Virtual instrumentation (VI) is a hot topic in the test and measurement industry today. The reason is simple: It is a user-defined rather than a vendor-defined technology. You are in control, using mainstream computers, off-the-shelf plug-in instrumentation cards and software to tailor a test system to meet your exact needs.
The benefits of VI can be many. To illustrate the ease of use, flexibility, and time and money savings afforded by VI, we’ll take you through the steps of integrating a VI system for ballast testing.
The Application
Our client manufactures magnetic-type high-pressure sodium street lighting ballasts. To design and develop ballasts for North American and international markets, a setup for product testing must accommodate:
Different types of core and coil ballasts such as reactors, autotransformers, constant-wattage autotransformers and constant-wattage isolated transformers.
Operating voltages from 120 V to 600 V and rated lamp wattages from 50 W to 400 W.
Capacitors for wattage control and/or power-factor correction.
Different lamp ignitors.
Open-circuit, short-circuit, lamp-starting and lamp-running tests.
Measurements at the ballast input and output ports must represent true rms values of current and voltage, true power and the ratio of watts to volt-amperes (power factor if the voltage and current waveforms are clean sinusoids). High-pressure sodium lamps are nonlinear loads, so current and voltage peak values and crest factors must be monitored along with total harmonic distortion.
System Integration Approach
Using conventional test equipment to establish a test bench with the required functionality was difficult with the tight budget of a growing company. To achieve these objectives, we opted for a VI approach.
Software
The LabVIEW® development platform from National Instruments served as the heart of the system. IET’s BallastVIEW application program was tailored to acquire signals, process data and present results to the user on the computer screen.
Hardware
Figure 1 shows the circuit diagram of the ballast instrumentation system composed of:
A 486 DX2-66, 12-MB RAM, 340-MB hard disk IBM-compatible PC running LabVIEW on Microsoft Windows.
A variac supplying AC power to the ballast under test.
A system resource interface (test fixture) containing switches and wiring required for the different test arrangements.
Transducers with suitable frequency response characteristics to sense current and voltage signals. Resistive voltage dividers and current shunts are well suited for this application.
RC anti-aliasing filters. The filter components were selected to avoid loading the board input amplifiers. Filter cutoff frequency halved the sampling rate (Nyquist frequency), which is necessary when performing Fast Fourier Transforms on digitized signals. RC filters protected electronics from the high-voltage spikes generated by the ignitor to start the lamp.
Analog Devices’ 5B signal-conditioning modules to amplify and isolate the filtered signals.
National Instruments’ Lab-PC+ to digitize the conditioned signals. The board was configured for four bipolar differential inputs. The sampling frequency was 7,680 Hz/channel. Acquisition was software-triggered on the rising slope of the input voltage.
Results
The front panel of BallastVIEW illustrates the test results of a 100-W high-power factor reactor-type ballast (Figure 2). The computer screen shows four VIs representing an input power analyzer, an output power analyzer, a spectrum analyzer and an oscilloscope. The digitized signals are multiplied by scaling factors (top left side of the screen) to reflect their actual value.
For the power analyzers, the power and power factor are displayed in the right column of the screen. The indicators (from left to right in each row) display the rms; maximum, minimum and peak average; the crest factor; the fundamental harmonic magnitude and the total harmonic distortion of each signal.
The normalized harmonic spectrum is shown in the bottom left corner. Each harmonic is normalized to the fundamental component of the respective signal. The percent harmonic magnitude can be checked by flipping the cursors of the harmonic magnitude indicator (bottom center).
The waveform graph (bottom right) shows the square wave of a typical arc voltage of a high-intensity discharge lamp. The input current is the signal rich in harmonics.
A shunt capacitor, connected across the ballast input terminals for power-factor correction, acted as a low-reactance path to high-frequency harmonic currents generated by an arc welder in the client’s factory. For a 60-Hz input, the magnitude of the odd harmonics between the 47th and 63rd was approximately 8%.
The credibility of the system was verified. The measurements agreed with test results from an independent test laboratory, utilities and customers of the ballast company.
Conclusion
The combination of an IBM-compatible plug-in data acquisition board, signal conditioning modules, LabVIEW and the BallastVIEW application program, results in a flexible, high-performance, easy-to-use and cost-effective system to test ballasts. The core of the BallastVIEW program constitutes the cornerstone to build applications for use in testing other products such as transformers, rectifiers, inverters and UPS.
About the Author
Ahmad Sultan, Ph.D., is the President of IET. He received B.Sc. and M.Sc. degrees in electrical engineering from Ain Shams University, Cairo, Egypt, and a Ph.D. from the University of Manitoba. Dr. Sultan worked at ASEA for three years and was an assistant lecturer at Ain Shams University from 1983 to 1988. IET Integrated Engineering Technology, 1627-77 University Cres., Winnipeg, Manitoba R3T 3N8, Canada, (204) 269-8063
Copyright 1996 Nelson Publishing Inc.
April 1996
|
|