SMPS input filter question.

[[email protected]]

Hi to all.
I am currently designing an input filter for a off-line switcher. The
switcher is a tinyswitch 278
running at about 11W. 220Vac to 13.8Vdc , 800mA.
When designing the filter , do you design for the average input
current , or must you design
for the peak input currents you see when the bulk input cap is "topped
up" every 10mS.
There will be a big difference in component (inductor) size depending
on which way you go.
The average current is only a few hundred mA , where as the peaks are
about 1,5A.
If desighning for the peak currents of 1,5A the filter components are
bigger than the whole
rest of the supply!!!
If this is the correct way to go than that is fine , but I don't want
to design for something that
is not necessary.

Cheers
Robin

 

[Eeyore]

Hi to all.
I am currently designing an input filter for a off-line switcher. The
switcher is a tinyswitch 278
running at about 11W. 220Vac to 13.8Vdc , 800mA.
When designing the filter , do you design for the average input
current , or must you design
for the peak input currents you see when the bulk input cap is "topped
up" every 10mS.
There will be a big difference in component (inductor) size depending
on which way you go.
The average current is only a few hundred mA , where as the peaks are
about 1,5A.
If desighning for the peak currents of 1,5A the filter components are
bigger than the whole
rest of the supply!!!
If this is the correct way to go than that is fine , but I don't want
to design for something that
is not necessary.

You need to design for the RMS current as far as heating effect is concerned but
the magnetics must not saturate at the peak input current.

Graham

 

[Phil Allison]

Hi to all.
I am currently designing an input filter for a off-line switcher. The
switcher is a tinyswitch 278
running at about 11W. 220Vac to 13.8Vdc , 800mA.
When designing the filter , do you design for the average input
current , or must you design
for the peak input currents you see when the bulk input cap is "topped
up" every 10mS.
There will be a big difference in component (inductor) size depending
on which way you go.
The average current is only a few hundred mA , where as the peaks are
about 1,5A.
If desighning for the peak currents of 1,5A the filter components are
bigger than the whole
rest of the supply!!!
If this is the correct way to go than that is fine , but I don't want
to design for something that
is not necessary.

** Standard practice is to use a " common mode choke" - which has two
windings carrying the AC supply current in opposite directions. This
prevents saturation with normal AC load currents.

http://www.coilws.com/toroidal/CM_Series_toroidal_c_choke.php

Even the smallest take 1 amp rms with ease.



....... Phil

 

[[email protected]]

<[email protected]>






** Standard practice is to use a " common mode choke" - which has two
windings carrying the AC supply current in opposite directions. This
prevents saturation with normal AC load currents.

http://www.coilws.com/toroidal/CM_Series_toroidal_c_choke.php

Even the smallest take 1 amp rms with ease.

...... Phil- Hide quoted text -

- Show quoted text -

Thanks for the help guys.

 

[Klaus Kragelund]

Thanks for the help guys.- Skjul tekst i anførselstegn -

- Vis tekst i anførselstegn -

Normally the CM choke has differential inductance also, so it will
saturate at a given current. You need to design for the peak current,
but often the CM choke is spec'ed with a RMS current and they can
handle a lot more peak current. The peak current is not shown in the
datasheet, so you need to test the actual peak current of your device
(the derate according to temperature, tolerances, etc). Normally
permeability rise with temperature

One core I just worked with 45% rise of permeability, so you need to
derate current accordingly. The same CM choke had a RMS current spec
of 0.5A, but a measured saturation current of 3.2A at 25 degrees

But, this is differential current. Be ware of the CM current. (caused
by stray currents to ground). The same CM choke measured only 7mA of
common mode saturation current. And this current running in your
product is difficult to measure. The end result is that the CM and DM
saturation currents add up. So if this choke is run at 3.2A peak DM
current, then it will tolerate 0A CM noise current.....

Regards

Klaus

 

[Phil Allison]

"Klaus Kragelund"

"One core I just worked with 45% rise of permeability, so you need to
derate current accordingly. The same CM choke had a RMS current spec
of 0.5A, but a measured saturation current of 3.2A at 25 degrees"


** A peak to rms ratio ( crest factor ) of 6 or more is extraordinary.

Never be seen in any normal AC supply current draw waveform.

The usual is about 2 or 3.

The OP has nothing to worry about.



........ Phil

 

[Klaus Kragelund]

"Klaus Kragelund"

"One core I just worked with 45% rise of permeability, so you need to
derate current accordingly. The same CM choke had a RMS current spec
of 0.5A, but a measured saturation current of 3.2A at 25 degrees"

** A peak to rms ratio ( crest factor ) of 6 or more is extraordinary.

Never be seen in any normal AC supply current draw waveform.

The usual is about 2 or 3.

The OP has nothing to worry about.

What about the CM current? (how can you know his application?)

About the crest factor - the project I working on now is a 50W pump
with a crest factor of 6. (electronics on a low-impedance mains
network). And this pump is not "extraordinary", standard RFI filter
with diode bridge and large cap.. When you design you need to take the
mains network into consideration, are you working on a soft net since
you have so low peak currents?

Regards

Klaus

 

[Phil Allison]

"Klaus Kragelund"
"Phil Allison"

What about the CM current? (how can you know his application?)

** The OP described it - shame you are a blind fuckwit.

About the crest factor - the project I working on now is a 50W pump
with a crest factor of 6. (electronics on a low-impedance mains
network).

** Yawn ............

**** off - you idiotic PEDANTIC IMBECILE !!



........ Phil

 

[Klaus Kragelund]

"KlausKragelund"
"Phil Allison"





** The OP described it - shame you are a blind fuckwit.


** Yawn ............

**** off - you idiotic PEDANTIC IMBECILE !!

....... Phil

Plunk ;-)