Tom,
Couple of questions for you if you don't mind?
While reading your last reply to Rob, it triggered a recollection on my part of the effects of "sealing current" on metallic loops. For onlookers, sealing current is a steady state battery on a loop which "binds," or "seals" bad splices and frame contacts that would ordinarily cause noise and "open" or swinging open conditions. You could think of the current as being responsible for "jumping" or "zapping" the chasm of bad connections and preventing discontinuities from occurring. In darkness, you could actually see a blue-orange "arcing" on a bad splice, if you set up the experiment properly. And the splice would hold firm, until, at some point in time after the current was removed, and then you would see a gradual deterioration until the circuit went "open" entirely (or on one side).
Years ago I did some experimental work along with the Labs on some live, problematic circuits in New York City << believe me, we had our share of them at the time>>, and I was convinced beyond any doubts, whatsoever, that sealing current was the remedy to many (not all, of course) metallic loop problems. Especially those that could be attributed to oxidation, loose contacts or splices, and the residual long term effects of humidity.
T-1 lines and DDS services are routinely provisioned with sealing current, although the current plays other important roles than just binding the circuit intact (such as signalling, loop powering and diagnostics). It is my opinion that this is what gives the DDS loop (even short head-to-head ones in the same central office) better error performance and low trouble rates.
Typically, sealing current works best (or at least it is more conveniently applied) on four-wire circuits, due to the ability of one to apply it in a balanced and neutralizing manner, between both directions of the circuit -- using center taps on the isolation coils. And while we were able to apply it to two-wire (single pair) lines, it was tenuous and had less engineering integrity than we would have liked. How about two-wire xDSL lines? Will they be dry or wet?
[For the benefit of those looking in, "dry" refers to the absence of any steady state voltage/current on the line, and wet alludes to a steady battery that has been applied to the line. POTS services are typically wet, due to the constant -48 volts applied on them at all times -- perhaps this is why dial tone lines do not go into noise conditions as frequently as analog <dedicated> data lines.]
Another question, at the risk of overstaying my welcome, is how many points of cross-connection are permitted on xDSL loops, from DCE to DCE? Is this a parameter that is, or will be, defined? I.e., will there be a rule that states that there shall be no more than x points of discontinuity, or y number of splices? Again, for the benefit of onlookers, every time a signal traverses a cross-connect or a splice/punch-down, a part of it is "reflected" causing undesirable transmission effects such as echoes, and increased attenuation and return loss.
Any information or opinions you may have on these issues would be greatly appreciated.
TIA, Frank |
