Pulse Width Modulation Simulator for Testing Insulating Materials

This invention relates to test instruments or testing life expectancy of magnetic wire used in dynamoelectric machines. While the invention is described with particular reference to the test instruments ability to test magnet wire intended for use in dynamoelectric machines associated with inverter controls, those skilled in the art will recognize the wider applicability of the inventive principles disclosed hereinafter.

It is well known that degradation of wire insulation can be caused by thermal, electrical, mechanical, or environmental stress, alone or in combination with one another. Thermal stress is introduced by aging, overloading and cycling. Electrical stress is related to dielectric, cracking, corona and transient spikes. Mechanical stress results from impact, stretching, and coil movement during coil winding, assembly, and machine operation. Environmental stress is introduced by moisture, chemical and foreign objects in the environmental area of the dynamoelectric machine. All of these stresses can be, and often are, present at the same time in applicational use. Motor manufacturers have long recognized the effects these factors have on magnet wire.

With the increased use of pulse width modulated (pwm) type of adjustable speed drives in the motor industry, the incidents of premature failure in wire insulation has increased. Dynamoelectric machines being controlled with pulse width modulated variable speed drives no longer experience the traditional sine wave voltage, which is a steady state condition with a maximum and a rms value. Rather, the machines experience a pulse width modulated voltage with significance harmonics and transients. We have found that motor failures in inverter applications result in premature failure of the wire insulation. To understand the failure mechanism of magnet wire insulation under such condition, we have devised a test instrument which simulates inverter output in order to evaluate wire performance in that environment.

The basis of the thermal aging model is a well known Arrhenius chemical reaction rate model: Life=A.times.e.sup.-(BT), where A and B are the constants determined by the activation energy in the reaction rate of the particular degradation reaction, and T is the absolute temperature. This model neglects threshold effects (aging starts only if a certain temperature is reached), and the model assumes that only one chemical reaction is occurring. Nevertheless, this model presently is an accepted standard in industry.

Voltage aging models also have been developed. The voltage aging models conventionally are divided into two groups, the inverse power model and the exponential model. The inverse power model of electrical aging is expressed as: Life=K.times.E.sup.-n, where K is a constant and a characteristic of the material system, n is an experimentally obtained constant depending on operating conditions, and E is the applied voltage.

An alternative, exponential life model is given by: Life=C.times.e.sup.-DE, where C and D are constants, and E is the value of the applied voltage.

The basic models, given by the above equations, have been changed in order to allow for the description of threshold degradation, or combined in order to provide a model for multi-factor aging.

We were aware of the research being conducted in this area. Motivated by the number of early failures occurring in industrial applications of power inverter drives, we have designed a testing system that allows for the varying of the parameters that are characteristic of power inverters, i.e., voltage, frequency, duty cycle, and rise time in combination with the conventional thermal aging problem inherent in insulation dialectics applied to magnet wire, or other portions of the insulative system applied to a dynamoelectric machine. As a result of these investigations, an improved insulation system for a dynamoelectric machine has been developed, as described in copending application Ser. No. 08/527,358, filed Sep. 12, 1995, the disclosure of which is intended to be incorporated herein by reference.

One of the objects of this invention is to provide a test instrument which permits improved testing of magnet wire intended for use in dynamoelectric machines.

Another object of this invention is to provide a test instrument which tests wire which is intended for use in applications having pulse width modulated drive systems.

Another object of this invention is to provide a test instrument which permits variations of pulse width, voltage, rise time and temperature applied to the test sample.

Another object of this invention is to provide a test instrument which permits testing of both voltage and/or current variation and a test sample.

Another object of this invention is to provide a test instrument which permits easy modification of the testing procedure for implementing additional factors which may affect insulation life.

Another object of this invention is to provide a test instrument which is relatively easy to operate.

Other objects of this invention will be apparent to those skilled in the art in light of the following description and accompanying drawings.

In accordance with this invention, generally stated, a test instrument is provided for duplicating inverter drive stress on magnet wire by permitting variations of temperature of the wire, and peak voltage, frequency and rise time of the electrical characteristic applied to the wire. A method of testing wire for use in dynamoelectric machines also is disclosed.

In the drawings, FIG. 1 is a block diagrammatic view of one preferred embodiment of the test instrument in the present invention;

FIG. 2 is a diagrammatic view of the wave form characteristic for the instrument shown in FIG. 1;

FIG. 3 is a block diagrammatic view for the current sensing circuitry portion the instrument shown in FIG. 1; and

FIG. 4 is a diagram of an integrator circuit for varying the rise time of the voltage applied to the test specimen and the device shown in FIG. 3.

Referring now to FIG. 1, referenced numeral 1 indicates one illustrative embodiment of test instrument of the present invention. The test instrument 1 includes a computer 2 adapted to include an analog to digital and digital to analog converter card 3, a digital input/output board 4, and a general purpose interface bus (gpib) 5. The computer 2 may comprise any of a variety of commercially available devices (PC's) modified as described above. Such devices commonly include a monitor 100, a keyboard 102, and a suitable memory 104 operatively arranged in the manner shown in FIG. 1 and described below. The memory 104 stores a program for controlling the operating characteristics of the test instrument 1.

The computer 2 has an output 10 from the analog to digital converter card 3 to high voltage power supply 12. An output 14 of the power supply 12 is operatively connected to a high speed switch array 16. The high speed switch array 16 may be constructed from integrated gate bipolar transistors (igbts) or metal oxide semiconductor field effect transistors (mosfets), depending upon the intended applicational use of the motor employing the magnet wire in the test. Other switch arrays are compatible with the broader aspects of this invention.

The switch array 16 has an output side 17 electrically connected to a test specimen 20, which may be environmentally segregated with respect to temperature in a heat chamber or oven 22. Also electrically connected to the specimen 20 is a current sensor 24. The current sensor 24 has an output side 26 electrically connected to a sample and hold device 28. An output side 30 of the sample and hold circuit 28 is electrically connected to the analog to digital and digital to analog converter card 3 of the computer 2. The digital board 4 has an output 40 operatively connected to the array 16.

An oscilloscope 32 is connected to the output side 17 of the high speed switch array 16 along a conductor 34. An output 36 of the oscilloscope 32 is connected to the interface card 5.

The operation of the instrument 1 is under control of the computer 2. A user can set up the parameters of the test wave form, such as peak voltage, frequency, duty cycle and rise time. Two control voltages are generated as a consequence of these settings. The DC voltage at the output 10, generated by means of the analog card 3, corresponds to waveform B in FIG. 2. The voltage of waveform B controls the output level of the bipolar high voltage power supply 12. Data also is entered at the computer 2 for controlling the duty factor and the frequency of pulses to be applied to the test specimen 20. This information is converted into the trigger pulse train, which is generated by means of the digital input/output card 4, and corresponds to the waveform A shown in FIG. 2. The high voltage from the power supply 12 is fed to the switching devices of the high speed switch array 16. The array 16 is controlled by the trigger pulses on the line 40. The array 16 generates the bipolar pulse train, waveform C in FIG. 2 at its output 17. Waveform C of FIG. 2 represents the output pulse train that is applied to the test specimen 20. The test specimen, which may be a twisted pair of wires, or a motorette in situations where insulated material used for phase to phase or phase to ground insulation also is being tested, is located within the heat chamber 22. The instrument 1 generates the output pulse train C that is fed to the test specimen 20 as long as the integrity of the insulation under test is preserved.

Breakdown of the insulation may be detected by a number of ways, two of which are preferred. The first method is based on voltage failure detection. The second method is based on current detection.

In the first method, voltage pulses are monitored using digital acquisition of the output pulses C at the output 17 of the array 16 by means of the oscilloscope 32. Once the output pulse is captured, it is transferred to the memory of the computer in digital form, using the general purpose interface bus or other appropriate parallel or serial links. The voltage wave form also can be digitized using fast analog to digital converters. The control program in the memory of the computer 2 compares the voltage waveform with a template generated at the beginning of the test for the sample under test. If the voltage deviates in amplitude by more than 20%, the program recognizes a failure of insulation. The percent setting for recognizing the failure of insulation is adjustable, and may be varied depending upon use of the magnet wire under test.

Referring now to FIG. 3, a block diagram for detecting insulation failure based on current sensing scheme is shown. Like numerals are used to describe like parts, where appropriate. The computer 2 is designed to trigger pulses which initiates the opening of suitable switches in the high speed switch or pulse generator 16. The charging current from the pulse generator charges a capacitor in the form of a twist pair wire specimen 20. The charging current of the capacitor, shown in FIG. 2 waveform as D, is transformed into a voltage signal by means of the current sensor 24. The voltage from the current sensor 24 is fed to the input side of the sample and hold circuit 28. A hold pulse for the sample and hold circuit 28 is generated by a variable delay circuit 42, which is controlled by the same pulse that was used for triggering the pulse generator 16. The delay between the rise (or falling edge) of the trigger pulse and the rise (or falling edge) of the hold pulse is controlled by means of the variable delay circuit 42, shown in FIG. 3. A network 47 is inserted on the output side 17 of the generator 16. In this manner, the current is always sensed at the same instant with respect to the rising edge of the trigger pulse. The sample and hold circuit 28 transforms the narrow current charging pulse into a direct current voltage level, FIG. 2 waveform E, that is convened into digital data using standard analog to digital converter 3. The time for the conversion is limited by the frequency of the pulses, and not by the fast rise time of the pulse. The digital data from the analog to digital converter 3 is fed to the computer 2. The computer monitors the change of amplitude of the current, and provides an indication of insulation failure if a certain predetermined value of the current is reached. The predetermined value of current is adjustable under program control at the computer 2.

As indicated above, change in the rise time is achieved by adding the RC network 47 to the output side 17 of the pulse generator 16. As will be appreciated by those skilled in the art, the circuit is a simple integrator, made up of a resistor 44 and a capacitor 46. The time constant of the circuit is determined by the product of the resistance of resistor 44 and the capacitor 46 values. The higher the time constant of the circuit, the longer is the rise time of the pulses used in the test.

Regardless of the method used to monitor failure in insulation, computer 2 tracks the time to failure. Once the failure is registered, analysis of data stored in the computer is accomplished.

Our experience has shown that there is a strong correlation between the life of magnet wire insulation, and other insulation employed in dynamoelectric machines being operated by inverters and the exponential model described here and utilizing voltage, temperature, and frequency of the pulses as variables. We also have observed that decreasing the voltage for the frequency under certain threshold values has positive effects on the life of the insulation. As indicated, we have found that voltage, temperature, and frequency, in that order, are the most important factors influencing insulation life.

Numerous variations, within the scope of the appended claims, will be apparent to those skilled in the art in view of the foregoing description and accompanying drawings. Thus, other types of pulse generators or switching devices may be employed in various embodiments of this invention, in addition to or in place of mosfet or igbt devices described. Likewise, while a range between 60 and 20,000 Hz has been described as the operational range of the switching array, other pulse width modulation designs may be employed, if desired. Other electrical devices are available for capturing data of the test specimen. The oscilloscope and current sensor is described in conjunction with FIGS. 1 and 4 may be replaced by other devices, if desired. While single line representations are employed to interconnect various component parts of our invention for the sake of simplicity, those skilled in the art recognize that the single line representations may be multiple electrical connections in actual embodiments of the invention.

These variations are merely illustrative.