Every method that reduces starting current does it by reducing the voltage the motor sees. Current falls in proportion to voltage; torque falls with the square of it. That single asymmetry is why soft starting is not free, and why the answer to “the breaker keeps tripping” is sometimes “then it will not start your load either”.
The breaker is only one constraint. Drawing several times full-load current for a few seconds pulls the supply voltage down for everyone on it — and because torque follows voltage squared, a deep dip can leave the motor unable to finish the very start that caused it.
At switch-on the rotor is stationary, so the machine looks like a transformer with a shorted secondary: low impedance, enormous current, power factor about 0.2. As it accelerates the current collapses.
I = P / (√3·V·PF·η). The plate rating is shaft output, so the motor draws more than that.
At a fraction a of nominal voltage a motor draws a× the current but produces a²× the torque.
Instantaneous magnetic trip at 3–5× rating (type B), 5–10× (C), 10–20× (D).
Locked-rotor current is on the nameplate as a code letter or an Ia/In figure. The spread matters more than the average: on a 20 A type C, a 7.5 kW motor at 6× stays clear of the magnetic trip and the same motor at 8× is inside its window. Design to your motor’s number, not to the rule of thumb.
The one thing to carry away. Every reduced-voltage method buys supply relief with shaft performance at that exchange rate, and physics fixes the rate.
The thermal element protects the windings from sustained overcurrent; the magnetic element clears faults instantly. A starting motor is a legitimate several-second overcurrent that must not look like a fault. Going up a breaker size to stop nuisance tripping raises the fault threshold and leaves the motor under-protected: change the type, or fit a motor circuit breaker, with separate overload protection sized to the running current.
The dip a start causes reduces the voltage that start has to work with, and torque follows its square: a 10% dip costs 19% of the torque. On a small transformer or a long rural feed this is the binding constraint, why the same motor is unremarkable in a factory and unusable on a farm.
A stationary rotor absorbs several times rated power with no cooling from the fan. One long start is survivable; three failed attempts in a minute is how windings burn. Permitted starts per hour is a real number on the datasheet, usually much smaller than people assume. Inrush lasts as long as it takes the load to reach speed: under a second unloaded, fifteen or twenty for a large fan with real inertia.
Star-delta is cheap with no electronics, but gives a third of the torque, an open transition that produces a fresh current spike partway through, and no control over the ramp: it survives mostly on existing installations and unloaded starts. A soft starter does the job properly; a VFD costs more again and pays it back in running energy on a pump or fan, and fixes a tripping breaker most thoroughly: constant volts-per-hertz keeps flux at its design value, so the motor makes full torque drawing typically 1.1 to 1.5 times running current, at the cost of the drive’s own input inrush, harmonics and EMC.
Because the two are protecting against different things. The thermal element responds to sustained overcurrent and is happy with a two-second inrush. The magnetic element responds instantly to a multiple of the rating, and a 6–8× inrush lands squarely in the type B window and often in the type C one. The fix is a device whose magnetic trip clears the inrush — a type C or D, or a motor circuit breaker — with separate overload protection sized to the running current, not a bigger breaker.
Rarely, on new work. It is cheap and it has no electronics, but it gives you a third of the torque, an open transition that produces a fresh current spike partway through, and no control over the ramp. A soft starter costs more and does the job properly; a VFD costs more again and pays it back in running energy if the load is a pump or fan. Star-delta survives mostly on existing installations and on unloaded starts — a compressor that unloads for starting, for instance.
As long as it takes the load to reach speed, which is a mechanical question, not an electrical one. An unloaded motor is up in well under a second; a large fan with real inertia can take fifteen or twenty. That duration is what decides whether the motor overheats, because a stationary rotor has no cooling and is absorbing several times its rated power. Repeated failed starts are far more damaging than one long one.
Yes, and more thoroughly than anything else, because it lowers the frequency along with the voltage. Keeping the volts-per-hertz ratio constant keeps the magnetic flux at its design value and the slip small, so the motor produces full torque or better while drawing little above its running current — typically 1.1 to 1.5 times. The catch is that the drive itself has an input inrush, it needs harmonic and EMC consideration, and it is the most expensive answer.
A standard cage induction motor with a locked-rotor current you supply, a stiff supply apart from the transformer impedance you enter, and reduced-voltage methods behaving ideally. Not modelled: the speed–torque curve of the motor or the load, so the page cannot tell you whether a given method will actually accelerate a given machine; rotor heating and permitted starts per hour; cable impedance and its own volt drop; the DC transient in the first cycles, which can briefly double the peak; open-transition spikes in star-delta changeover; drive input inrush and harmonics; or soft-starter ramp shaping.
This page has changed in 182 archived releases, but each of those was a site-wide pass, so none is attributable to this tool on its own and none is listed here. That is not a claim that the tool never changed — a release that reworked many pages at once may well have altered this one too. The changelog has them.