Question on Inductor choice on a Switched Converter

The design guidelines are trying to keep the converter working in continuous mode. That is there should always be some current flowing in the inductor. the lower the current the larger the inductor to achieve this.

Of course the ICs will also work in discontinuous mode, that is inductor current going to zero but that can affect stability.

So if i desgne a converter for max 300mA and the circuit sinks 50mA it could be unstable?

It is explained on page 18 of the LM2575 data sheet under the section inductor selection.

It could be *discontinuous*.

Usually that's not unstable, at least on a buck converter.

AFAIK many data sheets suggest a continuous (only) design just because it is a bit simpler to analyze and you don't have to characterize the behavior in two different modes.

I had this same question earlier. The easiest way to "understand it" is to scope Pin-4 (I think) while it is running under a constant load, and then swap out the coil values. You will see why it is the way it is.

Incidentally, if overall inductor size is the concern, a higher- frequency switcher might be able to fix that. But IIRC, the LM2575 is a fixed frequency switcher.

The inductance will need to be greater (more expensive, all else equal), but the energy stored at the same switching frequency will be less (less expensive).

So the inductor for the 50mA case won't necessarily be more expensive, and may well be both smaller and less expensive. It'd take some catalog crawling to verify this.

As pointed out, you are being encouraged to choose the inductor based on keeping the current continuous at the supply's minimum design current. This means choosing a peak-peak inductor current equal to twice that minimum current.

=A0

Thank you for all the replies. After taking a look at the datasheet, some other internet references and your answers i got the idea how it does works. I just have one final question. Is it possible to determine previously if a designe is going to be unstable at certain conditions? If yes, how?

Thank you!

h a

ng

no?

ger

log

=A0

Just another question. When chosing the Inducton Max Current, should i consider the max current or just the average current?

Thank you!

Maximum current including ripple. Otherwise you risk the inductor saturating.

BOth a

loing

se a

rk no?

bigger

al),

less

e,

log

d on

t.

Hmm thank you!

And finally, should i get worried about a 5A peak current disturb any digital part of my circuit with EMI? Thank you!