[RFI] DSL Filter Update

Jim Brown jim at audiosystemsgroup.com
Mon Mar 9 10:43:27 PDT 2009


On Sun, 8 Mar 2009 20:43:14 -0500 (CDT), Christopher E. Brown wrote:

>Depends on the DSL gear, ADSL1 cuts off around 1.1, VHDSL uses much higher 
>freqs.

>Common use these days in the US is ADSL2+, with ADSL1 hardware cycling out 
>of production.


>Also, remember that RF can impact *BOTH* ends.  I have seen a couple cases 
>where enough signal made it into the line to impact both the CPE *and* the 
>DSLAM, as well as causing increased error rates for neighbors in the same 
>cable bundle.  A serious common mode choke at both the CPE and at the NID 
>could be needed in severe cases.  At my home I have chokes at both ends of 
>the line, only a single cylinder but my original DSL issues were not so 
>severe.

Many thanks, Chris. 

For those of us not working in the telco world, please translate DSLAM, 
CPE, and NID. :)  

Your point about RF entering the cable at one point and traveling along the 
cable for some distance is well taken, although I would expect that to be 
most problematic below 4 MHz. I could also see the possibility data errors 
introduced by one subscriber modem throwing a monkey wrench into the overall 
data stream. 

Several thoughts.First, my measurements of common mode chokes that have 
historically been effective at protecting telephone lines show that they 
have VERY high choking impedances (10K Ohms - 80K Ohms). A single cylinder, 
even a rather large one, is typically a few tens of Ohms at HF -- pissing in 
the wind. It might be useful as part of a differential choke, but not as a 
common mode choke. 

As for protecting the upstream side -- if there were equipment within  
several hundred meters of the ham transmitter, I could see protecting it 
with a common mode choke and a differential filter. 

It would also be quite helpful if the wiring in the vicinity of the 
transmitter was a very high quality twisted pair (like CAT5). It's too bad 
that virtually all commercially available RJ11 cables use parallel 
conductors. I suspect this is a major contributor to RFI to equipment on 
telco lines. Is anyone aware of RJ11 cables built with twisted pair? 

73,

Jim Brown K9YC




More information about the RFI mailing list