On Tue, 25 Aug 1998, Muriel Bittencourt de Liz wrote:
> Hello,
>
> I'm working with EMC at the field of power electronics and i'm
> interested in developing filters for conducted EMI. I've looked for
> designs at some books, but they aren't very practical.
>
> Could anyone send me some information 'bout some practical "guide to
> design of EMI filters"? Any information will be welcome.
>
> Muriel
First, is the filter general, or specific?
If specific, first characterize your noise source.
After all you're making a filter and need to know the impedances and
magnitudes involved.
For example, a switching power supply has low impedance differential
voltage in the low frequency range (this is the voltage developed across
the esr of the capacitor during peak current demands). It also has very
high voltage high impedance tones as you go up in frequency (these are
from the transistor's and transformer's parasitic capacitance) which tend
to be "common mode".
Anyway, you get the idea. QUANTIFY and CHARACTERIZE your noise source.
Then most filters (determined by the characteristics of the noise source)
have two chokes that can handle peak AC current (It's difficult to get
these more than 20-50uH)
Next go to a y cap to gnd within the limits of regulatory compliance. Use
a y cap, not two caps, because the tolerance of unbalaned capacitances can
actually inject noise into your filter.
Place a cap between the lines. 1uF is fairly large physically.
Then put in your common mode choke (constructed *very* well) These are
easily 1-2milliH. And top off at the AC mains with the largest of your AC
mains caps (1uF is usual).
Put the whole thing in a metal box (you'd be surprised how much field gets
out the gap of a power transformer) and you need this to work up to 30MHz.
Finally, do an actual breadboard and modify parts values appropriately.
If the filtering is still insufficient you may have to repeat stages to
get the noise down. Again, you're just making a filter.
- Robert -
PS for anaylsis I use PSpice demo version. Just don't forget to
characterize the LISN, line cord, and conductive surfaces, too.
---------
This message is coming from the emc-pstc discussion list.
To cancel your subscription, send mail to majord...@ieee.com
with the single line: "unsubscribe emc-pstc" (without the
quotes). For help, send mail to ed.pr...@cubic.com,
ri...@sdd.hp.com, or roger.volgst...@compaq.co (the list
administrators).
