On 07/16/2012 10:26 PM, Jim Brown wrote:
> This is a differential mode filter, so it suggests that some or all of
> the telco cable feeding the modem isn't twisted pair, or isn't very good
> twisted pair. SO -- another solution could be replacing that poor cable
> with CAT5, using the pair within the CAT5 that has the highest twist
> ratio. Shielding is NOT important, but TWISTING is. Un-twisted cable is
> notoriously bad for RF rejection.
Thanks Jim.
I should probably have added the rest of my story. Along with the
common mode chokes (which I believe were 16 turns on a 2.4" #31
core, with several in series), I also replaced the wire from the
telco NID outside my house to the modem. The cable I used isn't CAT5
but has two rather tightly twisted pairs and a shield. This didn't
seem to help either. I disconnected all phone and DSL wiring from my
side of the NID and hung my modem right next to it on the outside of
my home. That didn't make any difference. I tried a few different
arrangements for providing power to the modem. Details forgotten as
it was several years ago, but nothing I did seemed to make any
difference. The level of 160m RF required to lose DSL remained very
near 50 watts throughout all this experimentation. After all that
and seeing the OZ1CTK filter magically cure the problem, my hunch is
that the real issue is outside my home and beyond my control. I may
be overlooking something.
What about the drop from the pole to my NID? That is not twisted.
I've forgotten the proper name for the stuff again, but it is a
parallel pair of #18 (or so) copper clad steel conductors with a
very indestructible black insulation. It's the stuff that looks like
zip cord. In my case there is about 150 feet of it.
Unrelated: Just because it made me feel better, I threw away the DSL
filters that were to be installed on my phones and put a DSL/POTS
splitter outside next to the NID.
73,
Paul N1BUG
_______________________________________________
UR RST IS ... ... ..9 QSB QSB - hw? BK
|