In bulk convertor, should I use the maximum Iout to calculate inductor?

[ridgemao1983]

When calculating the value of inductor in a bulk convertor, you need to use the Iout - regulator's output current. Should I use the predicted maximum Iout or the actual Iout to calculate the inductor?

The actual Iout is obtained by calculating ALL the loads on the hardware board.
The predicted maximum Iout is the actual Iout plus some margins.
I think I should use the actual value of Iout, am I right?

And I think the actual Iout can vary during board's operation, and the actual Iout reaches its maximum value when you turned on all the functions on the board. If I use the actual Iout to calculate the inductor's value, should I use its maximum value or average value?

 

[FvM]

If I use the actual Iout to calculate the inductor's value, should I use its maximum value or average value?

Do you refer to the difference involved with inductor current ripple? You usually have two specifications for rated inductor current, Irms (thermal) and Imax (core saturation defined). Both requirements should be considered.

 

[ridgemao1983]

Sorry, I didn't mean the inductor's Iout. When you design a bulk regulator, you need to consider the required maximum load, the maximum load current, It is the Iout I refered here, not the inductor's characteristic.

You need to use this "Iout" to calculate the inductor's value, I think I should set the Iout much closer to the actual load current.
But when you choose other components(eg. capacitor) of the bulk regulator, you need to add some margin to the maximum load current.

 

[FvM]

I this case I don't exactly understand the question. Assumed, the "actual" Iout is e.g. 10% of maximum Iout, do you want to use the tenfold inductor value? That sounds absurd. If a converter is operated at a fraction of rated current for most of the time, you'll possibly want to increase the inductance to reduce the ripple, but not beyond reasonable limits.