<sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">C</sub>
, suppressing the transient gate overvoltage. Thus, a modified dummy gate IGTT (DG-IGTT) structure with increased C
<sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">gc</sub>
is specially designed as a practical implementation of the VCE technique. Experimental results show that the DG-IGTT has a low gate overvoltage (V
<sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">GC-max</sub>
) of 31.2 V, whereas the V
<sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">GC-max</sub>
of conventional IGTT is over 136.7 V. In addition, the DG-IGTT in Kelvin source package can further achieve a much lower V
<sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">GC-max</sub>
of 18.5 V at ultrahigh di/dt of 32.4 kA/μs. The DG-IGTT successfully works at the repetition rate of 200 Hz, indicating the proposed VCE technique is promising for improving device robustness in ultrahigh di/dt and repetitive pulse applications.
