Hi Steve, <ADSL will be a telco Vietnam, far worse than ISDN was.> Can I plagarize that line someday? I think the extent that statement becomes true will depend on how widespread xDSL becomes over long copper pair loops (more than two or three miles from a central office).
<First, rough estimates are that 40% of the existing copper plant won't support any form of ADSL.>
That sounds about right. My telco owned ISP confirms that guesstimate for their proposed DSL rollout:
public.pacbell.net
<telcos don't know WHICH 40% will NOT support it. They can't even regional this. It might work OK for me, but it may not work for either of my next door neighbors.>
Yes, distance is the big, but not the only obstacle. The 16,000 ft. maximum range is mentioned in the above link is the big limiting factor. Digital signal attenuation takes over after that and the signal becomes too weak. Noise on adjacent cable pairs drowns the digital signal and it becomes a bit error on the other end. The quick, crude way to extend DSL on copper beyond 16,000' becomes part of that Vietnam syndrome you mentioned: install DSL repeaters. For a mass deployment, a very bad move. Of course, if DSL demand is big in a distant serving area, it justifies placement of a Digital loop carrier system with the ability to provide derived DSL services over short copper loops. The distance issue becomes mitigated, but this again takes substantial time and $$.
<xDSL killing load coils (placed to make the phone lines "cleaner" by "removing" copper's high frequency carrying capacity) have been installed across the country. And records have rarely been kept documenting which lines have load coils, and WHERE they are (often buried).>
For ADSL, load coils are almost a non-issue. First, load coils are only present on voicegrade copper pairs that serve customers located 18,000' or more from a central office via copper pairs (less than 25 percent of customer lines). More and more of the long, loaded copper pair loop plant is replaced by digital loop carrier as the loaded copper pair loops are used up. The DSL announcement I mentioned above (like other DSL deployments using similar technology) restricts availability to customers that live no more than 16,000 ft from a central office, so, see how load coils are not applicable? They simply would not be present on copper loops serving customers less than 18,000 ft from a CO. Even if DSL technology did permit transmission to customers on copper loops beyond 18,000 feet, it is a normal thing for telcos to determine this and de-load pairs. This is done on a daily basis for HDSL to provision T1 on copper pairs out to 30,000 feet, when HDSL repeaters (doublers) are installed every 10,000 ft or so, and loads are removed from the two HDSL pair. All telco outside plant assignment records and location records show load points by cable pair. You de-load by cable pair, so removal of loads on a pair for a digital service will not drastically affect an adjacent loaded pair for analog voicegrade service. Buried load coil cases? Well, there may be a few out there, but that is a rare species as far as I know. Even if the copper pair cable is direct buried, telco construction practices since the 1970's universally specify placement of a manhole, vault, or above ground crossbar on a pole stub for mounting load coil cases. Regardless, load coil cases are always installed at 6000 ft (+ or - 200') intervals, and I've never had trouble finding, designing, building, or removing them. The problem with loads is that in long copper pair loops, there is often a lack of spare cable pairs these days for digital services, and technicians end up having to de-load one pair or two pair at a time, instead of de-loading an entire group of 25 clean pair that is reserved for digital services.
I don't have much to add about your other conclusions. I agree that ADSL on existing copper pair telco networks on a large scale could get very messy. The one type of xDSL deployment that telcos might have more success with is a fiber to the neighborhood architecture, where the copper loop is only say 2000' to the home at most, with no bridge tap. That would remove many of the transmission and installation obstacles to a high volume of DSL customers, but again, we probably won't see this on a large scale for some years.
For the question of reliable ADSL service, if the service can be installed and pass an initial performance "soak" test, my guess is that it would follow the patterns of other telco provided services, which is in the 99+ percent average availabilty range.
Would I sign up for ADSL as described in the link above? Very doubtful. Cable modems are coming to my area within the next year for a $49 monthly flat rate. A separate topic!
Ray.
|
