[RFI] ebay line filters

[Jim Brown]

Hi Doug,

See comments below.

On Mon,8/15/2016 9:18 PM, Doug Powell wrote:
> You made my points.
>
> In the end a commercial equipment manufacturers can only design ‎and test to a standardized test setup. This is exactly what's described in CISPR 11, CISPR 22, ANSI/IEEE C63 and national derivatives of those standards. It is up to the final equipment manufacturer to use these filters in such a way as to take good advantage of them.  Companies like Corcom, Schaffner, Cosel, TE Connectivity and others cannot have official knowledge of how these filters are finally used in every application. If a Ham wishes to use a catalog part, they are pretty much on their own to identify a suitable model. Unfortunately, simply looking at data sheet specs is not going to reveal all the answers needed. Case by case experimental data is almost always needed.

WRONG.  You're talking about designing to the test (the EMC standards), 
which is the same as school teachers teaching to the test (another 
arbitrary standard). Neither necessarily produces a product that works 
in the real world.

While you've described the tests for compliance with standards, they're 
not meaningful. Those tests don't measure the mechanisms that cause RFI 
at MF and HF. The problem is IMPROPER DESIGN. Once you understand the 
mechanisms that CAUSE it, it's easy to correct them. Go study the 
material on my website about "The Pin One Problem," which is described 
in the context of both baseband and RF EMI to audio systems, but it is 
the same design/construction error I described to you in the last email. 
And, as we know, ALL passive networks are bidirectional, so the same 
error that couples AF and RF noise INTO the box also couples it OUT of 
the box. Designers simply need to avoid this design error to prevent 
putting common mode trash on the green wire.

Your description of "common mode" as the voltage between neutral and 
green is the POWER description of common mode, but that is NOT THE 
PROBLEM that causes RFI. RF noise on the MF, HF, and low VHF bands is 
caused by that common mode current, which is current flowing LATERALLY 
along the power line. So the power line is RADIATING because there is 
current flowing along it. It radiates like any other wire carrying RF 
current by simple antenna action.

There are two solutions. 1) Correct the Pin One Problem -- terminate the 
green wire correctly, terminate cable shields to the shielding enclosure 
DIRECTLY, NOT VIA THE CIRCUIT BOARD. 2) Choke the cable with the 
improperly terminated green wire (or the audio/video/coax cable) with 
the improperly terminated shield.

> That said, there are some generalities that may be applied as a result of past experience.  It is instructive and a bit of fun to look at Henry Ott's tongue in cheek article on maximizing your emissions.
>
> http://www.hottconsultants.com/techtips/maxemission.html.

I haven't seen this, but Henry and I have been friends for more than ten 
years. Henry is a guy who really understands EVERYTHING, and who is also 
able to teach it very well. BTW -- Henry is quite familiar with "The Pin 
One Problem" with respect to audio. Several years ago, I arranged for 
him to give expert lectures to Audio Engineering Society conventions on 
EMC.

73, Jim K9YC