Aircon DC reactor: the heavy coil fitted for the mains

The part is a length of wire wound onto an iron core, with no electronics inside it. Two leads go in and out. Daikin tells the technician, in the RXS09 and RXS12 removal steps, that the two reactor harnesses can go back either way, colour regardless. Very little else in that box tolerates being reversed.

Position is the next useful fact, because the board does not carry it. Daikin gives the coil a removal heading of its own, Removal of Reactor / Partition Plate, and one screw holds it. Mitsubishi Electric writes the same arrangement into the PUZ-HA service steps. There the reactor and the capacitor are attached to the rear of the electrical parts box. On Toshiba's larger multi-type outdoor units the coil gets a box to itself, sitting behind the inverter assembly with its own anchor plate.

Mass is why the mounting looks like that. A Toshiba outdoor inverter parts table lists a reactor for the compressor rated 5.8 mH at 16 A, and a second one for the fan at 5.8 mH and 14 A. Those figures belong to the MMY-MAP chassis in that document and to no other machine. Hitachi Industrial Equipment Systems sells DC reactors as an inverter option and prints a weight against each. The smallest of them, matched to a 0.2 kW drive, is 0.8 kg. That is different equipment, quoted to show the class of object and not the aircon part.

Some units carry two of them. Toshiba's wiring legend separates the reactor for the compressor from the reactor for the fan and lists both. Carrier labels two reactors beside the compressor on its 18K size. So on those machines, the phrase the reactor has already skipped a question.

Each maker labels it differently on the diagram, which is how one object reaches a homeowner under several names. Mitsubishi Electric marks it DCL, alongside a diode bridge, a main smoothing capacitor and a power module on the same legend. Daikin marks it L1R and expands that symbol as REACTOR. Toshiba writes reactor for compressor. Hitachi's industrial catalogue, selling the same idea to a different trade, calls it a DC reactor. Of everything on those legends, the reactor is the one item bolted outside the electrical box.

The purpose is written down, and it points away from the machine. An inverter makes direct current out of the mains before it drives anything. That conversion pushes current back into the supply at frequencies the supply never sent. Mitsubishi Heavy Industries names the consequence in a technical review of its own room aircon inverters. The converter causes the outflow of harmonic current to the power supply system, it writes, and this harmonic current can interfere with other equipment.

Reactors are the named remedy. Mitsubishi Electric publishes a harmonic suppression guideline for inverters, and the instruction there runs to one sentence. Connect an AC reactor on the power supply side of the inverter or a DC reactor on its DC side or both, it says, to increase the circuit impedance, suppressing harmonic currents. The same table states the effect. Harmonic currents are suppressed to about one half.

A drive maker selling the component on its own says it in a line. Hitachi Industrial Equipment Systems describes its DC reactor option this way: it suppresses harmonics generated on the inverter input side, and it is useful for power factor improvement. Two jobs, and neither of them is cooling a room.

Power factor is the second job and the easier one to picture. Current drawn in step with the voltage is current the supply delivers without carrying extra. Mitsubishi Heavy Industries describes its converter working the reactors so that a sinusoidal current with a power factor of approximately 1 flows. A service manual built by Gree and sold under the Comfortstar name reaches the same idea from the repair end. The PFC board is a component of the inverter air conditioner for power factor correction and voltage boosting, it says, and the reactor is one of core parts of PFC.

None of that makes the coil optional. In the designs these manuals describe it sits in the running power path, so a machine that loses one does not simply lose its manners. Mitsubishi Electric lists a disconnected reactor as a reason an outdoor unit never powers up at all. Why a maker fits the part and what happens when it goes are two separate subjects, and only the first is about the supply.

A harmonic limit applies to room aircon, and makers say so in writing. Mitsubishi Heavy Industries puts it flatly. Every room air conditioner is required to meet harmonic regulation values such as IEC, the paper states. Its own export models carry two chopper circuits in the converter to satisfy those limits.

Those limits live in IEC 61000-3-2. The fifth edition, published in January 2018, is titled Limits for harmonic current emissions. The bracket after that title carries the boundary: equipment input current at or below 16 A per phase. It applies to equipment intended to be connected to public low-voltage distribution systems, and it sets limits for the harmonic components of the input current.

That boundary sits inside the domestic range. The three export models tabulated in the Mitsubishi Heavy Industries paper run on single-phase 220 to 240 volts. Their maximum running currents are listed as 14.5 A, 17 A and 17 A. An ordinary household split sits either side of the figure in that standard's title.

One link in the chain stays general on purpose. No service manual read for this page cites a standard number against the reactor. Makers state a harmonic obligation, published guidelines name reactors as a way of meeting it, and the coil is in the unit. That is a reading of three documents, offered as a reading and not as a quoted instruction.

Supply quality changes the job the coil has to do. The same technical review notes that a distorted supply voltage distorts the current through the reactor with it, and the paper is direct about the cost. The distorted input current causes extra harmonic current flow, which increases the loss and deteriorates the efficiency. The machine still cools. It cools less efficiently, for a reason with nothing to do with the refrigerant side.

Several do, and that separates this coil from most of the power stage around it. In the service diagnosis covering Daikin's FTXS-L and RXS range, disconnection of the reactor sits among the supposed causes of an H0. That code is raised when supply and DC checks taken before the compressor starts come back wrong. The branch that follows ends at two instructions: connect the reactor properly, or replace the reactor. On the same range, H8 lists a defective reactor among its supposed causes.

Other makers arrive from other complaints. In the PUZ-HA self-diagnosis table, Mitsubishi Electric lists disconnection of the reactor. No electric power reaches the outdoor controller circuit board in that case, which presents to the owner as an outdoor unit that never starts. Carrier's ductless service manual gets to the coil three separate times. The routes there are a communication error, an overload current protection, and a voltage protection.

One manual carries a warning worth taking into any conversation about this part. The Gree-built document, having named the reactor as one of core parts of PFC, adds a flat prohibition. Do not remove the reactor and replace with short circuit by no means, it says. Bridging a coil out is the kind of shortcut that appears to work. The unit runs again, the fault clears, and the function the part was fitted for is gone with no sign of it on the outside.

No owner ever sees this coil, and no display names it. What appears is a code raised by whatever noticed the trouble, and across these documents that is a current reading outside its band, a voltage reading outside its band, a link that stopped answering, or no power at all. A code is evidence of detection. Evidence of diagnosis is a separate thing, and only the code reaches the room.

Heat is documented too, which is why the manuals slow the technician down. Toshiba instructs that the breaker be set off and at least ten minutes allowed before the service panel is opened. The reason given is heat: the fan motor, reactor and inverter will be very hot to the touch. Daikin sets its own wait before disassembly begins. This stage carries mains potential, and there is stored charge in the smoothing capacitor beyond it. All of that is written for someone in gloves with the supply isolated. No part of it belongs to an owner.

This is one of the few items in an outdoor power stage that makers say to replace on its own. Daikin's branch ends at replace the reactor. Toshiba gives it a part number and a procedure. Carrier's flow finishes the same way, and the Gree-built manual lists it as a stocked item. So a quote naming this coil is not automatically a board price, and it should not read like one.

The same documents also hold endings where nothing is replaced. Connect the reactor properly is one. Check the connection of the reactor is another. Those cost different money, and a quote charging a part price for a joint has billed the wrong outcome.

Where a unit carries two, one question comes before the others. Which coil, and what the machine reported before it stopped, is most of the diagnosis on that chassis.

Matching the label on the paper to the label on the diagram is worth the minute. DCL, L1R and reactor for compressor are three makers describing one object, and a technician working from the wrong manual writes down the wrong one. The model on the outdoor unit plate settles which document applies.

One thing this coil never does is explain itself to the room. It has no light, no sound and no smell of its own, so a page like this exists to make a quote readable, not to help anyone spot a symptom. Where the reactor is genuinely the finding, the machine reported something else first, and that something else is where the conversation starts.