[time-nuts] Frequency from telco
cfmd at bredband.net
Sun Sep 24 19:55:35 EDT 2006
From: "Poul-Henning Kamp" <phk at phk.freebsd.dk>
Subject: Re: [time-nuts] Frequency from telco
Date: Sun, 24 Sep 2006 23:30:55 +0000
Message-ID: <3148.1159140655 at critter.freebsd.dk>
> In message <0609242322.AA00795 at ivan.Harhan.ORG>, Michael Sokolov writes:
> >Hal Murray <hmurray at suespammers.org> wrote:
> >> Does anybody know how DSL works? If I poke around with a scope in my DSL
> >> modem/router, will I find a clock locked to the telco's master clock?
> >I doubt it. DSL lines come from DSLAMs, not from "real" telco equipment
> It's more complicated than that. Most DSLAMs can are hooked
> into the network sync because people once thought they would
> also be used to deliver isochronous services like voice.
Yeah, who would be so stupid to put voice over DSL???
(Irony dripping all over the place.)
> However, the CPE ("modem") in your end, has absolutely zero clue
> about this.
For ADSL DSLAMs having an ATM interface, chances are that it is being fed
through the SDH pipe, as the ATM specs now requires.
But then again, Gigabit Ethernet interfaces is popular these days, and then
there is no telco sync. Smaller DSLAMs have Fast Ethernet by the way, they are
really cute! :o)
If the DSLAMs where properly hooked up, it is a simple matter of finding the
pilot tone and lock on to that.
Besides, the sync signal may not be that great anyway, wandering around. That
is a real issue. There is no guarantee for UTC traceability, but most of the
time there is traceability to a cascade of Cesium beams.
The end story is: Don't keep your hopes up!
More information about the time-nuts