bud@kentrox.uucp (Bud Couch) (06/01/91)
In article <TERRYK.91May30205639@pinocchio.encore.com> terryk@encore.com (Terence M. Kelleher) writes: > >In article <1991May22.014949.14165@ims.alaska.edu> floyd@ims.alaska.edu (Floyd Davidson) writes: > > Just to add a little to this... Telephone switch systems are > also "fixed" at 10 pps, so changing it would most likely > cause errors (8-12 pps would probably work, out of that range > and the switch will think it isn't dial pulse anymore). > >In reality, telephone systems are extremely tolerent of the pulse >timings. On a Hayes compatible that I had the "pleasure" of being a >designer on, I expirimented with varying the times. It was quite >reliable from about 2pps to around 100pps. Of course, your line may >vary. > In reality, telephone systems, and I include the effects of the outside cable that gets your signal back and forth, are extremely variable. On telephone line equipment, trunk equipment, and carrier equipment that I have had the "pleasure" of being a designer on, the biggest single variable that had to be accounted for was the afformentioned cable characteristics. The specifications for pulse dialing in the U.S. are 8 to 12 pps, 58% to 64% break (-to-make ratio), *at the dialing point*. Further distortion is introduced by the cable line, but the receiver is designed to accomdate this. Testing on any given line may indicate acceptable performance, but this cannot be construed as providing carte blanche to use the tested dialing parameters elsewhere, or even at a different time. KEEP THE DIAL PARAMETERS IN YOUR MODEM BETWEEN 8 AND 12 PPS, 58-64% BREAK, AND YOU WILL NOT HAVE PROBLEMS. If you want to get real theoretical about it, refer to _A Theoretical Analysis of Telephone Dial Pulse Signaling_ by Muchalak, et al, in the IEEE Trans. on Comm. V COM20, No. 6, Dec. 1972. -- Bud Couch - ADC/Kentrox | And diff'ring judments serve but to declare | If my employer only knew... |That truth lies somewhere, if we knew but where| standard BS applies | -William Cowper |