In the mid 70s, my dad took me to one of these and demonstrated for me. He worked for Western Electric at the time, which AT&T eventually bought. But, it was interesting. I was lucky that dad shared his work with me. I knew he left in the morning and came home late evening and sometimes left for days. I always wondered why. When he explained this, he went backwards, showed lines, then home to show lines to the house and connected to the home rotary dialed phone. My daughter saw our old rotary dial, and she didn't know what it was.
Well, thank you. If you're interested in learning more details about SXS equipment, I have much longer videos that explain it all. Just search on my channel name "hicken65". Thanks for watching.
The release of tensioned springs and the force of gravity combine to create a speedy reversal of the Strowger two-motion switches in a simultaneous action that is both visually pleasing and audibly rewarding with the "kerchunk" finale (well, at least for those of us that like this stuff). Thanks for watching.
Hicken65, thank you for this wonderful description. In USAF, I was a central office tech 362x0. Stationed in Germany 1952-1954, I attended the Siemens u. Halske RP-40 school in Munich, which led to a lifetime career in electronics.
Well, your career certainly spanned an exciting and dynamic time period for the development of electronics. Glad you enjoyed the video and thanks for watching.
Started out in 1980 in New zealand on the BPO version of these "100 type" very similar except they had a uniselector as a line finder. Spent many hours adjusting and oiling switches! 24,000 line exchange, the noise was amazing, you could tell when a ad break came on the Tv from the increase in noise. The funniest were Rugby test matches, the place would go quiet and you could hear individual calls going through, like 3am but during the day. Amazing pieces of equipment, hopefully there is one preserved somewhere.
Yes, when working in a SXS CO you were certainly aware when something abnormal was happening. There are a number of preserved SXS CO's in North America in museums or private collections. If you expand my video description above, there is a link to a video of one of them.
What a wonderful video! Thanks for creating and sharing it. It brings back some nice memories as I have spent a decent amount of time in step-by-step offices. Those beasts were mechanical marvels and amazingly reliable considering the torture that they had to endure.
Thanks for this demo. I enjoy learning about this technology that was obsolete before I was born. I'll take some time and watch your other videos as well.
My father told me a story of when he was a kid in a city served by these switches be used a walkie talkie to coordinate with a fried to dial each other at the exact same time. This resulted in the phone making all sorts of strange noises. But after a few tries of this, a voice came on their line and shouted “will you kids stop playing with the phones!!”
Well, with one exception, this activity would not have caused any problem in the step-by-step switching equipment. Both calls would have progressed normally and after dialing the last digit, the two connectors would recognize that the the lines being called were in use and return busy tone to each caller. The exception was when the two callers shared the same party line. The two simultaneously active dials being a small fraction of a second out of synchronization would have resulted in non-conforming dial pulses and caused irregular activity at the first selector.
@@Hicken65 interesting - he described hearing noises like something was shorting out. Could a different type of switch cause this? They would time it so that the dial would send its last pulse at the exact moment. This had to be in the late 40s or early 50s. It must have caused some issue if someone at the central switch patched in and told them both to cut it out! lol
Well, two single party customers served by the same central office switching system and who called each other regularly would likely call each other at the same time once in a while. It would not cause a problem in the central office. Each call would reach a different connector and both connectors would recognize a busy line and return busy tone. The exact timing of the last digit to be dialed would have no relevance. Each line would appear busy at the connector as soon as the line received dial tone, before any digits were dialed at all. This was necessary so that anyone calling a line that was in the process of dialing an originating call would get busy tone and not disrupt the call in progress. So I have no idea what was happening in the scenario described to you by your father.
@@Hicken65 ok I asked my dad for more info - he said he toured the central office as kid (Stamford CT) - they used Strowger switches - he said when they did it, one would always get through to the other except when the voice came in. It may have been when the voice came on that they heard the strange noises. So perhaps it did not cause any actual issue but the repeated calls between the two parties caught the attention of someone working on the frame.
When I was in college in the early 1980’s in Springfield, MA, our local exchange 782 and 783 in the 413 area code used these steppers. There was no Touch Tone service offered and the clack on the line when dialing was very loud on a “modern” phone. I always recalled there was no clack for the last two digits. Just for fun, I hooked up an old model 300 phone and it had less clack sound. Guessing it was better adapted to the old stepper switches than the 500’s or Trimline phone I normally used.
Well, that's an interesting anecdote that I can't explain. When I plug 500 sets into my demo unit. they are much quieter than the 302's. In fact, I took varistors from the receivers in 500 sets and put them across the receivers in the 302's and it helped quite a bit to quieten things down. Thanks for watching.
Interesting, but I still can't figure it out. It's amazing telephone calls got completed for many decades this way, with all that electromechanical switching along the way.
The electromechanical telephone switching network was indeed an amazing achievement. If you're interested in figuring it out, you can watch the very long videos on my channel (Hicken65) that detail the operations of each of the three switch types. Thanks for watching.
PEnnsylvania 6-5ohohoh -A famous number by Glenn Miller and orch. It was the switchboard number of the Pennsylvania Hotel in NYC, across 8th avenue from Penn Station, where the band was appearing and staying.
Yes, Miller and many other artists (Dorsey Brothers, Woody Herman, Count Basie and Duke Ellington, to name a few) performed in the hotel's famous Café Rouge. You can still reach the hotel at PE 6-5000, but you'll have to prefix it with area code 212.
When I was a kid, the radio station would have contest, I would force the dial back rapidly to dial the number faster. After seeing how the system worked I am amazed at how well they were made back in the day...
The switches had some tolerance for imperfect dials that were pulsing over or under 10 pulses per second. If the pulse rate deviated too far, however, a call would generally not reach its intended destination.
@@dbeierl A number of switching systems around the world were capable of handling 20 PPS. The Bell System Step-by-Step that I'm demonstrating could not. Your 1968 Central Office was a different type.
Thanks. Of course most of this equipment was scrapped as it was being decommissioned in the 1980's and 1990's. Some of it was saved by collectors who have built demo units similar to mine. There are a few collectors with entire SXS central offices in their basements or sheds. There are also several museums with working displays of this equipment.
Ever since I was a little kid, I just wanted a video that said "Here it is, I'm gonna dial a number, here's what happens!" ... Well, 24 years of having an internet later, here it is.
Apparently so. I had another comment on one of my videos sometime ago from a person who had an old jukebox. He mentioned that it contained similar stepping switches. It would be nice if someone could video the inside of one in action and explain what was happening. Thanks for watching.
As a kid during the era when this type of system was in use, I wondered how the system knew if a dialed number was long distance or not. I made the assumption that the way it did that was based on the sequence of the first three numbers dialed. At that time, all Area Codes consisted of three numbers of which the 2nd digit was always either a 1 or a 0, and nobody had a telephone number with 1 or 0 as the second digit of the seven digit number. As an example, back then, in Connecticut (area code 203) you did not have to dial an area code for dialing a local number, 335-3396 in the Bridgeport area. If you wanted to call a number outside the local area, but still within Connecticut, you simply dialed 1+335-3396 for example, say from Hartford to Bpt. If you wanted to call outside Connecticut, the sequence was 1+212-555-1212 for example. So, by recognizing that the 3rd digit in the sequence was or wasn't a 1 or 0 determined if the call was headed outside the home area code. Is this correct?
Yes, you're correct.. The first digit "1" was the long distance access code. It was not considered to be part of the called telephone number. In a typical local step-by-step office, if the first digit received was a 1 and the second digit received was 2 to 9, you would be connected to a long distance trunk to a toll office.The full 7digit or 10 digit telephone number would be sent to the toll office (but not the access code"1").If the second digit of the received telephone number was 2 o 9, the toll equipment knew you were making a 7 digit intra-area call. If the second digit of the telephone number was 0 or 1, the toll equipment knew you were making a 10 digit out of area call. It made a difference in determining if the routing translation would be performed on the first 3 digits of the telephone number or the first 6 digits of the telephone number.
When working in a SXS office, you always knew when "something" was happening. There was a dramatic increase of clattering switches. It was time to turn on the radio to catch the breaking news.
So I've been watching and I've watched a few videos and I'm still confused. Does the phone not care what route it takes between the finder and the selector? So I get the line finder. And if I'm getting this right. All the numbers lead you to a specific bank to the line selector? Does that mean that The same line can be serviced by multiple selectors?
Each telephone line did not have its own dedicated switch or set of switches. To make efficient use of the switches, there was a great deal of shared access to them. For example, line finders were arranged in groups of up to 20 that had shared access from 200 lines and connectors were arranged in groups of up to 10 that shared access to the same 100 telephone numbers. Between the line finders and connectors were the ranks of selectors. Each line finder was tied directly to its own first selector. After that, the 2nd through 5th selectors were set up to be accessed on a shared basis from the preceding rank of selectors.Typically, SXS systems were provided with sufficient switches such that about 7% to 10% of the telephone lines could be in use simultaneously on originating calls and 7% to 10% could be in use simultaneously on incoming calls. A 50,000 line SXS office would have around 25,000 of these switches.
@@Hicken65 I get how it works now! Thank you so much for taking the time to give me a detailed explanation. You have no idea how much I appreciate this. I've been trying to figure out for years of these things work.
@@Rusty_Nickle And if you're really interested, I have much longer, very detailed videos on my channel. Just search on "Hicken65". Thanks for watching.
I remember seeing equipment like this In use at Linden st and N Silk st in Allentown pa back in the 90s when they had a door open in the windowless building! It was I believe Bell Atlantic now Verizon. I would love to see what they replaced all that equipment with? They even had live operators in that building back then for information, like time and weather or was a free call from many payphones! Today Verizon will never pay anyone to answer the phone in the USA!
The replacement was likely a Lucent 5ESS or a Nortel DMS100. As far as" love to see" it goes, it depends on whether you would like to stare at blue boxes or brown boxes. Trying to get a human being to answer on almost any business telecommunications system has unfortunately become a daunting task.
Thanks for posting this, it was fascinating. One point is unclear to me. When a digit is dialed, how does the relevant switch know when the last pulse for that digit has come arrived, and it can then seek an available circuit. I know the timing of the pulses is important. Does the switch simply have a timeout, and if another pulse doesn't come in time, that indicates that that digit is complete? Thanks!
The slow release C relay monitors for the end of the dial pulse train. When the A relay has re-operated and remains operated after the end of the last pulse, the C relay releases. The acceptable release time range is from 111 to 140 milliseconds. This closes the energize path for the rotary magnet. Operation of the rotary armature rotates the shaft and wipers to the first terminal of the bank on the selected level. C remains released unless this selector processes another digit in a digit absorbing application.
When the shaft and wipers begin to move vertically upwards, you'll see a horizontally mounted copper coloured wiper brushing against a set of vertically mounted copper contacts. There is one contact for each level of the bank. The operated line relay causes a ground potential to be placed on the contact associated with its level. The shaft continues its upwards motion until its horizontal wiper detects the ground, at which time it stops vertical motion and begins rotary hunting. The top set of contacts on the bank assembly are the sleeve contacts, which provide the service request status of each line on the bank. If a line is requesting service, the operated line relay will cause -48 volt potential to be placed on its sleeve contact on the bank. The rotating sleeve wiper looks for the -48 volt battery and stops rotating when it sees it. At this point, the line requesting service has been found. If you are interested in more detail, you can watch my very long video on this subject entitled "The Line Finder In The Step-By-Step Telephone Switching System". Just search You Tube for my channel name, Hicken65. Thanks for watching.
Hicken65 ... Thanks for the explanation. It is a really complicated system! It took me ages to spot the circular piece of metal with the finger out to the right parsing the 10 tiny flat contacts. I had already started to watch the other video you mentioned, but had to abandon it to get on with other stuff. It is bookmarked for when I can devote time to watch it all without interruption.
How did switching offices change when touchpad dialing was introduced? Also, why did the first through fourth selectors have multiple contacts on each row? was this so that any selector could be connected with up to 10 in the next position, thus almost "multitasking"m
Other switching systems, such as crossbar and ESS could handle touch tone. The SXS could not. However the telephone companies soon inserted an electronic device between the line finder and the 1st selector that could receive the tones and translate them into rotary dial type pulses that they would send to the 1st selector. The subscriber would have the convenience of touch tone dialing, but there was a noticeable delay before the distant phone would ring.
For details on the selector contacts, see my long video The Selector In the Step-By-Step Telephone Switching System Part 1. Just search my channel name, Hicken65, on You Tube.
In your examples, your step by step units are side by side to make one call from a telephone to another. In a central office, would this be the case, or would my dialing end up traveling between different areas of the CO ? For instance, after I dial first digit as 5, would the connection be then made to a step by step that is dedicated to handle the second digit following a 5., and if I then dial 2, would be connected to a step by step that handles the 3rd digitl for numbers that begin with 52 ? And how did that work when I dialed from one CO to another ? if I dial "52" and this belongs to a CO halfway across town, does this mean that after racteching up twice, it would hunt for a free line to the other CO, after which, the dialing pulses would be transmitted to the other CO? Upon receiving that connection, how would the other COP know that the first digits dialed were "52" ? Were there separate physical lines for each dialing prefix ? (if this is all explained in another viddeo, just give me a link or video title).
In large offices, there were dedicated bays (320 selectors per bay) for each rank (1sts, 2nds, 3rds, etc.) of selectors. So yes, your call would travel all around the CO. For calls to other CO’s, they could exit the initial CO on the first, second or third digit. The outgoing paths were called trunks. If you were calling 529-1000, which was in another CO and the 52 pulses were handled in the initial CO by a first and second selector, the remaining pulses (9-1000) would travel over the seized trunk and arrive in the distant CO on an incoming 3rd selector. Only calls that began with 52 would reach the third selectors in the distant CO, so it didn’t have to sort out what was the first 2 digits were. The 3rd selector would process the digit 9. A 4th selector would be seized and process the digit 1. A 5th selector would be seized and process the digit 0. And finally a connector would be seized and process the digits 00. Generally, switches called “repeaters” were used on the trunks to freshen up the pulses sent out by the subscribers dials. If you are really interested in this subject, I have very detailed videos describing each switch type. Just search TH-cam by entering my channel name, Hicken65. Below is a link to a video (not mine) inside an actual in-service small SXS central office, that gives some idea of what the equipment looked and sounded like. th-cam.com/video/InN6G2RnaCw/w-d-xo.html
Thanks for answer. When COs started to have multiple NNX, did they have to then build trunks for each NNX supported by the destination CO ? Or was a totally different method of telling the destination CO about the digits already dialed by originating telephone ? @@Hicken65
@@jfmezei Generally, separate trunk groups were not required. Most multiple NNX CO's had the same first 2 digits. In my previous example, if the distant CO had three NNX codes, they would likely have been something like 529, 528 and 527. The incoming 3rd selectors would receive 7, 8, or 9 and the calls would be directed to their correct destination.
@@Hicken65 I watched your video from start to finish, so interesting! The machinery looks incredibly delicate and those noises remind me somewhat of a typewriter crossed with a modem dail-up. Thank you for that link watched almost the whole thing. What a sight to behold! My digital programmer brain is having trouble fathoming all those unique analog connection paths, connected by all those physical switchboards among millions of relays, connectors, wipers, contactors and what must have been hundreds of kilometers of copper wire in that building alone. Not to mention the part count for all of those machines, it had to be in the trillions (or maybe millions?). Anyway, dazzling stuff. Analog excellence, what a rush. Many, many people had to have gotten deaf in that room! XD I mean this link BTW: th-cam.com/video/InN6G2RnaCw/w-d-xo.html It truly is an epic watch, and it needs a boatload more viewers! Maybe you could (if the uploader permits it) download it, edit it to be more concise and structured and then voice over and explain exactly what we're hearing/looking at, what its purpose is and what it is doing at the moment. I would love a video like that but I lack the knowledge about these old telephone systems. But, I am handy with computers so if you'd like and/or need my help downloading the video from TH-cam (in decent quality) and converting it into a format that is accepted by the editing software of your choice just drop me a message :D and if needed I could probably also help you with recording audio. Please let me know, it would be superb!
Sorry, I don't relish the thought of attempting to edit that video. However, for a full explanation of how the line finders, selectors and connectors work, you need to watch my long videos on these switches. Just search my channel name, Hicken65, on You Tube. Thanks for watching.
Well, the short answer is I don't know. To build a demonstration unit similar to mine, here's what you would need: 2 line circuits, each consisting of one line relay and one cutoff relay; one line finder switch and its jackspring and bank assemby; five selectors with their five jacksprings and five bank assemblies; one connector and its jackspring and bank assembly; a 48 Volt dc power supply, a ringing generator, a dial tone generator; an interrupter for the ringing; an interrupter for for busy tone; wire to interconnect everything; and mounting framework for all the equipment which you can build using angle iron. There is no single supplier, that I am aware of that is selling this stuff. Occasionally, bits and pieces show up on Ebay and occasionally, some collectors list parts for sale in collector forums or newsletters. Telephone collectors hold several shows each year which often include switch demo units. You need to research such organizations as Telephone Collectors International and Antique Telephone Collectors Association.
I understand that the selectors are in the same frame for demonstration purposes in the video, but in the real world wouldn't the selectors for each digit be located in various parts of the CO with the last connecting selector just serving the last nnn-nnXX digits? How would dialing the first two digits select a proper trunk to go to another CO?
Since the exchanges in an dialing area (area code) may be many, I'm curious how and when the outgoing trunk to another CO is selected -- I assume after the first two digits of an exchange are dialed then the appropriate trunk is selected to then connect the call to the appropriate CO for the last 5 digits to be decoded. That implies that there are trunks from each CO to all other COs in the area code, or was there another CO inter-switching method which was more efficient?
What does your demo setup use as sources for ringing voltage and dialtone? Does it use the standard motor-driven interrupters to make the busy signals?
The ringing generator is a Tellabs model 8102 made in 1998. It is a solid state generator that outputs 105 volts ac at 20Hz with a series -48 volts dc component. The dial tone/busy tone generator is a Lorain model F1 made in 1966. It is a solid state generator that outputs 600 Hz amplitude modulated by 120 Hz. The busy tone and ringing tone interrupters are both made by the Pylon Electronic Development company (year of manufacture and model unknown). They both have electric motors with brush and cam arrangements.
Well the part I didn't quite get, was how the system could handle more than 1 call at a time, and how there was an upper limit to the number of phones that could be called. I take it, this was a simplified documentary
It is indeed simplified. My demonstration unit was designed only to show how a call progressed through a SXS switching system. When a switch is used on a call it is tied up to that call for the entire duration of that call and so cannot be used to process any other calls. When a call is established through my demo unit, all 7 switches are tied up and so only one call can be established at any given time. In a very large central office, each established intraoffice call would also tie up 7 switches, but there were many thousands of these switches. Each telephone line did not have its own dedicated switch or set of switches. To make efficient use of the switches, there was a great deal of shared access to them. For example, line finders were arranged in groups of up to 20 that had shared access from 200 lines and connectors were arranged in groups of up to 10 that shared access to the same 100 telephone numbers. Between the line finders and connectors were the ranks of selectors. Each line finder was tied directly to its own first selector. After that, the 2nd through 5th selectors were set up to be accessed on a shared basis from the preceding rank of selectors.Typically, SXS systems were provided with sufficient switches such that about 7% to 10% of the telephone lines could be in use simultaneously on originating calls and 7% to 10% could be in use simultaneously on incoming calls. A 50,000 line SXS office would have around 25,000 of these switches.
Thanks for showing what you could, but the design of these switches still has lots of unexplained details, not least of which is that the system also does automatic billing.
There is a lot more detail in my longer videos. Just search on my channel name. Unfortunately I do not have any Automatic Number Identification or Automatic Message Accounting equipment available to make a video. Thanks for watching.
Yes, the wipers always hunt from left to right and always stop on the first idle trunk found. In my demo unit, I have artificially made the first 9 trunks appear to be in use in order to enhance the experience of watching the motions of the wipers.
I have another video, "Dialing Through The Step-By-Step Telephone Switching System" that describes the progress of a call to Beechwood 4-5789. Thanks for watching.
Is all that apparatus for every single line? I don't understand how it handles calls to consecutives numbers, by example, if phone lines xxx50 up to xxx59 can be use at the same time all the ten for receiving calls all at the same time? It works with parallelism of those mechanism?
My demonstration unit just has one switch of each type to show how a call progresses through a Step-By -Step system.In a central office, there is a lot of shared access utilized by parallel multipling of bank wiring so that multiple switches can share access to the same 100 or 200 circuits. For instance, in a typical setup, 200 lines share access to a set of 20 line finders and 100 telephone numbers share access from 10 connectors. In a 50,000 line office, there would be about 24,000 switches of all types, so just under half a switch per line.
wow, you mean that of 200 lines only 20 can do a call at the same time? that is like the 10% of sinmultaneosly usage!. But I imagine those 200 lines can receive a call regardless of all call lines are used.
Yes, that is correct, only 10% of lines could have simultaneous originating calls in progress and 10 % of lines could have simultaneous incoming calls in progress. This was typically sufficient. Of course, in serving areas that were business dominated, less sharing was done in order to handle the loads and still provide mandated service levels. The originating equipment did not interfere with the terminating equipment. If 20 of the lines in a group of 200 lines sharing access to 20 line finders were in use on originating calls, the remaining 180 lines were free to receive incoming calls.
Yes, there are several ranks of selectors. One of each type is in my demo unit.You would get slow busy tone, 60 interruptions per minute (IPM) if you dialed an in use number. You would get fast busy tone (120 IPM) if there were no available trunks anywhere between digits along the dial path. If there were no line finders available, you didn't receive any tone. You just waited until one became idle and then you received dial tone.
A couple of questions: Say my phone is 46789 and I wish to dial 34567 To reach me, are there 100 line connectors whose 8th row and 9th position output connects to my phone line? Conceptually, if my number begins with 4, does this mean that my line connector would be at end of aisle 4 in the CO? And if I pickup the phone, do I have my own line selector attached to my line (always the sime that hunts for a free trunk) ? Would the first line selector be next to it, but when I dial the first digit 3, I would be connected to the next line selector in aisle 3 that serves numbers that start with 3 ? (and from that point, would connect to selectors and final connector on aisle 3) ? Just trying to understand on a larger scale how it was all arranged as opposed to the demos with just 2 phones.
Connectors were arranged in groups in which all the connectors in a group had access to the same 100 telephone numbers. For residential customers, a connector group would consist of up to 10 connectors. In your example , the connector group for your telephone would serve telephone numbers 46700 to 46799, and yes, your telephone would be on the 9th rotary position of the eighth level on the connector bank. In a very large SXS system, say 50,000 lines, there would be somewhere in the neighbourhood of 24,000 switches of all types. These were mounted on equipment bays that were 11' 6'' high, and as you can imagine this took up a tremendous amount of floor space. In large offices, switches were mounted on dedicated bays for each switch type. That is, there were dedicated bays for line finders, for each selector rank (i.e. 1sts, 2nds, 3rds, 4ths and5ths) and for connector groups. Any given 7 digit telephone call would tie up 7 switches spread over 7 different bays that could be located anywhere in the central office. The only bays directly tied to telephone numbers were the connector bays, which typically contained 7 connector groups and thus 700 telephone numbers. The line finders were arranged in groups of up to 20 line finders that served 200 lines on a full sharing basis. That is any of the 200 lines could be connected to any of the 20 line finders in a group. When you originated a call, one line finder in your group would be activated, find your line and connect it to a first selector, which would give you dial tone as a signal that dialing could commence. If you dialed 34567, your call would tie up a particular line finder, 1st selector, 3rd selector, 4th selector 5th selector and connector. If you later made another call to that same number, your second call could be on a completely different set of switches that routed your call on a different physical path through the central office (although it would begin in the same line finder group and end in the same connector group). For a much better understanding of this topic, you can watch my very long, detailed videos on the linefinder, the selector and the connector . Just search my channel name, Hicken65, on You Tube.
@@Hicken65 Thansk for the explanation. But still confused on 200 telephones sharing 20 line finders. When I lift the phone, the videos I have seen show the line finder being the first item. But if any of the 200 lines can any up on any of the 20 line finders, what is the name of the device that get me connected to any one of the 20 line finders? Also would it be correct to state that if 20 people of the 200 in my group are already on the phone, when I pick it up, I would get no dialtone? Would I need to hang up and pick-up again to try to get a free one, or could I stay off-hook and wait until I got dial tone?
@@jfmezei Each telephone line has its own line circuit, which consists of two relays, the line relay and the cutoff relay. When you pickup the the telephone , the line relay operates over the electrical loop through your telephone set. Each line finder group has a group control circuit that contains the preference circuitry to select an idle linefinder. The operated line relay initiates a request to this circuit to begin the selection process. If all the linefinders are in use, waiting calls queue up and will get served as line finders become idle. So if you stay off-hook you will eventually get dial tone. Details on this are provided in my video "The Line Finder In The Step-By-Step Telephone Switching System" which you can find on my You Tube channel, Hicken65.
@@Hicken65 Thanks. I had seen the video where you have a phone, line finder, selctors and connectors, and the phone appeared to be directly connected to the line finder. But your response shows there was something before that which connected my phone to a the first availabel line finder.
How does the system keep track of what number is being dialed? The only way I can think to achieve this is with incredible, costly, "fan-out": at every level, there's a connector (or bank of them) to handle each dedicated collection of digits. But this would mean that for every digit added, you'd need 10* more circuits than the last, and in your typical 7 digit number, that would mean 1,000,001 connector circuits (10^5 connectors, the final connector that can handle two digits, and one connector at the start that doesn't need any memory). That seems like an awful lot of circuitry. Surely that can't be the way the exchange kept track of the number dialed at every stage.
While my video shows how a call flows through each of the 7 switch types, it does not explain any of the concepts of concentration, trunk distribution, bank multipling, trunk sharing or selector shelf subgrouping arrangements. In an actual central office there are hundreds or thousands of each switch type and a typical call is routed all around the central office instead of being confined to the small space of the demo unit. Let’s look at a theoretical scenario in which a SXS system has 10,000 lines and 10,000 station numbers. For simplicity, we will assume all traffic is intra switch, which means all calls originate from these lines and terminate to these numbers. Users only need to dial the 4 digit station numbers (i.e. 0000 to 9999). Assume a maximum of 10 % of the lines/numbers are in use at any given time in the busy hour (this is typical). I will need 1000 1st selectors to handle the originating traffic (i.e. 10% of 10,000). Since only 4 digits are being dialed, we will not need any 2nd, 3rd or 4th selectors. All calls will route from 1sts to 5ths to connectors. The 1st selectors will process the thousands digits. Each of the 10 levels of the 1st selectors will represent a particular thousands digit of the station number. Each level will route the call to its own rank of 5th selectors. Assume the traffic is evenly dispersed across the 10 levels (i.e. 10% per level). We would provide 100 5th selectors for each level. That would give us 1000 5th selectors to handle the traffic from the 1000 1st selectors. Let’s study a call to station number 0987. The first digit dialed will be “0”, which will utilize level 0 of the first selectors. Any given 1st selector can only access 10 trunks on level 0. Selectors are arranged in groups of 10 on a shelf. The 10 banks associated with that shelf of 10 selectors are all multipled together, so that the 100 trunks on one bank (10 levels x10 trunks per level) are multipled to its adjacent bank, and the next bank and so on, such that there are only 100 trunks that can be accessed by the shelf of 10 selectors (10 trunks per level). So we now know that on level 0, 10 selectors are sharing access to the same 10 trunks to a rank of 5th selectors. But we have 1000 1st selectors spread over 100 shelves. We know that 10% of the traffic from the 1000 1st selectors will be trunked on level 0, so we will need 100 5thselectors to handle that traffic. Since any given 1st selector shelf can only access 10 of these 5th selectors (associated with level 0), it is necessary to divide these one hundred 5th selector trunks into subgroups of 10 trunks each. This would give us 10 subgroups of 10 trunks to the 5th selectors from level 0 of the 1sts. We have 100 1st selector shelves. These also must be evenly divided into subgroups, with each subgroup of shelves having access to one of the trunk subgroups on level 0. This will give us 10 subgroups of 10 selector shelves, with each subgroup having access to one subgroup of 10 trunks to the level zero 5th selectors. This is accomplished by simply multipling the 10 trunks of any subgroup across the 10 circuit outputs of all 10 shelves of the selector subgroup, so that all 10 shelves (100 1st selectors) would be sharing access to the same set of 10 trunks to the 5ths. This set up is repeated for thousands digits 1 to 9 on the other 1st selector levels. This gives us 10 ranks of 100 5th selectors each, with each rank associated with a different thousand digit. The 5th selectors will process the hundreds digits of the station numbers. The 5th selectors are mounted on shelves of 10 with banks multipled in exactly the same manner as the 1st selectors. Let’s continue with our call to station number 0987. We are going to utilize level 9 of the rank of 5th selectors associated with thousands digit 0. We have 100 5th selectors spread over 10 shelves. Assume the traffic is evenly distributed across the 10 levels. Therefore, each level will require 10 trunks to connectors. Since the number of trunks for level 9 does not exceed 10, there is no requirement to create more than one subgroup. The 10 trunk outputs on level 9 of each of the 10 shelves of 5th selectors will simply be multipled together, to provide one subgroup of 100 5th selectors accessing 10 trunks on level 9. The 10 connectors associated with these trunks will serve station numbers 0900 to 0999. On our call, the tens digit 8 and the units digit 7 will be processed by the connector. This set up will be repeated for the other 9 levels of this particular rank of thousands digit zero 5th selectors. This will provide 10 groups of 10 connectors, with each group serving 100 station numbers. The whole process described above will be repeated 9 more times for levels 1 through 9 of the 1st selectors. For each 1st selector there is 1 line finder. So overall, this office would have 1000 line finders, 1000 1st selectors, 1000 5th selectors and 1000 connectors. I hope this helps your understanding of the SXS system. Thanks for watching.
I don't recall any time out feature for partial dials, so the switch train would just be held up until the caller hung up. There was optional equipment available for releasing the equipment if no digits were dialed, or if a caller didn't hang up after a call was finished.
ok thanks for the replies. So does one person's home connect to the exact same line finder every time? And same with the connector switch.. each house phone line is wired to one connector switch in the office?
Well, the famous Hotel Pennsylvania is located right across the street from Penn Station, in the Manhattan borough of New York City. Since the panel switching system was the Bell System’s “Big City Solution” to customer dialed calls, I would think you’re probably correct. However, my demonstration unit is not trying to convey an historically accurate configuration for that location. I’m just trying to show how a call progressed through a Step-By-Step system using two well known (in their day) telephone numbers to enhance the demonstration.
Did every telephone have one of these giant contraptions dedicated to it at the phone company's office, or did multiple telephones share these switches?
@@Hicken65 I remember watching an episode of The Six Million Dollar Man back in the 70's. He needed to "trace" a call and someone was like "that's not possible! No one can trace a call that fast!" But Steve Austin went into the room with hundreds of these switches and traced the call. At the time I had no idea what I was looking at.
The trunks are just 3-wire (tip, ring and sleeve) cables between each rank of selectors.These cables run from the bank terminals of one rank of selectors to the input side of the next rank of selectors.The path through a selector is from its input side, through operated relay contacts inside the selector, through the wiper cords, through the wipers, through the bank terminals on which the wipers have come to rest, through the trunk cable connected to those bank terminals and on to the input side of the selector to which the far end of the trunk cable is connected. This type of connection path builds up as each digit is dialed, so that after all 7 digits there is a continuous 3-wire path from the line finder to the connector switch.
In the mid 70s, my dad took me to one of these and demonstrated for me. He worked for Western Electric at the time, which AT&T eventually bought. But, it was interesting. I was lucky that dad shared his work with me. I knew he left in the morning and came home late evening and sometimes left for days. I always wondered why. When he explained this, he went backwards, showed lines, then home to show lines to the house and connected to the home rotary dialed phone. My daughter saw our old rotary dial, and she didn't know what it was.
What a great demonstration. You just answered so many questions I had about how it worked. Really great video!
Well, thank you. If you're interested in learning more details about SXS equipment, I have much longer videos that explain it all. Just search on my channel name "hicken65". Thanks for watching.
I waited anxiously for the reset at the end. Cool stuff.
The release of tensioned springs and the force of gravity combine to create a speedy reversal of the Strowger two-motion switches in a simultaneous action that is both visually pleasing and audibly rewarding with the "kerchunk" finale (well, at least for those of us that like this stuff). Thanks for watching.
Thanks, I enjoyed your well-produced video and your article in Switchers' Quarterly that brought me here!
Hicken65, thank you for this wonderful description. In USAF, I was a central office tech 362x0. Stationed in Germany 1952-1954, I attended the Siemens u. Halske RP-40 school in Munich, which led to a lifetime career in electronics.
Well, your career certainly spanned an exciting and dynamic time period for the development of electronics. Glad you enjoyed the video and thanks for watching.
Almost same thing with me, my start was at russian ATC-54M (which is Siemens & Halske copy)
Fascinating!!
I have stayed at that hotel twice in my life.
Started out in 1980 in New zealand on the BPO version of these "100 type" very similar except they had a uniselector as a line finder. Spent many hours adjusting and oiling switches! 24,000 line exchange, the noise was amazing, you could tell when a ad break came on the Tv from the increase in noise. The funniest were Rugby test matches, the place would go quiet and you could hear individual calls going through, like 3am but during the day. Amazing pieces of equipment, hopefully there is one preserved somewhere.
Yes, when working in a SXS CO you were certainly aware when something abnormal was happening. There are a number of preserved SXS CO's in North America in museums or private collections. If you expand my video description above, there is a link to a video of one of them.
What a wonderful video! Thanks for creating and sharing it. It brings back some nice memories as I have spent a decent amount of time in step-by-step offices. Those beasts were mechanical marvels and amazingly reliable considering the torture that they had to endure.
Wonderful! The Marvelettes call Glenn Miller.
Two of the most well known numbers from the era when telephone exchanges had names. Thanks for watching.
@@Hicken65 that changed in 1966
As a kid I got a tour of a small town central office. The only thing that I remember is the constant racket from the stepper swiches.
It could get quite loud in a central office. To some it was just noise. To me it was a symphony.
Central office, was where those 1st 2 letters came from?
Hollywood always used KL5☎️
LATER..555
Thanks for this demo. I enjoy learning about this technology that was obsolete before I was born. I'll take some time and watch your other videos as well.
Ed Sharpe
WELL DONE~
wish we had at SMECC a demo unit like that SxS demo setup you have! This is a wonderful tutorial! KUDOS!
Well, thank you for watching.
Thanks for this great video 👍👍👍 My first job in telecom was at SXS station
It seems that most people that worked in a SXS office have fond memories. Thanks for watching.
Rotary dial phones make sense now.
Glad I could help with your understanding. Thanks for watching.
My father told me a story of when he was a kid in a city served by these switches be used a walkie talkie to coordinate with a fried to dial each other at the exact same time. This resulted in the phone making all sorts of strange noises. But after a few tries of this, a voice came on their line and shouted “will you kids stop playing with the phones!!”
Well, with one exception, this activity would not have caused any problem in the step-by-step switching equipment. Both calls would have progressed normally and after dialing the last digit, the two connectors would recognize that the the lines being called were in use and return busy tone to each caller. The exception was when the two callers shared the same party line. The two simultaneously active dials being a small fraction of a second out of synchronization would have resulted in non-conforming dial pulses and caused irregular activity at the first selector.
@@Hicken65 interesting - he described hearing noises like something was shorting out. Could a different type of switch cause this? They would time it so that the dial would send its last pulse at the exact moment. This had to be in the late 40s or early 50s. It must have caused some issue if someone at the central switch patched in and told them both to cut it out! lol
Well, two single party customers served by the same central office switching system and who called each other regularly would likely call each other at the same time once in a while. It would not cause a problem in the central office. Each call would reach a different connector and both connectors would recognize a busy line and return busy tone. The exact timing of the last digit to be dialed would have no relevance. Each line would appear busy at the connector as soon as the line received dial tone, before any digits were dialed at all. This was necessary so that anyone calling a line that was in the process of dialing an originating call would get busy tone and not disrupt the call in progress. So I have no idea what was happening in the scenario described to you by your father.
@@Hicken65 ok I asked my dad for more info - he said he toured the central office as kid (Stamford CT) - they used Strowger switches - he said when they did it, one would always get through to the other except when the voice came in. It may have been when the voice came on that they heard the strange noises. So perhaps it did not cause any actual issue but the repeated calls between the two parties caught the attention of someone working on the frame.
@@Hicken65 that makes sense to me - perhaps his memory is fuzzy it was like 60+ years ago.
When I was in college in the early 1980’s in Springfield, MA, our local exchange 782 and 783 in the 413 area code used these steppers. There was no Touch Tone service offered and the clack on the line when dialing was very loud on a “modern” phone. I always recalled there was no clack for the last two digits. Just for fun, I hooked up an old model 300 phone and it had less clack sound. Guessing it was better adapted to the old stepper switches than the 500’s or Trimline phone I normally used.
Well, that's an interesting anecdote that I can't explain. When I plug 500 sets into my demo unit. they are much quieter than the 302's. In fact, I took varistors from the receivers in 500 sets and put them across the receivers in the 302's and it helped quite a bit to quieten things down. Thanks for watching.
Interesting, but I still can't figure it out. It's amazing telephone calls got completed for many decades this way, with all that electromechanical switching along the way.
The electromechanical telephone switching network was indeed an amazing achievement. If you're interested in figuring it out, you can watch the very long videos on my channel (Hicken65) that detail the operations of each of the three switch types. Thanks for watching.
@@Hicken65 Thanks.
Just fantastic and truly educational!! Bravo!
thanks so much!! My old man "Walter Pantle was the man in Pensacola Fl . on SXS.. Thank You for sharing!!!!!
You're welcome. Thanks for watching.
PEnnsylvania 6-5ohohoh -A famous number by Glenn Miller and orch. It was the switchboard number of the Pennsylvania Hotel in NYC, across 8th avenue from Penn Station, where the band was appearing and staying.
Yes, Miller and many other artists (Dorsey Brothers, Woody Herman, Count Basie and Duke Ellington, to name a few) performed in the hotel's famous Café Rouge. You can still reach the hotel at PE 6-5000, but you'll have to prefix it with area code 212.
Excellent explanation!!! Thank you!!!
You're welcome, and thanks for watching.
Great video, just what I needed. Thank you.
Well, glad I could be of assistance and thanks for watching.
When I was a kid, the radio station would have contest, I would force the dial back rapidly to dial the number faster.
After seeing how the system worked I am amazed at how well they were made back in the day...
The switches had some tolerance for imperfect dials that were pulsing over or under 10 pulses per second. If the pulse rate deviated too far, however, a call would generally not reach its intended destination.
@@Hicken65 I had an adjustable dial that could do 20 pps. I used it with complete success in Annapolis MD in 1968.
@@dbeierl A number of switching systems around the world were capable of handling 20 PPS. The Bell System Step-by-Step that I'm demonstrating could not. Your 1968 Central Office was a different type.
@@Hicken65 thank you.
Yes, the timeout worked by a slow release relay that would determine that the last pulse was received in a string.
Wonderful. That equipment must be rare as hens teeth.
Thanks. Of course most of this equipment was scrapped as it was being decommissioned in the 1980's and 1990's. Some of it was saved by collectors who have built demo units similar to mine. There are a few collectors with entire SXS central offices in their basements or sheds. There are also several museums with working displays of this equipment.
Hicken65 that’s like my stash of Amateur radio tube equipment. What engineering can accomplish!
Good video. Nice clear explanation.
Ever since I was a little kid, I just wanted a video that said "Here it is, I'm gonna dial a number, here's what happens!" ... Well, 24 years of having an internet later, here it is.
Thanks. It's always nice to hear from a satisfied viewer.
I like the song you use at the end, it reminds me of I think an Australian song 8675309 by Tommy Tutone.
I use an excerpt from the Tommy Tutone song in my video Putzing With The Payphone.
@@Hicken65 Actually I did find that later on, after my comment on the Payphone vid.
This was awesomr. Thx for posting.
You're welcome and thanks for watching.
It’s like a complicated jukebox.
Apparently so. I had another comment on one of my videos sometime ago from a person who had an old jukebox. He mentioned that it contained similar stepping switches. It would be nice if someone could video the inside of one in action and explain what was happening. Thanks for watching.
@@Hicken65 it was fascinating to watch and videos like yours are the best part of TH-cam.
As a kid during the era when this type of system was in use, I wondered how the system knew if a dialed number was long distance or not. I made the assumption that the way it did that was based on the sequence of the first three numbers dialed. At that time, all Area Codes consisted of three numbers of which the 2nd digit was always either a 1 or a 0, and nobody had a telephone number with 1 or 0 as the second digit of the seven digit number. As an example, back then, in Connecticut (area code 203) you did not have to dial an area code for dialing a local number, 335-3396 in the Bridgeport area. If you wanted to call a number outside the local area, but still within Connecticut, you simply dialed 1+335-3396 for example, say from Hartford to Bpt. If you wanted to call outside Connecticut, the sequence was 1+212-555-1212 for example. So, by recognizing that the 3rd digit in the sequence was or wasn't a 1 or 0 determined if the call was headed outside the home area code. Is this correct?
Yes, you're correct.. The first digit "1" was the long distance access code. It was not considered to be part of the called telephone number. In a typical local step-by-step office, if the first digit received was a 1 and the second digit received was 2 to 9, you would be connected to a long distance trunk to a toll office.The full 7digit or 10 digit telephone number would be sent to the toll office (but not the access code"1").If the second digit of the received telephone number was 2 o 9, the toll equipment knew you were making a 7 digit intra-area call. If the second digit of the telephone number was 0 or 1, the toll equipment knew you were making a 10 digit out of area call. It made a difference in determining if the routing translation would be performed on the first 3 digits of the telephone number or the first 6 digits of the telephone number.
Pensylvania Six-five-oh-oh-oh. Well, someone was going to do it.
Were you ever in a SXS office when a "shower" occurred? One would think the office was crashing down on them.
When working in a SXS office, you always knew when "something" was happening. There was a dramatic increase of clattering switches. It was time to turn on the radio to catch the breaking news.
So I've been watching and I've watched a few videos and I'm still confused. Does the phone not care what route it takes between the finder and the selector?
So I get the line finder. And if I'm getting this right. All the numbers lead you to a specific bank to the line selector? Does that mean that The same line can be serviced by multiple selectors?
Each telephone line did not have its own dedicated switch or set of switches. To make efficient use of the switches, there was a great deal of shared access to them. For example, line finders were arranged in groups of up to 20 that had shared access from 200 lines and connectors were arranged in groups of up to 10 that shared access to the same 100 telephone numbers. Between the line finders and connectors were the ranks of selectors. Each line finder was tied directly to its own first selector. After that, the 2nd through 5th selectors were set up to be accessed on a shared basis from the preceding rank of selectors.Typically, SXS systems were provided with sufficient switches such that about 7% to 10% of the telephone lines could be in use simultaneously on originating calls and 7% to 10% could be in use simultaneously on incoming calls. A 50,000 line SXS office would have around 25,000 of these switches.
@@Hicken65 I get how it works now! Thank you so much for taking the time to give me a detailed explanation. You have no idea how much I appreciate this. I've been trying to figure out for years of these things work.
@@Rusty_Nickle And if you're really interested, I have much longer, very detailed videos on my channel. Just search on "Hicken65". Thanks for watching.
How cool!!
Glad you liked it and thanks for watching.
I remember seeing equipment like this In use at Linden st and N Silk st in Allentown pa back in the 90s when they had a door open in the windowless building! It was I believe Bell Atlantic now Verizon. I would love to see what they replaced all that equipment with? They even had live operators in that building back then for information, like time and weather or was a free call from many payphones! Today Verizon will never pay anyone to answer the phone in the USA!
The replacement was likely a Lucent 5ESS or a Nortel DMS100. As far as" love to see" it goes, it depends on whether you would like to stare at blue boxes or brown boxes. Trying to get a human being to answer on almost any business telecommunications system has unfortunately become a daunting task.
And now it's controlled by computers what a time to be alive.
AI will soon take Over🤔😵
Thanks for posting this, it was fascinating. One point is unclear to me. When a digit is dialed, how does the relevant switch know when the last pulse for that digit has come arrived, and it can then seek an available circuit. I know the timing of the pulses is important. Does the switch simply have a timeout, and if another pulse doesn't come in time, that indicates that that digit is complete? Thanks!
Thanks! As I suspected, timing is everything. I'll watch the video you reference.
I have the same question but no answer.
The slow release C relay monitors for the end of the dial pulse train. When the A relay has re-operated and remains operated after the end of the last pulse, the C relay releases. The acceptable release time range is from 111 to 140 milliseconds. This closes the energize path for the rotary magnet. Operation of the rotary armature rotates the shaft and wipers to the first terminal of the bank on the selected level. C remains released unless this selector processes another digit in a digit absorbing application.
How does the line finder "find" the line calling line?
It didn't seem to traverse every contact on every level. It only traversed on one level.
When the shaft and wipers begin to move vertically upwards, you'll see a horizontally mounted copper coloured wiper brushing against a set of vertically mounted copper contacts. There is one contact for each level of the bank. The operated line relay causes a ground potential to be placed on the contact associated with its level. The shaft continues its upwards motion until its horizontal wiper detects the ground, at which time it stops vertical motion and begins rotary hunting. The top set of contacts on the bank assembly are the sleeve contacts, which provide the service request status of each line on the bank. If a line is requesting service, the operated line relay will cause -48 volt potential to be placed on its sleeve contact on the bank. The rotating sleeve wiper looks for the -48 volt battery and stops rotating when it sees it. At this point, the line requesting service has been found. If you are interested in more detail, you can watch my very long video on this subject entitled "The Line Finder In The Step-By-Step Telephone Switching System". Just search You Tube for my channel name, Hicken65. Thanks for watching.
Hicken65 ... Thanks for the explanation. It is a really complicated system! It took me ages to spot the circular piece of metal with the finger out to the right parsing the 10 tiny flat contacts.
I had already started to watch the other video you mentioned, but had to abandon it to get on with other stuff. It is bookmarked for when I can devote time to watch it all without interruption.
How did switching offices change when touchpad dialing was introduced? Also, why did the first through fourth selectors have multiple contacts on each row? was this so that any selector could be connected with up to 10 in the next position, thus almost "multitasking"m
Other switching systems, such as crossbar and ESS could handle touch tone. The SXS could not. However the telephone companies soon inserted an electronic device between the line finder and the 1st selector that could receive the tones and translate them into rotary dial type pulses that they would send to the 1st selector. The subscriber would have the convenience of touch tone dialing, but there was a noticeable delay before the distant phone would ring.
For details on the selector contacts, see my long video The Selector In the Step-By-Step Telephone Switching System Part 1. Just search my channel name, Hicken65, on You Tube.
In your examples, your step by step units are side by side to make one call from a telephone to another. In a central office, would this be the case, or would my dialing end up traveling between different areas of the CO ?
For instance, after I dial first digit as 5, would the connection be then made to a step by step that is dedicated to handle the second digit following a 5., and if I then dial 2, would be connected to a step by step that handles the 3rd digitl for numbers that begin with 52 ?
And how did that work when I dialed from one CO to another ? if I dial "52" and this belongs to a CO halfway across town, does this mean that after racteching up twice, it would hunt for a free line to the other CO, after which, the dialing pulses would be transmitted to the other CO? Upon receiving that connection, how would the other COP know that the first digits dialed were "52" ? Were there separate physical lines for each dialing prefix ?
(if this is all explained in another viddeo, just give me a link or video title).
In large offices, there were dedicated bays (320 selectors per bay) for each rank (1sts, 2nds, 3rds, etc.) of selectors. So yes, your call would travel all around the CO. For calls to other CO’s, they could exit the initial CO on the first, second or third digit. The outgoing paths were called trunks. If you were calling 529-1000, which was in another CO and the 52 pulses were handled in the initial CO by a first and second selector, the remaining pulses (9-1000) would travel over the seized trunk and arrive in the distant CO on an incoming 3rd selector. Only calls that began with 52 would reach the third selectors in the distant CO, so it didn’t have to sort out what was the first 2 digits were. The 3rd selector would process the digit 9. A 4th selector would be seized and process the digit 1. A 5th selector would be seized and process the digit 0. And finally a connector would be seized and process the digits 00. Generally, switches called “repeaters” were used on the trunks to freshen up the pulses sent out by the subscribers dials. If you are really interested in this subject, I have very detailed videos describing each switch type. Just search TH-cam by entering my channel name, Hicken65. Below is a link to a video (not mine) inside an actual in-service small SXS central office, that gives some idea of what the equipment looked and sounded like.
th-cam.com/video/InN6G2RnaCw/w-d-xo.html
Thanks for answer. When COs started to have multiple NNX, did they have to then build trunks for each NNX supported by the destination CO ? Or was a totally different method of telling the destination CO about the digits already dialed by originating telephone ? @@Hicken65
@@jfmezei Generally, separate trunk groups were not required. Most multiple NNX CO's had the same first 2 digits. In my previous example, if the distant CO had three NNX codes, they would likely have been something like 529, 528 and 527. The incoming 3rd selectors would receive 7, 8, or 9 and the calls would be directed to their correct destination.
@@Hicken65 I watched your video from start to finish, so interesting! The machinery looks incredibly delicate and those noises remind me somewhat of a typewriter crossed with a modem dail-up.
Thank you for that link watched almost the whole thing. What a sight to behold! My digital programmer brain is having trouble fathoming all those unique analog connection paths, connected by all those physical switchboards among millions of relays, connectors, wipers, contactors and what must have been hundreds of kilometers of copper wire in that building alone. Not to mention the part count for all of those machines, it had to be in the trillions (or maybe millions?). Anyway, dazzling stuff. Analog excellence, what a rush. Many, many people had to have gotten deaf in that room! XD
I mean this link BTW: th-cam.com/video/InN6G2RnaCw/w-d-xo.html
It truly is an epic watch, and it needs a boatload more viewers! Maybe you could (if the uploader permits it) download it, edit it to be more concise and structured and then voice over and explain exactly what we're hearing/looking at, what its purpose is and what it is doing at the moment. I would love a video like that but I lack the knowledge about these old telephone systems. But, I am handy with computers so if you'd like and/or need my help downloading the video from TH-cam (in decent quality) and converting it into a format that is accepted by the editing software of your choice just drop me a message :D and if needed I could probably also help you with recording audio. Please let me know, it would be superb!
Sorry, I don't relish the thought of attempting to edit that video. However, for a full explanation of how the line finders, selectors and connectors work, you need to watch my long videos on these switches. Just search my channel name, Hicken65, on You Tube. Thanks for watching.
How much could a complete system like this one could cost to build today with used parts?
Well, the short answer is I don't know. To build a demonstration unit similar to mine, here's what you would need: 2 line circuits, each consisting of one line relay and one cutoff relay; one line finder switch and its jackspring and bank assemby; five selectors with their five jacksprings and five bank assemblies; one connector and its jackspring and bank assembly; a 48 Volt dc power supply, a ringing generator, a dial tone generator; an interrupter for the ringing; an interrupter for for busy tone; wire to interconnect everything; and mounting framework for all the equipment which you can build using angle iron. There is no single supplier, that I am aware of that is selling this stuff. Occasionally, bits and pieces show up on Ebay and occasionally, some collectors list parts for sale in collector forums or newsletters. Telephone collectors hold several shows each year which often include switch demo units. You need to research such organizations as Telephone Collectors International and Antique Telephone Collectors Association.
I understand that the selectors are in the same frame for demonstration purposes in the video, but in the real world wouldn't the selectors for each digit be located in various parts of the CO with the last connecting selector just serving the last nnn-nnXX digits?
How would dialing the first two digits select a proper trunk to go to another CO?
Thank you. It's not usually explained well in any SXS video as the selectors grouped together is misleading.
Since the exchanges in an dialing area (area code) may be many, I'm curious how and when the outgoing trunk to another CO is selected -- I assume after the first two digits of an exchange are dialed then the appropriate trunk is selected to then connect the call to the appropriate CO for the last 5 digits to be decoded. That implies that there are trunks from each CO to all other COs in the area code, or was there another CO inter-switching method which was more efficient?
What does your demo setup use as sources for ringing voltage and dialtone? Does it use the standard motor-driven interrupters to make the busy signals?
The ringing generator is a Tellabs model 8102 made in 1998. It is a solid state generator that outputs 105 volts ac at 20Hz with a series -48 volts dc component. The dial tone/busy tone generator is a Lorain model F1 made in 1966. It is a solid state generator that outputs 600 Hz amplitude modulated by 120 Hz. The busy tone and ringing tone interrupters are both made by the Pylon Electronic Development company (year of manufacture and model unknown). They both have electric motors with brush and cam arrangements.
Great job with your demo system!
Was there a time back then when they didn't have an off hook tone? What year did they come out with the off hook tone?
From another video, it seems that the dial tone came along with dialing. Prior, the operator would answer.
How does the selector know when a line is in use and when one is available?
Hicken65 neat! thanks!!
Well the part I didn't quite get, was how the system could handle more than 1 call at a time, and how there was an upper limit to the number of phones that could be called. I take it, this was a simplified documentary
It is indeed simplified. My demonstration unit was designed only to show how a call progressed through a SXS switching system. When a switch is used on a call it is tied up to that call for the entire duration of that call and so cannot be used to process any other calls. When a call is established through my demo unit, all 7 switches are tied up and so only one call can be established at any given time. In a very large central office, each established intraoffice call would also tie up 7 switches, but there were many thousands of these switches. Each telephone line did not have its own dedicated switch or set of switches. To make efficient use of the switches, there was a great deal of shared access to them. For example, line finders were arranged in groups of up to 20 that had shared access from 200 lines and connectors were arranged in groups of up to 10 that shared access to the same 100 telephone numbers. Between the line finders and connectors were the ranks of selectors. Each line finder was tied directly to its own first selector. After that, the 2nd through 5th selectors were set up to be accessed on a shared basis from the preceding rank of selectors.Typically, SXS systems were provided with sufficient switches such that about 7% to 10% of the telephone lines could be in use simultaneously on originating calls and 7% to 10% could be in use simultaneously on incoming calls. A 50,000 line SXS office would have around 25,000 of these switches.
The Abelman Sisters ❤️
Thanks for showing what you could, but the design of these switches still has lots of unexplained details, not least of which is that the system also does automatic billing.
There is a lot more detail in my longer videos. Just search on my channel name. Unfortunately I do not have any Automatic Number Identification or Automatic Message Accounting equipment available to make a video. Thanks for watching.
is there any scenario where the selector wipers would stop prior to the 10th trunk (if its not completing the number)
Yes, the wipers always hunt from left to right and always stop on the first idle trunk found. In my demo unit, I have artificially made the first 9 trunks appear to be in use in order to enhance the experience of watching the motions of the wipers.
Beechwood 4-5789 You can call me up any old time, lol
I have another video, "Dialing Through The Step-By-Step Telephone Switching System" that describes the progress of a call to Beechwood 4-5789. Thanks for watching.
Ha, I remember that song. Wonder whose phone number that was....or is.... lol 😆
@@MarkMphonoman It was a working telephone number in many area codes and the subscribers that had it did probably did suffer from annoyance calls.
Is all that apparatus for every single line? I don't understand how it handles calls to consecutives numbers, by example, if phone lines xxx50 up to xxx59 can be use at the same time all the ten for receiving calls all at the same time? It works with parallelism of those mechanism?
My demonstration unit just has one switch of each type to show how a call progresses through a Step-By -Step system.In a central office, there is a lot of shared access utilized by parallel multipling of bank wiring so that multiple switches can share access to the same 100 or 200 circuits. For instance, in a typical setup, 200 lines share access to a set of 20 line finders and 100 telephone numbers share access from 10 connectors. In a 50,000 line office, there would be about 24,000 switches of all types, so just under half a switch per line.
wow, you mean that of 200 lines only 20 can do a call at the same time? that is like the 10% of sinmultaneosly usage!. But I imagine those 200 lines can receive a call regardless of all call lines are used.
Yes, that is correct, only 10% of lines could have simultaneous originating calls in progress and 10 % of lines could have simultaneous incoming calls in progress. This was typically sufficient. Of course, in serving areas that were business dominated, less sharing was done in order to handle the loads and still provide mandated service levels. The originating equipment did not interfere with the terminating equipment. If 20 of the lines in a group of 200 lines sharing access to 20 line finders were in use on originating calls, the remaining 180 lines were free to receive incoming calls.
and that it is in several layers? I remember to get the busy tone , on just the 3rd. number digit out of 6 or 7.
Yes, there are several ranks of selectors. One of each type is in my demo unit.You would get slow busy tone, 60 interruptions per minute (IPM) if you dialed an in use number. You would get fast busy tone (120 IPM) if there were no available trunks anywhere between digits along the dial path. If there were no line finders available, you didn't receive any tone. You just waited until one became idle and then you received dial tone.
A couple of questions:
Say my phone is 46789 and I wish to dial 34567
To reach me, are there 100 line connectors whose 8th row and 9th position output connects to my phone line?
Conceptually, if my number begins with 4, does this mean that my line connector would be at end of aisle 4 in the CO?
And if I pickup the phone, do I have my own line selector attached to my line (always the sime that hunts for a free trunk) ? Would the first line selector be next to it, but when I dial the first digit 3, I would be connected to the next line selector in aisle 3 that serves numbers that start with 3 ? (and from that point, would connect to selectors and final connector on aisle 3) ?
Just trying to understand on a larger scale how it was all arranged as opposed to the demos with just 2 phones.
Connectors were arranged in groups in which all the connectors in a group had access to the same 100 telephone numbers. For residential customers, a connector group would consist of up to 10 connectors. In your example , the connector group for your telephone would serve telephone numbers 46700 to 46799, and yes, your telephone would be on the 9th rotary position of the eighth level on the connector bank. In a very large SXS system, say 50,000 lines, there would be somewhere in the neighbourhood of 24,000 switches of all types. These were mounted on equipment bays that were 11' 6'' high, and as you can imagine this took up a tremendous amount of floor space. In large offices, switches were mounted on dedicated bays for each switch type. That is, there were dedicated bays for line finders, for each selector rank (i.e. 1sts, 2nds, 3rds, 4ths and5ths) and for connector groups. Any given 7 digit telephone call would tie up 7 switches spread over 7 different bays that could be located anywhere in the central office. The only bays directly tied to telephone numbers were the connector bays, which typically contained 7 connector groups and thus 700 telephone numbers.
The line finders were arranged in groups of up to 20 line finders that served 200 lines on a full sharing basis. That is any of the 200 lines could be connected to any of the 20 line finders in a group. When you originated a call, one line finder in your group would be activated, find your line and connect it to a first selector, which would give you dial tone as a signal that dialing could commence. If you dialed 34567, your call would tie up a particular line finder, 1st selector, 3rd selector, 4th selector 5th selector and connector. If you later made another call to that same number, your second call could be on a completely different set of switches that routed your call on a different physical path through the central office (although it would begin in the same line finder group and end in the same connector group).
For a much better understanding of this topic, you can watch my very long, detailed videos on the linefinder, the selector and the connector . Just search my channel name, Hicken65, on You Tube.
@@Hicken65 Thansk for the explanation. But still confused on 200 telephones sharing 20 line finders. When I lift the phone, the videos I have seen show the line finder being the first item. But if any of the 200 lines can any up on any of the 20 line finders, what is the name of the device that get me connected to any one of the 20 line finders?
Also would it be correct to state that if 20 people of the 200 in my group are already on the phone, when I pick it up, I would get no dialtone?
Would I need to hang up and pick-up again to try to get a free one, or could I stay off-hook and wait until I got dial tone?
@@jfmezei Each telephone line has its own line circuit, which consists of two relays, the line relay and the cutoff relay. When you pickup the the telephone , the line relay operates over the electrical loop through your telephone set. Each line finder group has a group control circuit that contains the preference circuitry to select an idle linefinder. The operated line relay initiates a request to this circuit to begin the selection process. If all the linefinders are in use, waiting calls queue up and will get served as line finders become idle. So if you stay off-hook you will eventually get dial tone. Details on this are provided in my video "The Line Finder In The Step-By-Step Telephone Switching System" which you can find on my You Tube channel, Hicken65.
@@Hicken65 Thanks. I had seen the video where you have a phone, line finder, selctors and connectors, and the phone appeared to be directly connected to the line finder. But your response shows there was something before that which connected my phone to a the first availabel line finder.
@@jfmezei You're welcome, and thanks for watching.
How does the system keep track of what number is being dialed? The only way I can think to achieve this is with incredible, costly, "fan-out": at every level, there's a connector (or bank of them) to handle each dedicated collection of digits. But this would mean that for every digit added, you'd need 10* more circuits than the last, and in your typical 7 digit number, that would mean 1,000,001 connector circuits (10^5 connectors, the final connector that can handle two digits, and one connector at the start that doesn't need any memory).
That seems like an awful lot of circuitry. Surely that can't be the way the exchange kept track of the number dialed at every stage.
While my video shows how a call flows through each of the 7 switch types, it does not explain any of the concepts of concentration, trunk distribution, bank multipling, trunk sharing or selector shelf subgrouping arrangements. In an actual central office there are hundreds or thousands of each switch type and a typical call is routed all around the central office instead of being confined to the small space of the demo unit. Let’s look at a theoretical scenario in which a SXS system has 10,000 lines and 10,000 station numbers. For simplicity, we will assume all traffic is intra switch, which means all calls originate from these lines and terminate to these numbers. Users only need to dial the 4 digit station numbers (i.e. 0000 to 9999). Assume a maximum of 10 % of the lines/numbers are in use at any given time in the busy hour (this is typical). I will need 1000 1st selectors to handle the
originating traffic (i.e. 10% of 10,000). Since only 4 digits are being dialed, we will not need any 2nd, 3rd or 4th selectors. All calls will route from 1sts to 5ths to connectors. The 1st selectors will process the thousands digits. Each of the 10 levels of the 1st selectors will represent a particular thousands digit of the station number. Each level will route the call to its own rank of 5th selectors. Assume the traffic is evenly dispersed across the 10 levels (i.e. 10% per level). We would provide 100 5th selectors for each level. That would give us 1000 5th selectors to handle the traffic from the 1000
1st selectors. Let’s study a call to station number 0987. The first digit dialed will be “0”, which will utilize level 0 of the first selectors. Any given 1st selector can only access 10 trunks on level 0. Selectors are arranged in groups of 10 on a shelf. The 10 banks associated with that shelf of 10 selectors are all multipled together, so that the 100 trunks on one bank (10 levels x10 trunks per level) are multipled to its adjacent bank, and the next bank and so on, such that there are only 100 trunks that can be accessed by the shelf of 10 selectors (10 trunks per level). So we now know
that on level 0, 10 selectors are sharing access to the same 10 trunks to a rank of 5th selectors. But we have 1000 1st selectors spread over 100 shelves. We know that 10% of the traffic from the 1000 1st selectors will be trunked on level 0, so we will need 100 5thselectors to handle that traffic. Since any given 1st selector shelf can only access 10 of these 5th selectors (associated with level 0), it is necessary to divide these one hundred 5th selector trunks into subgroups of 10 trunks each. This would give us 10 subgroups of 10 trunks to the 5th selectors from level 0 of the 1sts. We have 100 1st selector shelves. These also must be evenly divided into subgroups, with each subgroup of shelves having access to one of the trunk subgroups on level 0. This will give us 10 subgroups of 10 selector shelves, with each subgroup having access to one subgroup of 10 trunks to the level zero 5th selectors. This is accomplished by simply multipling the 10 trunks of any subgroup across the 10 circuit outputs of all 10 shelves of the selector subgroup, so that all 10 shelves (100 1st selectors) would be sharing access to the same set of 10 trunks to the 5ths. This set up is repeated for thousands digits 1 to 9 on the other 1st selector levels. This gives us 10 ranks of 100 5th selectors each, with each rank associated with a different thousand digit. The 5th selectors will process the hundreds digits of the station numbers. The 5th selectors are mounted on shelves of 10 with banks
multipled in exactly the same manner as the 1st selectors. Let’s continue with our call to station number 0987. We are going to utilize level 9
of the rank of 5th selectors associated with thousands digit 0. We have 100 5th selectors spread over 10 shelves. Assume the traffic is
evenly distributed across the 10 levels. Therefore, each level will require 10 trunks to connectors. Since the number of trunks for level 9 does not exceed 10, there is no requirement to create more than one subgroup. The 10 trunk outputs on level 9 of each of the 10 shelves of 5th selectors will
simply be multipled together, to provide one subgroup of 100 5th selectors accessing 10 trunks on level 9. The 10 connectors associated with these trunks will serve station numbers 0900 to 0999. On our call, the tens digit 8 and the units digit 7 will be processed by the connector. This set up will be repeated for the other 9 levels of this particular rank of thousands digit zero 5th selectors. This will provide 10 groups of 10 connectors, with each group serving 100 station numbers. The whole process described above will be repeated 9 more times for levels 1 through 9 of the 1st selectors. For each 1st selector there is 1 line finder. So overall, this office would have 1000 line finders, 1000 1st selectors, 1000 5th selectors and 1000 connectors.
I hope this helps your understanding of the SXS system. Thanks for watching.
@@Hicken65 see you really love it (just like me 😀👍)
@@anesthcat9420 Well, absence makes the heart grow fonder.
How does the system cancel a call incompletely dialed?
I don't recall any time out feature for partial dials, so the switch train would just be held up until the caller hung up. There was optional equipment available for releasing the equipment if no digits were dialed, or if a caller didn't hang up after a call was finished.
so was that whole big thing only able to handle one call at a time?
So in the central office, to handle one call it would require 4 of those switches?
ok thanks for the replies. So does one person's home connect to the exact same line finder every time? And same with the connector switch.. each house phone line is wired to one connector switch in the office?
PEnnsylvania 6-5000 would not have been a step by step office. It would have been in a panel office.
Well, the famous Hotel Pennsylvania is located right across the street from Penn Station, in the Manhattan borough of New York City. Since the panel switching system was the Bell System’s “Big City Solution” to customer dialed calls, I would think you’re probably correct. However, my demonstration unit is not trying to convey an historically accurate configuration for that location. I’m just trying to show how a call progressed through a Step-By-Step system using two well known (in their day) telephone numbers to enhance the demonstration.
Did every telephone have one of these giant contraptions dedicated to it at the phone company's office, or did multiple telephones share these switches?
No, they weren't dedicated. There was lots of sharing.
@@Hicken65 I remember watching an episode of The Six Million Dollar Man back in the 70's. He needed to "trace" a call and someone was like "that's not possible! No one can trace a call that fast!" But Steve Austin went into the room with hundreds of these switches and traced the call.
At the time I had no idea what I was looking at.
@@sam1174 I always enjoy watching old movies with scenes inside analog central offices.
That's answered in the video.
The only part of this that does not make sense is how these connect to the trunk cables.
The trunks are just 3-wire (tip, ring and sleeve) cables between each rank of selectors.These cables run from the bank terminals of one rank of selectors to the input side of the next rank of selectors.The path through a selector is from its input side, through operated relay contacts inside the selector, through the wiper cords, through the wipers, through the bank terminals on which the wipers have come to rest, through the trunk cable connected to those bank terminals and on to the input side of the selector to which the far end of the trunk cable is connected. This type of connection path builds up as each digit is dialed, so that after all 7 digits there is a continuous 3-wire path from the line finder to the connector switch.
Educational, but a snoozer