Buck Converter Inductor selection Maximum Switch Current Limit vs Output current

[newbie_hs]

I need to select an inductor for this buck converter. Its output current is 3 A and maximum switch current limit is 6 A. As per my knowledge I need to select an inductor whose current ratings (both DC and saturation) greater than 3 A.

My question is:

When selecting an inductor, do I need to select an inductor whose current rating (DC) greater than 6 A and current - saturation (Isat) greater than 3 A?

 

[FvM]

Inductor saturation current must be greater than actual converter peak current, Iavg + Iripple/2.

--- Updated Apr 8, 2023 ---

No. In the below picture, maximum load current is Iav, inductor peak current is Imax. See https://en.m.wikipedia.org/wiki/Buck_converter

 

[D.A.(Tony)Stewart]

Choosing an inductor L depends on Vin-Vou, dI/dt , load max. When transforming from a higher input voltage, permits a higher output current for constant power, thus with a soft-start, one should design not to exceed output load and have margin to saturation. If no soft start, then your design details are missing.

 

[newbie_hs]

Inductor saturation current must be greater than actual converter peak current, Iavg + Iripple/2.

Thank you.By "actual converter peak current",you mean the maximum load current

 

[mtwieg]

Thank you.By "actual converter peak current",you mean the maximum load current

When you say "output current is 3 A," we're inferring that's the maximum average load current Iavg.

The peak inductor current is higher than Iavg by Iripple/2.

So your inductor needs a rated DC current of at least 3A. If the peak current limit for the controller is 6A, then it's wise to use an inductor with a saturation current of at least 6A. In theory you could maybe get by with less than 6A, but that depends on more specific info.

However, you must also consider core loss. If you're actually switching at 1.25MHz, it's likely that core loss will be the dominant bottleneck, not saturation current.

 

[Easy peasy]

COTS inductors are really bad for losses - if we need 47uH 3A ave current 6Apk

we use 2 L's in //, each 100uH 3A, 6Apk i.e. 100% over-rated

any less will result in cores too hot

 

[D.A.(Tony)Stewart]

Peak/Avg current depends on the Q gain of the series resonance of L, C RdsOn circuit during startup for Buck regulators, so choosing a soft-start design can be critical. Otherwise with low ripple current it does not have to be much larger than the load.

Thus starting with a low duty cycle raises the average RdsOn and reduces the Q of the end result towards 0.7 for critical damping is ideal.