rees@pisa.ifs.umich.edu (Jim Rees) (01/31/91)
In article <16472@accuvax.nwu.edu>, 0004133373@mcimail.com (Donald E. Kimberlin) writes: [ an excellent introduction to 4-wire lines ] > So, a local telephone plant uses only one pair per subscriber. > In engineering terms, it is far from a perfect transmission line. The > main reason is that no transmission line operates at its normal > electrical "impedance" until it is a significant portion of an > electrical wavelength of the signal it carries. That's not exactly right. A transmission line has a characteristic impedance, which in the case of your local subscriber loop is probably around 100 ohms or so. I don't know the exact number but it can be calculated from the dimensions of the wire. If the line is exactly matched to the terminating impedance, then you don't get any reflections, regardless of how long the line is. The problem with local loops is that they aren't a good match to the endpoints, which have an impedance of around 600 ohms. But even if the characteristic impedance of the line were carefully set to 600 ohms, it wouldn't matter until the line becomes a substantial fraction of a quarter wave long, which would be at least several miles. Any resistance in the local loop will also make the loop deviate from an ideal transmission line. I think the biggest cause of reflections is probably the subscriber phone set, which doesn't present an exact 600 ohm resistive termination at all frequencies. I would guess that the resistance of the subscriber loop is secondary, and that actual transmission line effects are last. Can anyone confirm this?