Aircon Soft Starter: An Inverter Unit Already Has One

A fixed-speed compressor makes its largest demand on the supply in the first moment it runs. Everything upstream carries that moment together: the cable, the switch guarding it, and the supply feeding the flat. All three were sized with this instant in mind, and none notices the machine again until the next one.

Heavy current for a brief moment pulls voltage down along the whole run. Manufacturer literature treats that sag as a design concern, not a curiosity. A Mitsubishi Electric split-unit service manual tells installers to add counter-measures where several units share one supply, naming what they prevent: the main voltage drop, and the rush of starting current.

Owners meet this as something visible. Lights dim for an instant as the compressor picks up, a fan hesitates, or a device sharing the circuit blinks. That flicker is the surge arriving; on a sound installation it passes immediately and leaves nothing behind.

Simultaneous restarts are the sharp version. The same manual notes that when several air conditioners sit on one supply and power returns after an outage, the starting current of all the compressors may flow at once. Its stated remedy is a system that lets the units start one by one.

Look at what that remedy leaves alone. Nothing about any individual compressor changes. The starts are spread out so the total never lands in one instant, and the manual separately advises keeping other appliances from switching on then. The strain being managed belongs to the installation, and timing manages it.

This is the problem a soft starter is sold against. Whose problem it is decides almost everything about whether the device is worth having. Answer it wrongly and the part lands on a machine with nothing to gain.

A soft starter holds current down while the motor comes up to speed. Danfoss describes the traditional method in its soft starter operating guide: constant current starting raises the current from zero to a specified level, then keeps it stable until the motor has accelerated. The natural peak never arrives, because the device does not permit it.

The intent runs opposite to the other parts in the start path. A start capacitor and a start relay add turning force so a stationary shaft breaks away. A soft starter withholds supply so the wiring sees less. One group argues for the motor, the other for the circuit, and knowing which side a proposed part sits on tells you what it was bought to solve.

Because the two pull against each other, softening has a floor beneath it. Danfoss states the consequence: if start current is too low, the motor does not produce enough torque to reach full speed. Its starters carry a protection named excess start time, which trips the unit when the motor never reaches speed within the window.

On an inverter unit the function is already present, supplied by the drive that runs the compressor. Copeland's application guidelines for its variable-speed scrolls put it in one line: with the drive, a soft start can be implemented by ramping up the compressor at a defined speed. The capability is a setting inside equipment the unit already contains.

A variable-speed compressor is never thrown straight onto the mains. Its speed is commanded by the drive at every moment, so the start is a controlled ramp by construction. The same guidelines list a soft start-and-stopping routine among the functions any third-party control system must provide, placing the ramp inside the drive's job.

An external device would duplicate something: it would sit ahead of the drive and limit what reaches it, while the event it was bought to soften happens past that point, inside equipment already commanding the start.

Telling which machine you own is usually straightforward. The outdoor unit's nameplate and the model documentation say so, and variable-speed models are marketed on the fact. Where a system came with the flat and no paperwork survives, the model number stamped on the casing is enough to look it up.

Copeland names the case, and it is a supply case throughout. Its guidelines say that for extreme electrical conditions such as weak power supplies, single-phase soft starters are available on request, while three-phase versions come from the wider market. The trigger is the state of the electricity reaching the machine.

The default runs the other way, which a quote rarely mentions. Copeland writes that the compression components inside its scroll compressors start unloaded, even when system pressures are unbalanced, so low-voltage starting characteristics are excellent. Starting components, it adds, are normally not required. On a scroll, needing no help at the start is the expected condition.

That turns the question around, and it should be. Anyone proposing one should be able to say what about this installation makes it the exception. A long or weak supply run, a circuit already close to its limit, or a small generating source behind the meter are answers that survive a follow-up question.

Selection and fitting belong with the people who size the circuit. The device is matched against the compressor's current figures and the starter maker's guidance, and installing it is licensed electrical work in Singapore. Treated as a design decision taken while the circuit is planned, it lands correctly. Treated as an accessory bought after something went wrong, it usually lands on the wrong machine.

A soft starter lowers what a healthy start demands and does nothing about a fault creating the demand. Danfoss draws the boundary in its manual: the soft starter is not a safety device and provides no electrical isolation or disconnection from the supply. It manages a load; it does not watch for anything going wrong.

This is where caution is owed. If a breaker keeps tripping because a start path has failed or a machine is labouring to turn, capping the current can move the trip point far enough away that the tripping stops. The symptom leaves; the condition stays, sealed inside a part nobody can open, and the owner now has good reason to believe the matter was settled.

There is a second-order version of the trap. A machine already short of turning force is the last one that should be given less of it, since capping current caps torque by the same manufacturer's account. Softening the start of a straining compressor lengthens the stretch it spends below speed, which is the stretch doing damage.

None of that argues the device is dishonest. It argues that a symptom disappearing is a poor test of whether anything got fixed. A part sold on the promise of making a trip go away deserves a harder look than one bought to make a machine work again.