28-09-2012, 12:22 PM
AN IMPULSE GENERATOR SIMULATION CIRCUIT
AN IMPULSE GENERATOR.pdf (Size: 2.7 MB / Downloads: 323)
ABSTRACT
This thesis describes the creation of a simulation circuit to match the output of a Marx type Impulse Generator. The goal was to estimate the stray capacitance and insert that capacitance into the simulation circuit to effectively produce an output similar to that of the generator. An actual three-stage impulse generator was used as the base. Several different levels of impulse voltage were tested, and the output waveforms were captured. Research was conducted to formulate the stray capacitance and identify the locations of these capacitances in the generator itself. The simulation circuit was then subjected to several iterations, adjusting the capacitance values to attain an output as close as possible to that of the actual generator.
Conclusions of the research indicate that an effective simulation circuit can be created to give an output that is close to, but not exactly that of, the actual generator. In the research, several areas of error were identified in the actual generator that were not present in the simulation circuit. These areas are discussed in the thesis.
HISTORY AND BACKGROUND
Introduction
The purpose of this research is to develop a SPICE(1)* simulation circuit that will generate a Marx-type Impulse Generator (IG) output wave shape. To conduct this study an existing IG set at the Cutler-Hammer Technology & Quality Center (TQC) in Pittsburgh will be utilized for the experimental results and the base for the simulation to match. This being the case, the calibration of this IG will also be consulted for the various correction factors on this equipment. This calibration is performed yearly by Dr. Roy Voshall of Gannon University. The Cutler-Hammer IG is used to perform design testing of electrical distribution equipment rated from 600 to 38000 volts.
Full wave and chopped wave impulse tests are methods to demonstrate the ability of high-voltage equipment to handle lightning strikes and switching overvoltages, as defined in IEEE Standard 4.(2) In order to evaluate the complete capability of the equipment to withstand these surges, the entire waveform is required. The full waveform can be captured and used in this evaluation by the use of an oscilloscope and a voltage divider.(3) A calibration method is used to maintain the test equipment and certify its ability to perform this testing; this method is IEEE Standard 4.
There are many factors that lead to error in the measured impulse waveform. The IEEE Standard 4 is meant to maintain a level of acceptable quality in the testing. Each impulse system is typically tailored for the class of equipment to be tested.
Impulse Generator and Equipment
The Impulse Generator used in the testing was a Hipotronics™ Series 100. It is a three-stage generator resulting in a peak capability of 300kV. The control system installed as part of the impulse generator is a Hipotronics™ model 970IG-DS, and is a programmable digital design. An impulse generator has two resistors per stage. The front resistor allows the front of the wave to reach peak in the desired time. The second resistor, referred to as the tail resistor, is required for the half-voltage level at the end or tail of the wave shape. Each stage front resistor value is 17.65Ω; the tail resistor value is 120Ω.
The oscilloscope is a Tektronix™ model TEK 544A. It is a digital oscilloscope with dual trace memory capability. The oscilloscope is connected to the voltage divider using a Tektronix 100X probe, Model P5100.
Divider Construction
The construction of a resistive voltage divider for high voltage measurements appears straightforward at first consideration. Typically, it consists of two resistances in series whose divider ratio reduces the applied value of voltage to a lower voltage value that is measured by an oscilloscope, as shown in Figure 3. Since this thesis deals with fast impulse types of voltage signals, those characteristics that result in measurement error will be addressed.
At the top of the divider is a toroidal shaped electrode that is used to alter the geometry, effectively controlling the electric field to reduce the gradient at the surface of the divider top. This toroidal electrode is normally of large diameter and is sized for the peak voltage rating of the divider.
In applications of impulse voltages, the stray capacitance of the resistor must be considered. The resistance ratio is to be constant over a wide range of frequencies during the impulse to eliminate distortion in the waveshape. The construction of the resistor is such that the inductance and capacitance are reduced to minimal levels by using wire woven into glass-fiber fabric and winding the resistor non-inductively.(7) For thermal reasons, wire is utilized to handle the high rate of energy transfer to the resistors from the generator. In this construction a capacitance exists from the high voltage arm-to-ground. A typical divider with the resistors stacked vertically has a value of capacitance-to-ground from 15 to 20pF/m of height.
Calibration of Divider
The output voltage of the impulse generator is typically calibrated by using sphere gaps. The spheres are of standard dimensions. The most commonly used is a diameter of 25 cm. One of the spheres is grounded, and the other is connected to the high voltage side of the impulse generator. The relative humidity and temperature of the test cell at the time of test determines the required gap for a given voltage level. Three successive readings are made at the test level. These readings cannot vary by more than +/- three percent. Setting the gap of the spheres based on the voltage level and factors of relative humidity and laboratory temperature will produce a breakdown at a known voltage across the gap.(9) In this arrangement, the oscilloscope is utilized with the divider to record the breakdown voltage. After three readings that are within tolerance, the voltage level is typically slightly decreased for one shot. There should be no breakdown across the gap.
Although the above method is used primarily to test the ability of the impulse generator to produce the accurate desired voltage level, it also verifies the divider-to-oscilloscope measurement equipment as well.
Losses and Error
Errors and loss factors affect components in the system other than just the voltage divider. The types of errors are those involving the coaxial cable connection from the divider to the scope, electromagnetic field disturbances coupled directly to the scope, connection between generator and divider, stray ground capacitance in the divider, and the divider ratio. In this section, the error components for the entire system will be discussed.
The high voltage is reduced by the divider and transmitted to the oscilloscope via a coaxial cable connection. The cable connection is made on the low voltage arm of the divider, and this cable acts as a distributed transmission line. A traveling wave will be reflected from the end of the line if there is a difference between the characteristic impedance, Zo, and the terminating impedance. For this reason, the connections at both ends of the coaxial cable are made with an impedance that is equal to the cable characteristic impedance, Zo. Also, it has been found that the frequency-dependent transmission error can be decreased by shortening the cable connection.