Story of the AFCI Tester
Arc Fault Tester from Fox Meter Inc.
A commentary by John Grant, 8/15/2002
In the spring of 2002, my wife and I had an addition built onto our house. During one of the electrical inspections a conversation with the electrical inspector ensued. He was curious about my background and knowledge of residential wiring. After he learned that we owned Fox Meter Inc. (an industrial instrumentation manufacturer), he mentioned that we should take a look at designing an arc fault circuit breaker tester. He said that he had been trying to locate a tester and as far as he knew, none existed. The journey begins.
At this point in time I had never heard of an arc fault breaker. Some quick searching on the Internet provided a beginning education into AFCI technology. Having been in electronics engineering for the last 30 years, I figured it would be an easy task to test these breakers with an arcing condition and view the resulting current signature on an oscilloscope and determine what conditions would cause them to open.
A trip to our local electrical supplier produced AFCI breakers from Square D, Cutler-Hammer and G.E. (expensive little devils). Back at our lab, I set up some tests using a load bank and some crudely constructed arcing devices, aluminum rod with wood rasp, carbon arc rod with copper and charcoal with stainless steel. With breakers wired into our load center we proceeded to produce all kinds of arcs from 5 amps to 15 amps (series arcing). I produced arcs that caused the fixture to burst into flames and nothing happened with any of the breakers. The only way that I could cause a trip was with a 50 mA ground current or to exceed the thermal magnetic rating of the breaker. My immediate assumption was that I was doing something wrong, so I enlisted some engineering friends to evaluate my test procedures. The result of this exercise was more head scratching and questioning the integrity of the fore mentioned firms. I also gained a little insight into why no AFCI field testers existed.
Being an engineer, curiosity is an obsession, so I proceeded
to try and talk to the manufacturers of the breakers, no information was forthcoming.
We purchased UL 1699 for $350 (over priced for what you get). After reviewing UL 1699
it was our opinion that the specifications verified that our tests should have caused
the breakers to open. We later learned that the review panel for UL had not accepted UL
1699. There was no mention of this in the UL specification. UL 1699 primarily deals with
series arcing conditions and unwanted nuisance tripping. The only specification that seemed
to apply is the 5 amp ground current arcing condition. Even this specification seems meaningless
because all the breakers we tested tripped at 50 mA. From what we had learned to this
point it was our opinion that it would be virtually impossible to design a load center
mounted AFCI breaker that would sense a series arc.
As a side note, during this time our chief engineers home caught fire. Thank God no one was home at the time. The fire started in a UL approved Childs lamp fixture, it could not be determined if it was a series or parallel arc that caused the fire. The standard 15 amp thermal magnetic breaker had tripped, but all of the wiring and fixture were totally destroyed. We of course asked ourselves if an AFCI breaker could have prevented the fire. Our answer was probably not. If the fire started as a series arc (electricity is the source of ignition and not the fuel for a fire) an AFCI breaker would not have helped. If a series arc caused heating to the point that insulation melted and a parallel arc or more likely a short developed, then an AFCI could be a life saver.
We were on a mission now, we were more determined than ever to trip one of these breakers with an arcing condition, but we still did not know enough to do this. More searching on the web produced a white paper from Square D that explained that these breakers are designed to trip on parallel arcing conditions in the range of 75 amps. Our justification and determination follows. Given sound wiring practices there are two fault conditions that can affect a branch circuit. The first is a short between line and neutral or ground, this usually results in the immediate tripping of the thermal magnetic portion of the breaker, the magnetic solenoid is the immediate response device backed up by the slower response thermal device. The second fault condition is the sputtering, sizzling parallel arc that can be a source of ignition. The problem with a parallel arc is that the high current draw is not continuous, rather a voltage sensitive condition that causes a momentary short that clears itself until the next cycle. These current pulses can be as high as the load center is capable of delivering, most homes are 200 amps, yet the average power at the breaker is well below its rating. This is the fault condition that AFCI breakers are designed to detect. Duplicating the conditions of a parallel arc can be mind numbing, but stuff happens in the real world.
At this point we took a deep breath and decided to build up a test bed that utilizes what we refer to as Resistive Pulse Technology (RPT). With this tester we were able to generate clean, resistive, bipolar loads form 5 amps to 100 amps and pulse widths form 10 micro-seconds to 1000 micro-seconds. The Cutler-Hammer and G.E. breakers started tripping at 35 amps @ 200 micro-seconds and the Square D tripped at 75 amps @ 300 micro-seconds. These tests became the foundation for the AS1000 design.
The AS1000 is a small (3 X 7 X 1.5), rugged, hand held unit with a standard removable power cord. The unit in its most basic sense is a high wattage, precision resistive load under embedded microprocessor control. Its primary function is to simulate a parallel arcing condition that is a must trip condition for load center installed AFCI breakers. The unit performs four tests. First, there are the standard indicators for a properly wired outlet, we use three LED displays. Second, is a push button that initiates a nuisance test. This test is a 2 second series of 120 Hz, 25 amp, 200 micro-second pulses. An AFCI breaker that trips during this test may be too sensitive and may trip with noisy appliances. Third, is the arc test, this is a half second series of 120 Hz, 75 amp, 300 micro-second pulses. The AFCI breaker should trip as soon as the button is pushed. The last test is the ground test. This test consists of a half second series of 120 Hz, 5 amp, 200 micro-second pulses into the ground wire. The AFCI breaker should again trip as soon as the button is pressed. The 5 amps is a UL specification, all the breakers we tested tripped at 50 mA. This unit will test 15 amp and 20 amp breakers.
The AS1000 sells for $167.00 and delivery is running out to 4 weeks due to the large demand. We are doing things that should improve deliveries to stock to 1 week. Fox Meter Inc. is a small woman owned business; all of our products are manufactured in the United States of America with pride, by American workers.
This unit is due to be available in the 3rd quarter of 2002. It will include all of the features of the AS1000 with the exception of the arc and ground tests. The AS2000 will have automatic ramp tests for both arc and ground currents with bar graph displays for the associated trip currents. The arc current will ramp from 5 amps to 100 amps making the tester suitable for series and parallel arcing. Also the ground current ramp will go from 1 mA to 100 mA making the tester suitable for AFCI and GFCI testing. The selling price should be in the $200 to $400 range. All and any comments from the field are welcome.
This unit is planned for later in 2002 and will incorporate
all of the features of the AS2000 with digital displays instead of bar graph displays.
Also, the unit will incorporate a line loss function that utilizes a high speed data acquisition
capability that will stress wiring and connections in a branch circuit up to 100 amps.
This should be an invaluable tool to determine the health of house hold wiring especially
in the older home.
The arena of AFCI breakers seems to be clouded in mystery and misinformation. UL seems confused about how to spec these devices, NEC requires their use and manufacturers are hand-cuffed by a litigious society that demands total safety and security. I believe that manufacturers have done their best to bring an added dimension of safety to house hold wiring. We believe that the development of outlet AFCI breakers that can sense series arcing at the outlet and its loads is possible and we believe that all the manufacturers are working on such a device. We would like to see the addition of an indicator at the AFCI breaker that would display that an arc trip occurred. All in all, when you step back and look at the entire US electrical distribution system, from power generation to the outlet in a small childs bedroom, we enjoy a level of safety and reliability that is unheard of in todays world. There will always be confusion when technology touches everyday lives and it is our sincere desire to provide good test systems and information to the professionals in the industry.
Mike's Comment: Only in America can this happen. Some guy who knows nothing about the NEC and UL comes up with an AFCI tester before the big testing companies!
Copyright © 2002 Mike Holt Enterprises,Inc.