[Asterisk-Users] Re: Woodpeckers
Stephen R. Besch
sbesch at acsu.buffalo.edu
Fri Feb 20 11:53:16 MST 2004
Chris Albertson wrote:
> --- Steve Underwood <steveu at coppice.org> wrote:
>
>
>>A power spectrum plot will tell him he has a 60Hz hum. I think he
>>already knows that. I think he can definitely consider solutions
>>without
>>following your suggestion. :-)
>
>
> No, It's not a "60Hz hum". Yes, 60Hz is getting into the line
> but the existing filters are removing the 60hz. What he hears
> is most likely 120Hz, 240Hz or something else or most likely
> a combination of various multiples for 60hz.
>
> I'd bet that the tiny speaker inside a telephone handset can not
> even reproduce a 60Hz tome. Yes you can hear a hum but it's
> the overtomes of 60 that you hear. Many people can not even hear
> down to 60Hz, some can but not everyone.
>
????The lowest note on the piano is A0=55 Hz. I don't know a single
person who cannot hear A0. Most people can hear down to 30 Hz, a lot can
hear down to 20 Hz and some can even hear as low as 10 Hz.
> If you were to design a filter wouldn't it be nice to know some
> thing about the noise? Is there a big peak at 360? how broad is
> that paek 5hz or 20hz?
It's alsmost assuredly power line stuff - very narrow peaks (0.1 Hz or
less). Harmonics derive from non-linear bits in the signal path -
rusty/corroded connections, salt in contact with wires, weird magnetic
induction effects, etc. Even if rather badly distorted, the energy at
the higher peaks won't be too large, and anything above the third or
fourth harmonic will be negligible.
I would expect the power spectrum of a
> "hum" to have multiple peaks.
>
I love these discussions. Anyone who has actually tried to realize real
filter designs should be having a real laugh right about now. While it
is be possible to filter the dominant 60 Hz fundamental component (where
80-90 percent of the energy should reside), and even, with some effort,
to knock off the most energetic harmonics, why bother. The far end will
still hear the hum. On the other hand, if you actually load the phone
line to shunt the common mode portion of the 60 Hz signal, you will
badly mess up line balance (echo anyone). If you shunt the differential
mode portion, unless you really use sophisticated circuitry, you will
kill the ringing voltage - as was pointed out, getting even 2 poles of
rejection with passive components at 60 Hz requires some pretty hefty
stuff (90 Hz is not even an octave away), and active solutions are a
real joy, considering that all of the easily designed topologies use
op-amps with +/- 15 volt supplies (remember that 90V ringing voltage?) -
and it still won't be bi-directional like passive designs - which means
that you will need two hybrids with a filter on each 2-wire leg between
them. I could go on (and on). Ah, the joys of filter design. This kind
of approach is a bandaid for something that is really someone else's
(i.e., the phone company's) problem. I echo (sic) the advice of several
other posters: go after the phone provider.
Incidentally, my brother works for Verizion, and they have fond names
for squirrels and other critters that attack aerial lines - it keeps
them in work after all. When on line duty, he spent most of his time
tracking down line faults just like the one(s) described here. Verizon
was (is?) heavily invested in line quality and considered it a priority
to resolve customer complaints. I find it amazing that your provider is
so recalcitrant, when in fact it is their problem.
Oh, by the way, it has been my experience that, sometimes, when
woodpeckers attack houses, it is because there may be food (i.e. insects
- like carpenter ants or termites) to be had, in addition to perhaps
just being on the hunt for a place to live. If there are a lot of
smaller holes, especially if they seem to attack only localized areas of
the house, then perhaps you should consult an exterminator as well. I
know that this last bit is a somewhat off topic, but hell, it does
involve a possible phone call or two, and probably one using * to boot!
More information about the asterisk-users
mailing list