# Inductor saturation vs RMS current

Looking at the specs for power inductors, there are two major ratings for consideration, saturation current and RMS current.

From my understanding, saturation current is when the magnetic flux of the core cannot hold any more energy(hence saturation). If saturation current is the point where the inductor cannot hold more energy, then why is the RMS current higher than I(sat)? Take the RFS1317-473KL for example, which has a RMS current of 4.5A but an I(sat) of 3.9A.

"Saturation current" is defined as the DC current that causes the inductance to drop by some percentage X%. So for every value of X% there is a "saturation current" specification. It would make no sense to specify one without the other. For the inductor in your question, they give three values.

These values correspond to points in the L(I) curve:

One of the important uses of saturation current is if you use it with DC-DC chip having cycle by cycle current limiting.

The worst case is maximum input voltage and minimum output voltage, ie output unexpectedly shorted while it is running. When this occurs there is a race between the chip's current limiting trying to turn off the MOSFET, and current increasing in the inductor according to:

$$\ \frac{di}{dt} = \frac{V}{L} \$$

If the inductor saturates deep enough before the MOSFET turns off, then L decreases, and the rate of increase of current shoots up. When that happens, if you probe current in your MOSFET, you'll see something like this:

So it's important to check at which maximum current the current limiting will trip and how long it will take to turn off the MOSFET, take into consideration that current keeps rising until the MOSFET turns off, and end up with an estimation of the final current value when it does turn off, in the worst case. Then ensure the inductor is not saturated at this point. A bit of inductance loss is OK, but if it has lost most of its inductance and become a wire, the MOSFET will smoke.

The other spec, RMS current, is about self heating due to coil winding resistance.

Suppose you make a LC filter. Under normal conditions it will pass 1A, but if there's a short on the output, the switching power supply may limit current to 2A.

In this case, for the filter to be effective, the inductor must not be saturated at the normal current. This doesn't need to apply at the fault current, because the filter becomes irrelevant if the output is shorted.

In case of a fault, inductor RMS current rating determines if the inductor overheats and burns or desolders itself from the board, or not.

Because copper costs money, it is possible to find inductors with lower RMS current rating than saturation current. You'd use these when trying to penny-pinch a DC-DC with hiccup protection where the inductor needs to not saturate for the current limiting to work, but this high value of current will only occur on peaks and will be followed by a turn-off period, so even if the RMS rating is exceeded during a few µs, it is not exceeded continuously.

In the provided reference's chart of Typical L vs Current refer to 100uH and its data. The drop in inductance for the provided DC Isat data can be easily seen on the chart. While the AC RMS current is for temperature rise from losses therefore the RMS current is unrelated to the saturation current data.