High-current grounding impedance tests have been specified in safety standards for many years. There are two, independent sources for these tests.
One source is CSA Standard 0.4, Bonding and Grounding of Electrical Equipment, which specifies a test current of 30 amperes for 2 minutes.
The other source, I believe, came from USA requirements for home appliances such as refrigerators, and specifies a test current of 25 amperes for 1 minute.
Both tests measure the grounding circuit impedance at either 25 or 30 amperes. The 25-amp test requires that the impedance be less than 0.1 ohm at the end of the test period. The 30-amp test requires that the voltage drop across the grounding circuit be less than 4 volts at the end of the test period (less than 0.13 ohm).
(A derivation of the grounding impedance limit value, 0.1 ohm, is presented in Technically Speaking, The Product Safety Newsletter, Volume 9, Number 1, January-March, 1996.)
The grounding circuit impedance tests presume that a fault will occur in a product between the mains live (or line) conductor and grounded parts. This fault will subject the grounding circuit to a very high current until such time as the relevant overcurrent device operates and disconnects the mains.
Overcurrent devices (e.g., fuses, circuit-breakers) do not operate at their rating. For example, a 15-ampere circuit-breaker does not operate at 15 amperes. But, it does operate at currents above the rated current. The operating time of any overcurrent device is a curve relating current and time. The higher the overcurrent, the shorter the operating time. Most overcurrent devices are calibrated at twice rated current. The maximum operating time at twice rated current is either 1 minute, 2 minutes, or 4 minutes, depending on whether the device is a fuse or circuit-breaker or circuit-breaker type.
For the purposes of safety, the construction of a product must enable fuse or circuit-breaker operation in the event of a fault from the live (or line) conductor to grounded parts. Therefore, the construction must be capable of carrying twice rated current of the overcurrent device for at least one minute.
This requirement means that the construction of the product grounding circuit must be reasonably robust. It must be at least as robust as the mains circuits.
The resistance of most electrical conductors is directly proportional to the conductor temperature. That is, as the temperature increases, the conductor resistance increases. For the purposes of safety, the grounding circuit resistance must not exceed the specified value, usually 0.1 ohm. To stay below this value, it is imperative that the conductor temperature be controlled.
Conductor resistance is inversely proportional to the cross-sectional area of the conductor. That is, as the cross-sectional area increases, the conductor resistance decreases. To control the resistance, it is imperative that the wire size be controlled.
And, conductor resistance is directly proportional to the length of the conductor. That is, as the conductor length increases, the resistance increases. In most product constructions, the grounding circuit wires are relatively short, so the length usually is not a significant contributor to the resistance.
So we need to consider the three parameters, temperature, cross-sectional area, and length in order to be assured that the grounding circuit resistance does not exceed the specified value, 0.1 ohm.
When a conductor is subjected to a current, some power is dissipated in the resistance of the conductor according to:
Power = I x I x R
Power is the measure of the electrical energy converted to heat in the conductor. So, the action of conducting current causes the wire to heat. That heating in turn causes the resistance to increase. If the value of R gets too high, the system can get into a positive feedback mode, where the R continues to increase until the conductor melts and the circuit opens. If this should occur to the grounding circuit, the product will become unsafe because the overcurrent device will not operate.
Typically, mains cords and wiring is No. 18 AWG (approximately 0.75 square millimeters). When subjected to 25 or 30 amperes, the wire will become warm, possibly even too hot to touch. But, it will not exceed the 0.1 ohm limit value. And, it will not cause its insulation to melt.
On the other hand, No. 22 AWG, when subjected to 25 amperes, will get too hot to touch, and will cause its insulation to melt. It will also melt the insulation of adjacent wires, causing unpredictable consequences of short-circuits. After 2 minutes at 30 amperes, No. 22 AWG is likely to melt.
The high-current grounding circuit test is a good test to confirm the adequacy of the design of the grounding circuit. The issue I want to address is whether the high-current test is useful as a production-line test.
Almost all safety certification houses demand a production-line test of the grounding circuit. Most do not specify the current for the production-line test. However, some safety certification houses demand that the 25-ampere test be a production-line test.
Can a 25-ampere production-line test find manufacturing defects that cannot be found by a low-current test?
One manufacturing defect is damage to the cutting or breaking of individual strands of a grounding wire when crimped or connected to a terminal. We simulated a bad crimp by cutting individual strands of a 36-strand, No. 18 AWG wire. We stripped the insulation from about 3 mm of the wire, cut one strand and peeled the two ends back to the edges of the insulation. Then, we measured the resistance using an ordinary digital ohmmeter, and the impedance using a high current sourcing milliohmeter (ac). We applied the 25 amperes for two minutes, taking millohmeter readings every 10 seconds.
We repeated the test, cutting one strand at a time.
During the two minutes, the resistance would increase. This is expected because the wire is heating due to the power dissipated in the wire.
As we continued to cut individual strands, the initial resistance increased, but not significantly.
The ohmmeter resistance indicated 0.2 ohms for
every test.
The high current sourcing milliohmeter indicated less than 0.1 ohm throughout every test until we cut the 31st strand. With four strands remaining, the strands melted at about 1 minute. (No. 18 AWG wire is comprised of 36 strands of No. 34 AWG copper wire.)
By comparison, No. 30 AWG in free air will melt at
5 amperes.
How can 5 strands possibly carry 25 amperes for
2 minutes without melting?
The answer is that the 5 strands were only 3 mm in length. They were well heat-sunk by the remaining strands, which were held next to the 5 strands by the surrounding insulation. Copper is a very good thermal conductor. The heat-sinking kept the strands from reaching the melting temperature.
This test shows that the 25-ampere test is not likely to find a wire with cut or broken strands as may occur due to defective crimping or due to excessive bending.
ACKNOWLEDGMENTS
Thanks to Eric Davis of Hewlett-Packard’s San Diego Division for testing the wire.
This work duplicates work done by Hewlett-Packard’s Vancouver Division in 1985, supervised by Ken Curtis.
Copyright 1997 by Richard Nute Originally published in the Product Safety Newsletter, Vol. 10, No. 1, January – March, 1997
