SMPS O/P filter inductor selection

[JEWilson]

I require to replace some chokes in PSUes

What is the best selection method?

I've seen data that Isat should be greater than Ipeak

Should I take it as read, Ipk is as shown on the PSU rail spec.?

Also, should I select the highest inductance as I can afford
given requirements and form factor?

For example,
The first inductors is 8 mm dia and 10mm high
This is the -5.0V rail on a Enermax PSU
The rail O/P current states 1.80 A max.

The second inductor is smaller at 6 mm dia and 8mm high
This is the +5.0VSB rail on a Chieftec PSU
The rail O/P current states 2.00 A max.

So, would it be the case, I select the form factor and
ensure the saturation current exceed the stated maximum
and try to get the highest inductance value as is possible
at a price TBA?

Any advice appreciated - thanks

[gonzo0815]

Re: SMPS O/P filter inductor selection

why do you want to replace them? are they burnt? haven't seen much burnt inductors in ATX PSu`s, where i would have bothered to fix it.

[JEWilson]

Re: SMPS O/P filter inductor selection

They are broken as a result of my poor handling!

So... need replaced!

Of course, as they have no markings on them and I do
not have a inductance meter, I have been looking at Pi Filter calculator applets

I have the O/P load resistor value for the Enermax (15R) -5.0V rail.
Caps C1 and C2 respectively, are 470uF 16V and 100uF 35V

I am a bit hazy on this and have not used filter calcs for many
years, so some advice would help here

Thx

[gonzo0815]

Re: SMPS O/P filter inductor selection

I think your approach seems not that bad, i don´t think that the exact value is critical.
Only the mag amp inductor (well usually the biggest inductor with more then one winding) is probably a little more important.
But there i haven`t much clue what it should be.

[JEWilson]

Re: SMPS O/P filter inductor selection

Having done a little research and having used pi filter design calc's

A synthesis of a pi LC low pass filter can be realised.
Butterworth, Chebychev and Bessel alignments amongst
other can be had but these relate to, for example, to RF and
audio filtration and control freq. response.

Using inductors in the O/P pi filter for SMPS I believe, is to
reduce ripple. To this end, it has been suggested, in the
event an inductor fails, replace it with a wire link but at
the expense of an increase in the ripple on that rail.

Also, it has been suggested some flexibilty may be available
in respect of the L1 and C1, C2 paramters such that, if
you increase rail capacitance, a smaller form factor and
in turn, rated inductance can be selected.

The rule of thumb as suggested, is to ensure the saturation
current of the inductor exceeds the peak current O/P of
the rail under investigation and that you should select the
highest inductance you can afford and of course, phy. form
factor comes into this.

Is this an accurate maxim for SMPS O/P rails?

For the Enermax, I replaced ; the O/P caps with Pana FC some
time ago. However, I quickly reaslised, post research,
the fact, the +5.0VSB as well as the -5.0V rail did not have the same
cap values in positions C1 and C2.
Probably, due to production issues as opposed to design I surmise.

For instance, the +5.0VSB rail with a advertised peak O/P current of 1.8 A has;
1. 1 of 470uF 16V and
2. 1 of 1000uF 10V

The -5.0V rail with an advertised peak O/P current of 1.0 A has;
1. 1 of 470uF 16V and
2. 1 of 100uF 35V

So, first thing I did was ensure;
+5.0VSB has 2 of 1000uF 10V
-5.0V has 2 of 470uF 16V

As I do not know the inductance in both these rails, I have looked
at phy. form factor as a guide.

If I attempted to design a Butterworth low-pass pi filter with C1 and C2
at 1000uF at the load at 10 ohms, dependant on freq response,
you would require an inductor of up to 200mH - this inductor would be huge!
So working at DC with an AC ripple component with OP rail filter considers
alternative characteristics such as ripple reduction and cost constraints.

Any thoughts and suggestions advanced, greatly appreciated