French
Choosing between a thermal-magnetic and an electronic trip unit is a selection question — one we cover in our MCCB selection guide. This article is about the question that comes after: the breaker is on the panel, the dials are staring at you, and you have to decide what to set Ir, Isd, Ii and Ig to.
Those four settings decide two different things. Ir protects the cable from overloading. Isd, Ii and the delays decide whether a fault on one feeder blacks out one machine or the whole board. Get the first wrong and you cook a cable. Get the second wrong and a fault in a packaging line takes down the plant.
Electronic trip units use the same naming convention across most manufacturers, which is helpful once you know it:
| Symbole | Fonction | What it protects against | Set relative to |
|---|---|---|---|
| Dans | Sensor or rating plug — the base value every other setting refers to | — | Fixed by the trip unit ordered |
| Ir (L) | Long-time pickup | Sustained overload | Multiple of In, typically 0.4–1.0 |
| tr | Long-time delay | How long an overload is tolerated | Seconds, usually quoted at 6 × Ir |
| Isd (S) | Short-time pickup | Moderate fault current, with a deliberate delay | Multiple of Ir |
| tsd | Short-time delay | The delay that lets a downstream device clear first | Typically 0.05–0.5 s |
| Ii (I) | Instantaneous pickup | Severe bolted faults, no intentional delay | Multiple of In |
| Ig (G) | Ground-fault pickup | Earth leakage / insulation failure | Multiple of In, where fitted |
There is one subtlety worth internalising: Ir and Ii are usually expressed as multiples of In, while Isd is usually expressed as a multiple of Ir. Confusing the two reference bases is the single most common arithmetic error in trip unit setting, and it produces settings that are wrong by a factor of two or more.
Ir is the long-time pickup — the current above which the breaker will eventually trip on overload. The instinct is to leave it at 1.0 × In, because that is where it ships. That is usually wrong.
Ir should be set to the design current of the circuit the breaker protects, not to the maximum the frame can carry. A 250 A breaker feeding a 160 A load, left at Ir = 1.0, gives that load no meaningful overload protection at all: the cable would have to carry more than 250 A before the breaker even begins to think about it.
Work the other way round. Establish the design current Ib, confirm the cable can carry it, then select In so that Ib sits comfortably inside the Ir adjustment range, and dial Ir down to Ib. If the load is 320 A, a 400 A sensor with Ir set to 0.8 gives you 320 A — a real overload threshold.
The long-time delay tr is usually an inverse-time characteristic: the higher the overcurrent, the faster the trip. Set it long enough to ride through legitimate inrush — motor starting, transformer energisation — but short enough to stay inside the thermal withstand of the cable.
The short-time zone is what separates a breaker that simply protects from a breaker that coordinates. Isd sets the current above which the short-time function arms; tsd sets how long it waits before acting.
That delay is the whole point. If a fault appears on a downstream feeder, the downstream breaker should clear it. The upstream breaker sees the same fault current, but because its tsd delay is longer than the downstream device’s total clearing time, it waits — and stays closed. Only if the downstream device fails does the upstream breaker trip as backup.
Typical short-time delays run from about 0.05 s to 0.5 s. A commonly used starting point for an upstream MCCB coordinating with downstream MCCBs is 0.2 s, but the correct value comes from the time-current curves, not from a rule of thumb.
Many trip units offer an I²t ON / OFF choice on the short-time delay. With I²t ON the delay shortens as current rises, which limits let-through energy. With I²t OFF the delay is fixed regardless of magnitude, which is easier to coordinate against. Choose deliberately; do not leave it at whatever the factory set.
The instantaneous function trips with no intentional delay. It exists for severe faults where speed matters more than discrimination.
Set it high enough that normal inrush does not nuisance-trip the breaker, and low enough that it still clears a solid fault on the circuit it protects. On some units the instantaneous function can be switched off entirely — which is how a Category B device achieves a pure time-delayed response. Do not disable Ii unless the coordination study genuinely requires it and the equipment has a declared short-time withstand to carry it.
Ground-fault protection detects residual current returning through earth. On a four-pole MCCB with a neutral CT, the trip unit summates the phase and neutral currents; any imbalance is leakage.
Ig must be set above the normal leakage and capacitive charging current of the installation, or it will trip on healthy circuits. Large cable runs and variable-frequency drives both generate standing leakage, which is why Ig is normally set with a deliberate delay as well — long enough to let a downstream RCD clear first. For the detection principle itself, see our article on residual current protection.
Say a sub-distribution board is fed from a 630 A incomer, and one of its outgoing feeders supplies a 320 A load over a cable rated 400 A. The feeder is a 400 A frame with an electronic trip unit.
Now the incomer. For the two to discriminate, the incomer’s short-time delay must exceed the feeder’s total clearing time, and its instantaneous pickup must sit above the prospective fault current at the feeder — or be set high enough that current selectivity holds. In practice: incomer Isd at a comparable multiple with tsd = 0.3 s gives the feeder 0.2 s of margin to clear first.
| Method | How it works | When it applies | Limit |
|---|---|---|---|
| Current selectivity | Upstream instantaneous pickup sits above the maximum fault current at the downstream device | Where there is meaningful cable impedance between the two levels | Fails at high fault levels or with short cable runs |
| Time selectivity | Upstream short-time delay exceeds downstream total clearing time | Main method on electronic trip units with an S function | Requires the upstream device to have a declared Icw to ride through the delay |
| Energy selectivity | Comparing let-through I²t of the two devices | Used by some manufacturers for current-limiting breakers | Needs manufacturer-published let-through data |
| Zone selective interlocking | Downstream device sends a restraint signal to hold the upstream delay | Larger systems where delays would otherwise push arc-flash energy up | Requires a communication link between devices |
The critical constraint is the one people discover late: a time delay is only safe if the upstream device can physically survive the fault current for that long. That capability is the rated short-time withstand current, Icw. Most MCCBs have modest Icw values, which is exactly why full time-graded selectivity often pushes a design towards an air circuit breaker at the incomer — a point we work through in MCCB vs ACB.
A thermal-magnetic unit typically gives you one adjustment — the thermal setting, often 0.7 to 1.0 times the nominal rating — with the magnetic element fixed by the factory or selectable in coarse steps. There is no Isd, no tsd, and no independent Ii.
The practical consequence is that thermal-magnetic breakers achieve selectivity only through current discrimination, and only where cable impedance is generous enough to make fault levels at the two devices differ substantially. Where you need tight, repeatable coordination across several levels, specify an electronic trip unit. Our SNSX 400–630 A et SNS 800–1600 A ranges cover the frames where electronic trip units earn their cost.
In is the rating of the trip unit’s sensor or rating plug — the fixed base value. Ir is the long-time pickup, set as a multiple of In, and it defines the actual overload threshold. In is chosen when the breaker is ordered; Ir is adjusted on site.
On most electronic trip units Isd is a multiple of Ir, while Ii is a multiple of In. This varies by manufacturer, so always check the trip unit’s own dial markings or manual before calculating.
Some electronic trip units allow Ii to be disabled to achieve a pure time-delayed response. Only do this where the coordination study requires it and the switchgear has a declared Icw high enough to withstand the fault for the full delay.
Because the two are not coordinated at that fault level. Either the upstream instantaneous pickup is below the prospective fault current at the downstream device, or the upstream short-time delay is shorter than the downstream device’s total clearing time. Compare the published time-current curves at the actual fault level.
Partially. They rely on current selectivity, which works where cable impedance makes the fault level at the two devices clearly different. They cannot do time-graded coordination because they have no adjustable short-time delay.
A breaker with correctly set Ir, Isd, Ii and Ig is a protection device. The same breaker left at factory settings is a switch with a number on it. The difference is a settings schedule produced during design, checked against the time-current curves, and verified at commissioning.
If you are specifying MCCBs for a panel and want the trip unit ranges confirmed against your coordination requirements, send us the single-line diagram and the fault levels. Contact our engineering team and we will confirm which trip unit and frame fits.
