EMC Filters

To: tesla-at-pupman-dot-com
Subject: EMC Filters
From: "Jim Lux" <jimlux-at-jpl.nasa.gov> (by way of Terry Fritz <twftesla-at-uswest-dot-net>)
Date: Tue, 01 Feb 2000 17:48:27 -0700
Approved: twftesla-at-uswest-dot-net
Delivered-To: fixup-tesla-at-pupman-dot-com-at-fixme

The jan/Feb 2000 issue of Compliance Engineering (http://www.ce-mag-dot-com/ is
the web site, but the magazine isn't there) has an article on "Choosing and
Installing Mains Filters"  (p 68 to 75) which provides a lot of practical
applications info with standard EMC filters.

An interesting point they make is that the filters are typically specified
(and tested)  with 50 ohm in and out impedances, which aren't particularly
representative of real applications where the source impedance may be quite
low (0.01 ohm) and the load is around 100 ohms ( 100 Watts, roughly).  In
such applications, the filter's attenuation may be approximately the same at
higher frequencies (>10 MHz), but may actually have "resonant gain" for
frequencies of 100 kHz to 1 MHz.


What this means is that some filters may actually make the RFI problem worse
by acting like a "matching network" between the source (the TC) and the load
(the power line) and transferring more power.

For reference, I've put some schematics of line impedance stabilization
networks (used in EMC testing) on my web site.
For 0.15 Mhz to 30 MHz http://home.earthlink-dot-net/~jimlux/hv/images/lisn1.png
For 10 kHz to 150 kHz http://home.earthlink-dot-net/~jimlux/hv/images/lisn2.png
Both as a visio drawing http://home.earthlink-dot-net/~jimlux/hv/images/lisn.vsd

These networks provide a "standardized" source impedance to the EUT
(equipment under test) that is representative of some standard line source
impedance (at RF)...  Note they've got lots of reactive components, all of
which can interact with a reactive load.


Jim Lux
phone:818/354-2075  fax:818/393-6875
Spacecraft Telecommunications Equipment Section
Jet Propulsion Laboratory   M/S 161-213
4800 Oak Grove
Pasadena CA 91109