Clamp Meter Around the Whole Cord: Three Measurements That Proved Nothing
A technician attends a flat where a 1.5-ton inverter split air conditioner (nameplate: 5.9 A running, 230 V, 50 Hz) trips its 20 A MCB once or twice a week, usually in the afternoon. He clamps his digital clamp meter around the three-core flexible cord of the indoor unit - earth conductor included - on the 40 A AC range and reads 0.2 A. He tells the owner the air conditioner is drawing almost nothing and the MCB must be faulty. He then switches off the MCB, puts his multimeter on the 200 Ω range across live and neutral at the 16 A socket and reads 9.4 Ω, concluding that the circuit is healthy. He also uses the multimeter's continuity buzzer to "test insulation" and gets no beep. Before leaving, he replaces the 20 A MCB with a 32 A one "to stop the tripping".
1. Critique each of the three measurements. What does each one actually measure, and why is the conclusion drawn from it wrong?
Read the worked answer
1. Clamping the whole three-core cord
The jaws enclosed live, neutral and earth together. The live and neutral currents are equal and opposite, so their magnetic fields cancel inside the toroid; the meter can only see the small difference - leakage or earth current. 0.2 A is therefore roughly the leakage, not the load. The reading is meaningless as a load measurement.
Also, the indoor unit's cord does not carry the compressor current in most splits, and an inverter drive draws a non-sinusoidal current. A mean-responding clamp meter under-reads such waveforms badly; only a true-RMS instrument gives a usable figure, and even then the compressor is on the outdoor unit's supply.
2. Multimeter across live and neutral at the socket
With the MCB off but the neutral not isolated, and with the air conditioner, its electronics and any other load still connected, the meter reads the parallel resistance of connected loads, not the circuit's insulation. 9.4 Ω is consistent with electronics and a fan winding. It is not an insulation test at all, and it was performed without proving the circuit dead.
3. Continuity buzzer as an insulation test
A multimeter applies only a few volts. Insulation that breaks down at operating voltage can look perfect at 3 V. "No beep" says nothing about insulation resistance; the test must be done at 500 V DC and expressed in megohms.
4. The 32 A MCB
This removes protection: the 2.5 mm² circuit likely has an Iz around 20 A or less, so the cable can now overheat and the fault current required to trip the device rises. It also leaves the original fault undiagnosed.
2. Set out the correct diagnostic sequence for the intermittent tripping, naming the instrument for each step.
Read the worked answer
Correct sequence
- Gather the pattern. Trips in the afternoon only, once or twice a week: that suggests heat - a distribution board on a sun-facing wall, a conduit in a hot loft, a condenser coil that is dirty or a unit that is short of airflow. Ask what else runs at that time.
- True-RMS clamp on the live conductor only, on the outdoor unit's supply: record running current against the 5.9 A nameplate and watch the start current for several starts. Add a power meter if the waveform is inverter-driven.
- Check the MCB itself. Measure the ambient at the board with a thermometer; an MCB in a hot enclosure derates. Look for a loose busbar screw or a discoloured terminal. Test the device's thermal calibration or swap it with a known good one of the same rating.
- Insulation resistance at 500 V DC (insulation tester): L–E, N–E, L–N ≥ 1 MΩ, everything disconnected. This is the test the buzzer replaced and failed to replace.
- Loop impedance and prospective fault current to confirm the 20 A device can actually clear a short circuit at the far end.
- RCCD trip test at IΔn and 5 × IΔn.
- Then the appliance itself: condenser cleanliness, fan operation, refrigerant charge, run capacitor, and the compressor's start current. High current in hot weather with a dirty condenser is far more likely than a faulty MCB.
3. Match the right instrument to each job, and state one limitation of each.
Read the worked answer
| Job | Instrument | Limitation to remember |
|---|---|---|
| Prove a circuit dead | Two-pole voltage indicator, proved on a known live source before and after | Non-contact voltage pens can indicate on a neighbouring conductor and must never be used as the sole proof of isolation |
| Insulation resistance | Insulation tester at 500 V DC (1000 V for some equipment) | Must have all loads, electronics and surge devices disconnected, or the reading is meaningless or damaging |
| Earth electrode resistance | Fall-of-potential earth tester, or a clamp earth tester on a suitable electrode | A clamp tester needs a parallel return path; it cannot measure a single isolated rod |
| Loop impedance / fault current | Loop impedance tester at the point of use | Readings drift with supply voltage; note the value at the time of test |
| Load current | True-RMS clamp meter, one conductor only | Mean-responding meters under-read inverter and electronic loads |
| Low-resistance joints and continuity | Low-resistance ohmmeter (four-wire where available) | A multimeter's buzzer cannot resolve fractions of an ohm, which is exactly where joint faults live |
| Socket wiring | Socket tester | Detects polarity and earth faults only; it does not measure earth loop impedance or insulation |
Rule of thumb for the candidate: choose the instrument by the quantity being measured, not by what is already in the bag, and record the instrument's serial number and calibration date on every test sheet.