Aircon clamp meter: what an inverter does to the reading

A clamp meter reads the magnetic field around a conductor. The jaw closes on the outside of a cable, and nothing in the circuit is opened to let the tool in. That single property is why the instrument exists at all.

The standard governing this class of tool says so in its own scope. IEC 61010-2-032 sets safety requirements for hand-held current sensors used on circuits without physically opening the current path being measured. Current clamps, clamp multimeters and current probes all sit under it.

One conductor at a time is the rule the physics imposes. Hioki puts it flatly in the instruction manual issued with its CM4371-50. Clamping around two or more conductors in a bundle stops the instrument measuring any current at all, on single-phase and three-phase circuits alike. The two fields oppose each other and the jaw sees close to nothing.

The opposite failure is quieter and harder to catch. That same manual lists a case where a value appears on screen with no input present. A transformer or high-current circuit nearby throws a strong magnetic field, and the jaw answers to that instead. A wireless device with a strong electric field does it too.

So a number on the display is a statement about the field at the jaw. Turning it into a statement about the machine is the part that happens afterwards, in what the report says next.

A variable-speed compressor has no single running current, because it has no single running speed. Daikin's service manual covering the FTXS-L and FDXS-L series sets out how that speed gets chosen. Frequency is determined by the difference between the room thermistor temperature and the target temperature.

That one sentence governs the whole reading. A room sitting far from its set point asks for a high frequency. A room close to it asks for a low one. Both are the machine behaving correctly, and they produce different currents on the same afternoon.

The control then trims that frequency further. Daikin sets the upper limit as the lowest of several separate limits at once: compressor protection, input current, discharge pipe temperature, and freeze-up protection. Heat pump models add heating peak-cut and defrost to the same list.

Current is not merely an output of that loop. It feeds back in. Under the heading of input current control, the manual describes the microcomputer calculating input current while the compressor runs, and setting the frequency upper limit from what it finds. The zones are named stop, dropping, keep and reset.

Where that limit bites moves as well. The manual states the current at which the control acts drops once outdoor temperature rises past a level that depends on the model. It prints separate thresholds for each capacity class, and separate ones again for cooling and for heating.

A further layer sits under the owner's thumb. Daikin's ECONO setting limits maximum power draw, and the manual notes the maximum running current in that mode varies with the outdoor unit attached. A button on the remote changes what a correct reading looks like.

Specification sheets do print a running current, and the figure is genuine. Daikin gives Running Current (Rated) for every model in the series. The value sits inside a table whose conditions are printed directly underneath it.

Those conditions are narrow. Daikin fixes four air readings behind the cooling figure: 26.7 °C dry bulb against 19.4 °C wet indoors, then 35 °C dry bulb against 24 °C wet outdoors. Heating carries its own set again, and a pipe run of 7.5 m belongs to the same note.

Four air measurements and a pipe length stand behind that one current figure. A Singapore flat meets all five at once by accident, if it meets them at all. The ledge runs its own temperature, the room is wherever the occupants keep it, and the pipe run is whatever the installer needed on the day.

Those tables carry a second admission worth reading. Beside each rated figure for capacity and power consumption sits a stated low and a stated high. The machine is documented as working across a span, and the rated figure marks one point inside it.

That span is where the practical trap sits. A figure taken near the bottom of it and a figure taken near the top can both belong to a healthy machine. Setting one against the other, or against the single rated number, opens a gap that looks like evidence. The comparison only carries weight once both readings arrive with the state they were taken in.

This is the argument aircon pressure readings already makes about pressure, and it does not need making twice. A number without the conditions it was taken under cannot be set against anything. What follows below is the part that belongs to current alone.

Two clamps on one cable can disagree, and the split comes from how each arrives at its number. Fluke separates the two families in its application note written for HVAC technicians. Average responding tools capture the rectified average of an alternating current and multiply it by 1.1.

That multiplier carries an assumption inside it. Fluke calls the displayed figure a calculated value based on an assumption about the shape of the current, and notes the method works on a pure sine. Where the current is not a sine, the assumption fails without announcing itself.

Aircon equipment is where it fails hardest. The same note names solid state adjustable speed motor drives among the loads causing the problem, and describes non-linear loads drawing current in short pulses instead of the smooth sine an ordinary induction motor draws. An inverter outdoor unit is that load exactly.

Fluke publishes the size of the error. Average responding meters may read anywhere from 5 % to as much as 40 % low on line side currents where loads are non-linear. Its own illustration puts a caption on a distorted branch circuit, where the average responding clamp reads low by 32 percent.

Instrument makers say the same thing from the other side. Hioki states that its true RMS tool reads distorted current well. What it shows will differ from a clamp that uses the averaging method. One input condition sits above the accuracy table, and it reads sine wave input.

Crest factor is where a true RMS tool meets its own edge. Fluke's calibration manual for the 374, 375 and 376 prints crest factor figures tied to stated currents, then instructs adding 2 % to the accuracy above a crest factor of 2. Correct has a boundary too, and the maker prints where it sits.

A current figure answers a small set of questions well. Whether the motor is drawing at all separates a dead circuit from a live one. Whether the figure sits far above the maker's full load rating separates a machine in trouble from one merely working hard on a hot day.

Circuit loading is the third of them. Fluke notes that fuses, bus bars, conductors and the thermal parts of breakers are all rated in rms current, because what limits them is heat. Checking a circuit for overload means reading rms current and setting it against the rated value for the component in question.

Naming a cause is where the figure stops. Fluke's own list behind excessive compressor current runs to shorted or grounded windings, a bad capacitor, a faulty relay, and bearing fatigue. Three of those are electrical and one is mechanical, and the clamp reports the same high number for all four.

Mechanical trouble reaches the meter by a long road. That note says compressor electrical problems are often caused by mechanical failure or by service errors, and that worn bearings drive current higher than normal. Low supply voltage does the same thing, by making the motor pull more than it should.

A reading inside range clears nothing either. Fluke states plainly that low amps are normal during low load conditions. On a machine choosing its own load minute by minute, a comfortable number can mean a healthy compressor, or a drive quietly holding one back.