The PBX Scene: October 1986

Some Recent Events

We are suddenly confronted with a rash of new small PBXs from the makers of large PBXs: Rolm's Redwood, AT&T's System 25, NEC's Impulse, the new Saturn I from Siemens, and Mitel's digital version of the SX-200. All these seem to illustrate a little-explored aspect of "economies of scale," namely, the prorating of the cost of research and development over number of units sold. As the market continues to fragment before competing vendors, and each vendor has to pay about the same in R&D to bring a product to market, it is obvious that the cost of R&D per unit sold has to go up, demonstrating once again that competition sometimes causes some aspects of costs to rise, not fall. But given the competitive situation, manufacturers will do everything they can to cut R&D costs, along with everything else. Thus it is not surprising to see many features and types of hardware used on large PBXs move to small PBXs nearly unchanged. However, we may be surprised at some of the things NOT moved to the smaller PBXs. Assume nothing about small PBXs on the basis of their big brothers.

What is more surprising is the way considerable additional effort is being spent to get a first-cost price advantage. In particular, we find extra line and trunk cards being designed so that the customer can add circuits in smaller increments than, say, 8. That is, if you need 26 trunks and the uniform line/trunk slots come in blocks of 8, a card will be provided that has only 4 circuits on it, at a lower price than one with 8. In any kind of growth situation this is poor economy. But it does allow the first cost to be less. Another approach, of course, is to revert to NON-universal card slot for trunks as compared with lines so that trunks, more expensive than lines, can be added in smaller increments. The reader is urged to make every effort to understand the potential growth in a given installation, to be able to take full advantage of some of these new opportunities.

In addition to small PBXs, many of which cannot handle data, we also see many small LANs being announced in the computer magazines, and many big microcomputer systems handling multiple (say, 20 to 20) users. If, as some insist, small businesses do not need voice-data capability, these small data systems are not going to find much of a market. On the other hand, maybe the computer and data people know something the PBX people do not.

The problem is connecting the multiple users together, either to each other or to their shared computer. With the telephone wiring already in place, a properly designed voice-data PBX should be able to run rings around installing new wire for a LAN, and should be a good way to connect dumb terminals (CRT-based telephone sets?) to the AT, the AT&T, or even the VAX. But the telephone industry has a long history of trying to provide POTS at minimum cost while charging as much as possible for "value added" services (remember Touch-Tone, Centrex, Call Waiting, etc.). In general, all such value added services either save or make money for the telephone industry, so they charge the customer extra to pay for reducing their costs. The ultimate along these lines may be the ISDN.

But ISDN is taking a long time to arrive, and many small PBXs that incorporate several different ways of handling voice and data are pouring onto the market: Tadiran, SRX, Cyber Digital, etc. are joining CXC, Ztel, and Telenova, and both Redcom and Toshiba have added data capability to their small machines in the very recent past. Watching the daily announcements in the small PBX area, comparing the giants with the small start-ups, is the most interesting thing going on in the PBX scene today.

Alternate Voice-Data Revisited

The original version of data over the phone system was alternate voice-data. You would call up the computer of your choice with your telephone and, when you heard its modem squeal, you would operate a switch or stick your phone in an acoustic coupler to connect your terminal to the line. This tied up your phone, but all you needed was the modem and you were in business.

Danray, with an analog PBX, first clearly demonstrated the desirability of simultaneous voice-data, and showed how existing telephone wiring could handle both at the same time at a reasonably low cost. Most of us then began to think of telephone inquiries to a human being who would then access a data-base in real-time and relay the appropriate information to the caller. That is, we visualized information coming from electronic rather than paper data bases. Reservation and ordering systems illustrate the situation at one level, and Northern Telecom's Share, where two or more individuals can observe and modify jointly an existing data base goes a bit further. But the case for voice and data simultaneously seemed to be proven.

Now comes Harris, pushing their new 20-20 switch in its PBX mode. It has "Data for Free." That is, an RS-232 connector in the back of their new Optic Teleset lets voice OR data to use the digital channel from the set to and through the 20-20. No extra ports on the PBX. No extra hardware except the connector and a little logic. A separate control channel handles signaling and supervision, provides a call-waiting display to show the user on a data call when a voice call is being offered, and vice versa. One can put the data call on hold to take the voice call, then return to data.

The idea is very appealing. If people are not putting much data through their PBXs as yet, maybe the cost is too high for what they are presently trying to do. And if Alternate Voice-Data can be added at almost no cost, as Harris had done, maybe the key to handing data in the office of the present is here. It does what many of us are doing now with modems, maybe less expensively, and certainly at higher speeds (at least within the PBX). If you can dump a letter into the data base computer for your boss to review at 19.2 Kbps, the transmission won't tie up the phone very long.

Conferencing vs. Teleconferencing

The Harris consultant seminar on the 20-20 in my area was held on Aug. 26 at a hotel on the far side of Philadelphia, in a place called King of Prussia. K of P is located at the joining of two major arteries of auto traffic in the area, the Pennsylvania Turnpike and the Schuylkill Expressway. Thus the meeting should have been easy to get to, right? Wrong. Both the Schuylkill and the Turnpike are under major construction, and driving time in each direction was both nerve-wracking and lengthy. We were in the car for over 3 hours, for a one-hour seminar. The seminar was presented via satellite from Harris' headquarters in Florida, and technically, it worked pretty well. But getting there and back was a whole career.

Thus, when Xerox announced a teleconference for consultants to introduce their new desk-top printing products, I jumped at the chance. They sent me the paper copies of their slides by Federal, and at the witching hour I dialed the magic number. Right from my own desk in my office in South Jersey. A nice young lady connected me to a big conference bridge, and the meeting started (about 10 minutes late, but that was ok because we were told we could call collect). Anyhow, we all introduced ourselves to each other on cue from the conference-master, and the program began. Just voice leaves a bit to be desired, even with copies of the slides, but the program wasn't bad. However, about half-way through, my line was disconnected and I got dial tone. I went back, and was reconnected with no problem. But the second time I was dumped, I figured I had had enough. Xerox had some interesting things to say, and the interactive conference was interesting, but the if the medium was at least part of the message, one might suppose a VCR tape in full color and stereo sound, mailed to each consultant, would have done a better job, and perhaps at less cost.

What's in this Release

This time, we have an update on the NEAX 2400 and a first look at the NEC Impulse and the Rolm Redwood. And we also have some material on the AT&T 5ESS switch PBX. 5ESS is a gigantic CO switch that can do local, tandem and toll switching, can support operator positions, and do just about everything else that a switch can do. However, bringing it into the PBX market will be interesting, particularly when it will also be the main AT&T offering to telephone companies for Centrex service. It is also the gateway to ISDN, and will, when the standards are finalized, support 2B+D voice-data telephone sets. At present it is just getting started in the large PBX field, but it does not support the electronic sets used by Systems 25, 75 and 85. But it will obviously develop to do whatever is ultimately desired.

[ Top ] [ Next ] [ Table of Contents ]

Copyright 2005 Lee Goeller. All Rights Reserved.