Editor's note: Step by step switching is hard to visualize from a written description. And nearly as difficult when you are looking at a stepper switch working. The mechanism is so strange looking and its parts move so fast that it's like trying to understand a mechanical stop watch. Check out the external links below first it you want a better idea of what Crowsen is writing about. After I get this website redone, around the first of the year, I do intend to get back to this article to illustrate it. Use the search engine above to find out a great deal more about steppers at this website.

http://www.seg.co.uk/telecomm/automat1.htm

http://www.sigtel.com/tel_tech_sxs.html

Tom: I would like to respond first to the issue of tracing and not tracing phone calls.(internal link) My experience is from New Zealand, but the technology used is similar but not the same as in the USA. As a general rule, up to about 1990, a significant proportion of our telecom network used electromechanical switches, the same would probably apply in USA. The significant point of that is these switches had no memory -- an event stimulated a response from an electromagnetic device, this in turn stimulated another, and so on. The device had no ability to recall what it had done. So if a kidnapper rang a person and demanded a ransom and hung up and the call train was lost, there was almost no possibility of discovering the origin of the call. That's why in 1960's and 70's TV programs such as Hawaii 50 and the like the person rung was told to keep the kidnapper talking as long as they could. While the kidnapper was talking the switch technicians would run around like mad things shouting above the noise as they narrowed down on the source of the call. With the advent of computerised switches and signalling protocols it was possible to obtain the phone number of the caller before the person answered the call. This meant that instead of having to laboriously examine hundreds of electromagnetic devices to determine the origin of the call, it was possible to have the answer almost instantly.

Your understanding is correct about how Strowger works, but there are a few fine points. The British system, (the jargon I use here is British Post Office / New Zealand Post Office ) which was very similar to the ones used in USA, used a device called a Group Selector as the basic building block. They used a different name in the American systems, but the concepts are the same. Once dial tone had been obtained the customer would start dialing, the important part was the restore of the dial. This is when the "carriage" would "go vertical".

After the dial on the phone finished restoring the "carriage" would "go rotary". The "wipers" (the electrical contacts on the carriage) would move around the "bank". The top set of wipers were called the "P" (for privacy) wipers, they would check for an earth on each contact as they swept around the bank. The earth meant that circuit was busy. Upon finding an absence of an earth (I forget what we called that condition) the "selector" would "extend" the "loop" (this was the call from the group selector back to the customers phone) to the next group selector or final selector (the last stage in the call train). For about half a second the group selector would keep its own busy earth on the circuit while waiting for the next stage to kick in with their "holding earth". This holding earth is what we used to trace the call on.

As an added dimension, there were 20 circuits that could be chosen by each group selector on a "level" (this was the digit dialed), and each one was in a group with 19 others (making a total of 20). But nobody dials the same number every call, so these groups of 20 were concentrated with others to form quite large groups. For example, you may have 240 switches that accessed 100 switches in the next stage of dialing. Usually the group of 20 would have between one and about 4 contacts that were unique to the group, then they would have a similar number of pairs, quads, and so on until the last few of the 20 contacts were spread over the whole of the switch.

Thus, when tracing backwards, it was easy if that switch used an early choice contact, but it was awful if a late choice contact was used. How did we know where the call came from? It was recorded on the label - each group selector or Final selector had its exact name and place (Rack - shelf - position) and in red lettering was the source of the call. For example (my memory is a bit faded but it was of this nature) it might say "6A A5" then in red lettering "1st B A1 - Ist B D10 O/L 2" (O/L = outlet) This meant the switch name was Rack 6A meaning the caller had dialed a 6 to get there, it was the first rack in the 6's, it was located on shelf A (at the bottom of the rack), on position 5. To find where the call came from you had to go back to the "First Groups" (the first groups were the ones where you got dial tone from), find rack 1st B, look on shelves A through to D looking for a switch on level 6 with the "wipers" on the first set of contacts ... ah ha, you say, but the label said outlet 2, surely you mean the second set of contacts. Not quite, but if you did indeed say that it means you are very astute and haven't got lost yet.

If you remember, there are two wires going into your house from the phone company, the system Strowger designed only needed two wires inside the switch for speech, hence speech was bidirectional. The "P wire" was added inside the switch, it was also used for "metering" i.e. determining how much to bill the customer. The NZPO didn't use this feature because some politician said everyone would get free local calls if they voted for him, his name is lost to history, but he got elected. That's an aside, thus we have three wires, but a close examination of the bank would show 6 contacts, thus the group selector and final selector had access to two destinations in each rotary position - there are 11 rotary positions on each level of the bank, 10 for regular use, the 11th was called "overflow" which gave (surprise surprise) overflow tone. To find out whether it was the even or odd contact that was used the technician would test by placing a "link" into two points on the interface block (given a name that escapes me) and a bell would ring to indicate the even contact. Thus in the example above, if the tech found two switches on 1st B shelf B which were on level 6 and the first rotary position (i.e. it could be either outlet one or two) he would have test each to see which rang the bell. The one that did was the source of the call for 6A A5.

Regarding the noise, I have permanent hearing damage. I play guitar, so it can't be to bad though. I had a hearing test done about 10 years ago and the lady said that I had worked in a noisy environment. Small switches in the country side were worse because they used cheap building materials such as "jib board", a type of plaster / cardboard wall covering, which has excellent sound reflecting properties. Bigger step by step switches tended to be slightly quieter.

One of the interesting tasks we had to do was "traffic recording" which involved measuring how many calls were made each 30 seconds or minute. It used a series of small single dimension rotary switches we called "Uni's" or Uniselectors. They only rotated. A whole shelf or series of shelves would all drive in unison at about 5 steps a second, so the whole thing sounded like a group of drummers beating out a 5/5 beat.

The effects of technology limitations is more widespread than most people imagine. For example, the 56K modem that I use to connect to the ISP, has its origin way back when toll calls were frightfully expensive. They worked out that the maximum bandwidth needed for clear intelligible speech (for men, of course) was about 4KHz, allowing for cheap filter design, this meant an effective range of 300 Hz to 3.3 KHz. Not exactly FM!!! When PCM - that's the stuff they record CD's with (they borrowed all the technology from the telco industry) came along, they kept the same bandwidth, then, when digital switches came along, surprise surprise, they used the same bandwidth again.

The original Strowger system would have handled a much wider bandwidth (albeit, with much more noise interference) than a modern switch. But before you leap up and down, the "impact" noise on the Strowger switch was considerable (it would knock out fax calls!) - that's the source of the "tick tick tick tick" noise (say it quickly) on phone calls of that era. Also the maintenance costs were much higher. For example Telecom NZ kept the phone bill the same from 1986 through to 1990, this was also the period when we had the highest inflation. The bill stayed the same because that was when they were throwing out most of this step by step stuff for computerised technology.

Another effect of technology limitation is the toll call or long distance call. Local calls used to mean a call that was un-amplified, long distance meant an amplified call. Basically you could get about 20 or 30 miles without amplification, probably a bit more with a two wire bidirectional amp (true - it amplified sound in both directions without having to do a split! - and don't bother asking how, it was explained to me in great detail and I still don't know how it worked). To go beyond that required the call to be split and then modulated ( and of course demodulated - hence the name MODEM - I may be slightly wrong here - the name means what I said, but it may have been used with telex calls over long distance, hence the name because the modulator and demodulator were in the same box) and then modulated again and again and again into groups (12 calls) and supergroups (60 calls) and then broadband ... gads, where do they get these names from.

Today, the computerised switch splits the call the moment it arrives and it stays split until it gets to the called parties phone line. Amplification is not an issue, because of fibre optics that link even local switches. In NZ we don't use any amplification on fibre within a metropolis, only on long distance stuff, but its all digital anyway, so no amplification only "regeneration". And don't be tricked into thinking that just because you rang your neighbour your call didn't travel very far. In rural areas a call could easily travel across the toll boundary to the big city, be processed, and sent back, and the caller pays local rates and doesn't even no. And if they ring the city? Well, what do you expect to pay, half the local rate, get real, its toll rates if you please!

A third effect is the lack of portability of phone numbers - this is still the case in NZ, and probably in the USA as well. The low price of RAM or high speed information retrevial from Hard Disks (yet to come) is a modern phenomena and a more advance signalling protocol is required to fix that problem. Remember this: The basic computerised switch in NZ was designed before the IBM pc!!! True. In fact, I remember the day my boss and I found a fault on a "card" that did the quantising (that's where you turn speech into a digital signal), we pulled out the old, and it was packed with transistors, we pulled the new one out of a Chinese take away bag (static protection) and it had one big chip in the middle and almost nothing else. We checked the serial numbers - the same. We looked at the two boards - totally different - checked the numbers again - still the same - put the new one in - and it tested perfect! I suspect the first board was probably designed just after the first CD player while the second was a few years after CD Walkmans with all their IC's.

And your humble IBM power cord? It came from test instruments used in the telco industry. The decibel as well.

Next time you look at the pictures of Neil Armstrong on the moon remember this, that was the best technology that money could buy at the time. That required super cooled receivers in the receiving stations on earth.

So next time someone plugs in a kettle with the humble PC cord attached and bemoans their phone bill, just remember - somewhere in the ancient history of the last 30 years is a link!!! I hope this hasn't bored you.

And so what do I do now? I drive a bus. I don't talk about this at work because most of them never had the opportunity I had, and because I'd find it too upsetting.

Stephen Crowsen.

Pages: (1)_(2)_(3)_(4)_(5)_(6)_(7)_(8)_(9) (10) (11) (Communicating) (Soundwaves) Next page -->(Life at Western Electric) next page -->

Pages: (1)_(2)_(3)_(4)_(5)_(6)_(7)_(8)_(9) (10) (11) (Communicating) (Soundwaves) next page -->