Canadian Broadband → Re: Feed-Back from a grateful customer!

{Whisper mode}

Hi Rocky,

If you mean this stuff above then no, but at least
it has removed any remaining doubt in my mind that
the shunting capacitors which are marked "C" might
be interacting directly with devices in the house.

I discovered that "hammer pulse dialing" cured the
problem on a temporary basis and even sometimes it
sufficed to only grab the line a few times to gain
INet access again but i no longer experience these
semi-permanent disruptions, the condition to which
you refer is over yet the cause hasn't been found.

It seems like Bell's technician was right to worry
about the high-impedance short he could measure on
the line:  it must be related to humidy & weather!



There's one interesting point to make here:  i 1st
saw in this cumulated error rate graph the history
of a failure influenced by temperature (going cold
as sunlight became darkness);  now i realize there
probably were few or no phone calls initiated from
the phones to "wash away" the mysterious shorts...



Of course, by writing "influenced" i mean that the
correlation between failures and the weather isn't
perfect:  in Louiseville this week-end was made of
spring meltdowns, snow, ice and rain, as i recall,
but i can't exclude other explanations such as the
electrical noise from seasonal activities in local
industries, eventually, heavy P2P transfers from a
bunch of leachers generating lots of cross-talk or
maybe even some hacker attacks, etc...  Go figure!



Oh, i almost forgot to comment about my error rate
curve relatively to the periods which occur during
shortages:  it's dead flat because access is lost.



Once i'm satisfied that the usual noise present on
this line will be handled by the ST546v6 correctly
i'll take my GNet BB0060B MoDem away but i'm still
trying to decide how i can implement an equivalent
diagnostic tool to 'DMT v2.19' in the form of some
Slax Linux "Live CD" to be run on a dedicated P-1.
Please drop me a note if you hear of such threads!





Addendum
¯¯¯¯¯¯¯¯

I've put back on line my SpeedTouch 546 during the
last week-end so here's what i could come up with:

           Up Time (in seconds):  8415 20510 61864 100764 127309 169262 196597 213668
    Received (local) CRC errors:   479   502   724   8174   8994  11753  11928  12115
    Received (local) HEC errors:    87    92   250   2534   3023   3777   3827   3908
Transmitted (remote) CRC errors:     0     2   102    102    104    108    108    108
Transmitted (remote) HEC errors:     0     0    62     62     64     66     66     66
          Total errored seconds:    77    95   219    639    932   1346   1384   1486

You'll find attached a graphic which i produced by
feeding my calculator with these numbers, it seems
there was a burst of errors around the 3rd record.