May 1,'00
Circuit Breaker Testing

New electronic circuit breakers with arc detection capability require new automated testing procedures.  Arcing currents are characterized by arbitrary on/off times and high frequency components.  These complex waveforms can be generated by computerized test equipment but the power levels formerly required made this very difficult and expensive.

New techniques have been developed by Zlan for manufacturing and testing electronic breakers which greatly reduce the required test power.  With these new techniques test equipment can now be implemented economically using conventional technology.  These new procedures make it much easier to evaluate changes in the design, thus reducing development cost.  Also, it is now possible to do comprehensive testing on breakers after assembly to assure they conform to specifications, thus providing a safer and more reliable product for the consumer.  All of these advantages are obtained without increasing the manufacturing cost. 

Electronic circuit breakers with arc fault detection capability require new test methods to assure they are capable of performing as intended.  Since the arcing signals are random by their very nature, it is necessary to use computer controlled waveforms in order to get repeatable results.  If the testing waveforms are not repeatable, it is extremely difficult to evaluate the effects of design changes and the performance of the breakers cannot be assured with confidence.  Some typical arc waveforms are shown below.

Figure 1 - The need for testing with controlled waveforms is illustrated in this example of an actual waveform generated by a knife-edge cutting into an extension cord. The peak current was 100 amperes. The test began with a large current for one-half a cycle and then there was zero current for more than 500 msec. If the test timer started at the first current pulse, it would appear that the breaker's trip time was more than 800 msec. In actuality it would have tripped in less than 500 msec if the current had not have been zero for so long. Unpredictablity of real-world arc current makes this type of testing very tedious and time consuming when tests have to be repeated an indeterminate number of times to get valid data. The question is "what is typical?" Arcing waveforms used for testing need to be carefully studied off-line to make an objective determination of their usefulness.

Figure 2 - Carbon rod arcing characterized by high frequency aberrations.

Figure 3 - Current waveform due to a loose wiring connection. Both high and low frequency aberrations are present.

Until recently automation of the arc testing process was hindered by the difficulty of generating complex high current waveforms at high voltages.  The peak power required, which might be on the order of several kilowatts, is hard to control with the fidelity necessary to represent arcing currents.  Triacs and SCR's make good high power switches but they don't have the ability to turn off at arbitrary times during the cycle.  Unfortunately, analog amplifiers with this power capability would be quite expensive and difficult to construct.

Steady-state load currents for testing the basic current overload performance can be defined by only a few parameters, including: amplitude, phase and conduction angle.  This makes the testing of breakers using popular thermal and/or magnetic principles fairly straightforward by using triacs or SCR's to control the test waveforms.  On the other hand, arc waveforms are much more complex and theoretical models may not be satisfactory.  Fortunately it's possible to obtain the desired test repeatability by digitally storing the waveforms of real arc signals and then recreating them electronically.  This results in applying exactly the same stimulus to the breaker that it would see in a real-world arcing condition.  The advantage of this approach is that test results are repeatable and comparisons can be made during product development to evaluate the effects of changes.  Also, various products from different manufacturers can be compared objectively by testing laboratories.

  1    2   3    [Next]   Home 

© 2000 Zlan Ltd.