[RFI] ebay line filters

[Jim Brown]

Hi Doug,

You're talking about 1) properly built equipment and 2) measurements in 
a lab.

The problem is that much (most?) equipment is NOT properly built -- the 
green wire SHOULD go to the shielding enclosure, the shielding enclosure 
of a filter integral to a product SHOULD be bonded to the shielding 
enclosure, a filter external to a product must have it's shielding 
enclosure bonded to the shielding enclosure of the product, and for that 
external filter, all three leads (phase, neutral, and green) must be 
"zero length" (that is, very short).  What happens far too often is that 
the green wire goes to a trace on a circuit board that may or may not 
every see the shielding enclosure, and there is impedance between that 
connection to the circuit board and the shielding enclosure. There's 
also equipment where the green wire is insulated from the shielding 
enclosure by paint. All of these conditions put the green wire above the 
chassis by some amount, so if the product is noisy, that noise ends up 
on the green wire, which causes current on the green wire, which causes 
the noise to radiate.

Second, real world conditions are not like that test lab setup. The 
green wire can be pretty long, which makes it a more efficient radiator 
at RF. And those lab conditions aren't looking for radiation, they're 
looking for voltage between Phase and Neutral (which they call 
differential) and between Neutral and Ground (which they call Common Mode).

73, Jim K9YC

on Mon,8/15/2016 4:26 PM, Doug Powell wrote:
> Good points,
>
> It is indeed true that common mode noise is with reference to the green wire ground, but back at the measurement point and not the case of the filter. In the case if Mil Std filters this may be on a grounded sheet metal table.  In the case of European approvals this measurement point is on the floor of the semi-anechoic chamber ‎and the test sample is 80 cm above on an insulated table. Establishing an RF ground up there is extremely difficult and gets worse as you go up in frequency. So, if you buy a filter approved to IEC, CISPR, or EN standards you are likely to get better results than one that is compliant with Mil Std.
>
> All the best, Doug
>
>
>    Original Message
> From: Jim Brown
> Sent: Monday, August 15, 2016 4:24 PM
> To: rfi at contesting.com
> Reply To: jim at audiosystemsgroup.com
> Subject: Re: [RFI] ebay line filters
>
> On Mon,8/15/2016 6:49 AM, Doug Powell wrote:
>> One thing you should realize about line filters is they are optimized for conducted line emissions coming from the product where they are built in. This means the line / load labeling ‎will determine your best orientation for connection.
> There's another MAJOR problem with commercial line filters -- most of
> the trash conducted from equipment is on the GREEN WIRE, which goes
> right past the filter. The specs for these filters CLAIM common mode
> suppression, but what they're calling common mode is voltage between
> neutral and ground. That's NOT what WE call common mode, and what WE
> call common mode is what causes most RFI -- radiation from the power
> conductors as a long wire antenna.
>
> The most effective line filter for MF and HF radio is a common mode
> choke formed by winding multiple turns around a #31 Fair-Rite ferrite
> core, using the guidelines in k9yc.com/RFI-Ham.pdf for coax the same
> size as the power cable. See Chapter Seven the "Choke Cookbook." Also
> see http://nccc.cc/pdf/CQP-RFI2013-2.pdf
>
> 73, Jim K9YC
>
> _______________________________________________
> RFI mailing list
> RFI at contesting.com
> http://lists.contesting.com/mailman/listinfo/rfi
>