[plug] modem dropouts, radio interference

[Denis Brown]

David Buddrige and others have commented about radio interference to
modems.  Given that David is quite close to our 6WF transmitting tower (and
sundry other signal sources), this is likely to be the culprit.  The idea
of inductor / capacitor filters seems likely as a fix.  Perhaps this is a
starting point (if you have the patience / interest to pursue it) and is
based on a similar experience I had many years ago as an apprentice in the
radio & television game, back when there was a 6PR transmitting tower down
at Applecross.  The client's stereogram was getting flooded with 6PR and I
successfully treated it with a 'trap" at the desired frequency...

1. Temporarily move the computer system, modem, etc to be near the main
house telephone point (i.e. temporarily eliminate the extension cords.)
Given the quick and repeatable dropout experience, that should be
sufficient to verify the effects of the extra cables in a very short time.

2. Assuming things are more reliable in the new (temporary!!!) location,
roll up the extension cords and connect them into the telephone circuit.
Ideally, to prevent them acting as an antenna, physically measure half-way
and roll one half clockwise, the other half anti-clockwise.  May be
overkill but this will nearly cancel the induced interference.  If the
modem remains reliable, the extension cables, per se, are probably okay.

3. (optional) Re-roll the cables in the one direction so they act as an
antenna and try various orientations of the roll.  Some, broad side on to
the transmitter, should produce little interference-related dropout (the
antenna picks up little signal) while at right angles to the tower there
should be lots of interference.  With the roll lying on the floor, there
may be some interference based on near-field effects.  Sheesh... 600 m from
the tower!  You probably walk around in the backyard just carrying a bare
fluoro tube to get light at nights!!

4. Appease your wife by moving the computer back to its usual location and
visit your local friendly DS store (or similar).  What I suggest (others,
suggestions please!) is a series-tuned trap at each end ot the extension lead.

At the risk of telling you what you may already know...  Series-tuned traps
shunt signal currents at the resonant frequency.  According to my
calculations, 400 picoFarads and 4.5 milliHenrys or 330 pF and 5.8 mH
should be close enough.  Maybe you'll get close enough by say 390 pF and
4.7 mH which should be standard values.  Could someone with a less-rusty
calculator check the maths, please?  The inductor values seem a bit high to
me.  If DS or Jaycar don't stock the parts, try Radio Spares in Osborne Park.

Anyway, you buy 2 inductors and 2 capacitors.  Connect an L and a C in
series and connect the free ends (one L, other C) to the active wires in
your extension cable, one LC at each end.  There will be no unpleasant
effects on the operation of the telephone or modem because the capacitor is
effectively an open circuit at direct current and very low frequencies
(dialing, ringing and modem signalling tones.)  At the resonant frequency
(6WF transmits on 721 KiloHertz if memory serves) the inductive and
capacitive reactances cancel out so effectively the traps absorb all the
energy.  It worked well for my 6PR experience!

Another correspondent suggested L and 2 C's.  That would be a "Pi" filter
and they are very good, too.  Someone with less rusty maths may be able to
give you the values and I think it'd be good to have one at each end of the
cable, especially at the modem end.  The Pi filter works by blocking signal
currents (at the resonant frequency) from flowing along the cable,
especially at the resonant frequency.  They are especially good for
preventing locally-generated signals from being radiated BY the connecting
cables.  What we have here is the reverse -- we already have the
interference on the cable, we want to prevent it getting into the modem.

Hope this has been interesting / worthwhile.  Hope to hear of a win, David!

Cheers,
Denis