| gabriel rosenkoetter on Tue, 11 Feb 2003 09:23:04 -0500 |
|
On Mon, Feb 10, 2003 at 11:27:19PM -0500, Paul wrote:
> Are you saying that the polarity of the line could be the problem?
No. Polarity shouldn't matter with the telco hardware involved,
it'll compensate. (At least, I've never given any thought to which
wire I patched to which and I've never had problems, and you can
even beige box without any thought as to which wire's which.) I'm
saying stupidity could be the problem.
Here comes the amateur telco rant:
You've got a cable, right, and it's got five pairs in it (orange,
blue, brown, green, violet, and their related whites--remember, no
stripes in this cable). So, in my first case, Mr. Verizon tech
attaches those cables like this in the NID:
NID
O --[ ]-- my DSL
BrW--[ ]--
Bl --[ ]-- my phone line
BlW--[ ]--
Br --[ ]-- nothing
OW --[ ]--
G --[ ]-- neighbor's phone
GW --[ ]--
V --[ ]-- nothing
VW --[ ]--
You can perhaps see how I ran into trouble with the orange pair in
the above...
(Note that this isn't a shocking error to make, as DSL lines--when
they're a loop separate from any phone line, as my ADSL was, having
been installed before line-sharing was common practice for ADSL
lines around here--carry no dial tone. So this guy didn't bother
to tone out all the lines manually and didn't realize that he'd
disconnected something because, so far as he could tell, the line
was just dead.)
The second and third errors (the third error really just being
correction of the second which didn't take my having worked around
the second error into account) involved a 50-pair NID in the basement
of another building (there are a total of 4 NIDs between the pole
and the DSL bridge in my old apartment).
Old-style 50-pin NIDs are a set of screw down access points, 25
pairs on the left, 25 on the right. Originally, I'd been on pair,
oh, 38 or so in this NID, which didn't make a lot of sense (or line
up with the NID to which that one was directly connected by a
stretch of 50-pair cable, which was always a bit curious to me).
The plan, clearly, was to switch me to pair 26 (which *did* match
with the other end of that 50-pair), but pairs were mismatched in
that 50-pair patch to the next NID, so that you had something like
this:
NID 1: NID 2:
26 Bl OW 26 O OW
27 O BlW 27 Bl BlW
So I toned this out, found the error in NID 1, and switched it. Then
Verizon came back the next day and rewired the back of one or the
other NID (I actually don't *know* which one was off), and I had to
go back and switch the patch on NID 1 again.
On Tue, Feb 11, 2003 at 01:10:36AM -0500, William H. Magill wrote:
> One easy way to see if you are getting any carrier at all from the
> DSLAM... pickup the phone and unplug the splitter, turn off (or
> disconnect from the phone line) your local modem. If you have a DSLAM
> on the other end that is talking to you, you will hear standard modem
> hash. But even if you do have carrier, it could still be "wedged."
If you've got a separate loop for the DSL, the DSLAM won't be
generating tone on it. That only happens with line share installs,
and if you hear tone, it just means the phone line portion of the
line's working right. Paul, if you've got a line share and you're
having my kind of problem, you wouldn't have a working phone right
now either, so William's suggestion that the DSLAM needs a kick in
the pants seems far more likely in your case (since nearly all ADSL
installs since mid- or late-2000 in the Philly have been line share
installs; you still get a separate loop for SDSL, and there's still
no tone on it unless someone at the CO gens it).
> Don't forget -- if you go after the 192.168.0.1 address, you will need
> to change your own subnet mask and gateway manually to get to it if you
> have a real one active, or have nothing configured.
Nah, just add a static route to it. Under Linux, that'd be something
like:
route add 192.168.0.1 dev eth0
Win32 has a route command too, I believe, but I've never known how
to use it.
On Tue, Feb 11, 2003 at 01:28:06AM -0500, William H. Magill wrote:
> ??? I thought that all ADSL modems were REAL modems -- talking DMT
> (Discrete Multitone) with additional features added. DSL actually being
> an analog service, not digital like ISDN. That's why the frequency
> filters (microfilters) -- voice/fax/classic-modem goes below 3.3k hz.
> and data above 35Khz. (We may be saying the same thing here, just using
> different points of view.)
I'm on thin ice replying to this, but I was under the impression
that we'd pushed DMT about as far as it could go with 56k modems and
that DSL "modems" were actually talking PPPoE (and maybe ATM across
that). But please don't quote me on it without checking a real
reference (I'm sure IEEE has one).
> Westel apparently does offer a variety of models, from simple dumb
> modems, to those which act as bridges, routers or firewalls, depending
> upon the software / firmware load.
Well, then I'm certainly out of my water. The only DSL "modems" I've
had have been acting as bridges.
--
gabriel rosenkoetter
gr@eclipsed.net
Attachment:
pgpFqaXxjozCk.pgp
|
|