Reference / Paper · 2025
The step-by-step telephone switching system: Handling of toll calls
Toll calls in SxS offices.
Reference / Paper · 2025
Toll calls in SxS offices.
The step-by-step telephone switching system: handling of toll calls Douglas A. Kerr Issue 2 July 22, 2025 FOREWORD In the 1920s and 1930s, the US had a well-developed, and very complex, system for handling long distance (“toll”) telephone calls, much of it operated by The Bell Telephone System. Step by step (SXS) switching equipment played two major roles in that scheme. Even during the era in which connections through the “interior” of the toll network were set up with manual cord switchboards, the completion of a toll connection to a subscriber served by an SXS system was done through a special part of that system, using special versions of the SXS switches. Then as the system evolved, special types of SXS switching equipment were used in the interior of the toll network, allowing the toll operator who initially handled the call to, by dialing, set up the entire toll connection all the way to the called station. These SXS switching arrangements use control and signaling arrangements quite different from those of the SXS equipment used for local calls. These in part arose to support special modes of call handling often involved in the long distance service. In this article we get an insight into how these SXS switching arrangements worked. It begins with extensive background on the surrounding context. The basic overall architecture of a toll call, as of the period of interest, is shown. Then the switch trains involved are described in detail from a functional viewpoint. The signaling and control protocols for the trains are described in detail in an appendix. 1 1.1 GENERAL Caveat This entire area is extremely complicated, and comprehensive, and authentic documentation is hard to come by. The author has relied on a wealth of information, including patents from the era of interest, but not every explanation “reaches from cover to cover”. Accordingly, in my detailed descriptions there will be some loose ends. Copyright © 2018-2025 Douglas A. Kerr. May be reproduced and/or distributed but only intact, including this notice. Brief excerpts may be reproduced with credit. The step-by-step telephone switching system–handling of toll calls 1.2 Page 2 Reader background It is assumed that the reader is generally familiar with the operation of the step-by-step (SXS) telephone switching system. If more information is needed, I commend the reader to the lead article of this series, “The step-by-step telephone switching system–Overview”. 1.3 The hardware The SXS switches described in this article use basically the same mechanisms, and operate in generally the same ways, as the switches used in “local” SXS switching systems, although there are some very significant operational differences, the most important of which will be described at length in this article. 1.4 The term “toll” Long distance calls early became described as “toll” calls in recognition of the fact that they were normally charged for “by the each”, the charge (a “toll”) generally depending on the duration of the call and the distance between the two stations. This was as distinguished from local calls, which generally were either “free” (that is, an unlimited number were included in the monthly change for the service) or charged for as a small fixed amount each after some “free” monthly ration has been used. Then, the modifier “toll” became used in the names of almost everything involved in long distance service (toll switchboards, toll operators, toll central offices, toll cables, toll switching machines, toll transmission plans, etc.). 2 2.1 TECHNICAL BACKGROUND The step by step local switching system Figure 1. Local intraoffice step by step connection For reference and comparison, Figure 1 shows in simplified block diagram form an intraoffice local connection between two telephone stations served by the same step by step (SXS) switching system central office. The step-by-step telephone switching system–handling of toll calls Page 3 The example presumes six-digit telephone numbers, with the first two digits indicating the serving central office (which in this example is the same central office that serves the calling station). Then, to complete the picture, Figure 2 shows in the same way an interoffice local connection between two telephone stations served by SXS switching systems in different central offices, with that same numbering plan. Figure 2. Local interoffice step by step connection 3 3.1 DELIVERY OF A TOLL CALL—THE TOLL TRAIN General If we consider, for example, the early 1930s, switching in the interior of the toll network was, with a very few exceptions, done on a totally manual basis, with special cord switchboards. Consider that context for now. Suppose now that the called subscriber in a toll call is served by an SXS office. Then the call is “delivered” from the last manual toll switchboard in the interior of the connection to the called line by a special train of SXS switches. Although this train, the “toll train” 1, directly parallels a corresponding part of the local switch train, there are many important differences in its operation (which is in fact one point of this article). In Figure 3 we see one example of a toll connection, manually switched in the toll network proper, delivered to a called station served by an SXS local office. 1 It is sometimes more-descriptively called the toll completing train and I will generally use that term here for greatest clarity. It is often sometimes called the toll switching train The step-by-step telephone switching system–handling of toll calls Page 4 Figure 3. Completion of a manual toll call In this example, the connection in the toll network proper is a simple one, involving only two manual toll switchboards, known as the outward and inward toll switchboards, with a direct toll trunk between them. (Of course in many real cases, some intermediate switching may be involved, but we do not need to think about that here.) The operator at the outward toll switchboard is responsible for the call, and will administer it from beginning to end. The operator at the inward toll switchboard is responsible for extending the connection to the toll train at the destination office, but she then stands back until the call is finished and the outward toll operator releases the connection. In the example, the calling station is also served by an SXS office. As was often the case in this situation, the caller dials 110 to reach the “long distance operator”. The first dialed 1 takes the 1st selector to level 1, where it reaches an auxiliary selector. Several other services reached by this selector also have three-digit codes beginning with “11” (e.g., “information”2, 113). The strategy for handling that scheme is that the auxiliary selector would ”discard” a dialed (second) digit of “1”.3 Then, in this example, 2 Known in modern times as “directory assistance”, to avoid any possibility that customers would think that they could call this number to ask about the population of Argentina. 3 This was actually called “digit absorption”. This strategy was used to save switch stages. The step-by-step telephone switching system–handling of toll calls Page 5 the third digit, “0”, takes the auxiliary selector to level 3, where it accesses a trunk to an outward toll switchboard.4 The caller passes the wanted city and telephone number to the outward toll operator. From her “Rate and Route Guide”, she finds what group of trunks she should use to extend the connection to that city. In this example, I assume that there are direct trunks to the destination city–that is, to the inward toll switchboard there.5 The outward operator selects an idle trunk in that group and plugs into it with her “calling” (front) cord. Through signaling that is beyond the scope of this article, this alerts the operator at the destination inward toll switchboard. The outward operator passes the called number to the distant inward operator. The inward operator plugs into a toll switching trunk6 to the office (SXS in this example) serving that number. The inward operator dials (perhaps four digits) into the toll train at that office. There, two selectors (of two different special types) and a connector (of a special type) establish a connection to the called line. I have marked the switches of the toll completing train, which are quite different in the details of their operation from the corresponding switches in the local train, with a little black triangle in the upper left corner. I will discuss those differences at length later. Note that the trunk labeled “toll trunk” could be appropriately called an intertoll trunk, since it runs between toll offices. But that term did not prominently appear during the era of the architecture shown. Keep it in mind, though. 3.2 Intertoll dialing Beginning in perhaps the mid 1930s, switching in the toll network proper began to be “mechanized”. The type of switching equipment that could be most expediently adapted to this was SXS equipment. Figure 4 shows the result of a connection made in this way, again for the simplest connection in the interior of the toll network. 4 This is called a “recording-completing” trunk for historical reasons that are beyond the current scope of this article. 5 That inward toll switchboard may in fact serve many cities in the same area. It will have “toll switching trunks” to each of them. 6 The name presumably refers to the fact that the trunk leads to a “switching system”. The step-by-step telephone switching system–handling of toll calls Page 6 Figure 4. Intertoll dialing For the most part, the toll train now operates in the same way as was described above. Here, the outward operator dials a series of arbitrary digits to establish a connection, through SXS switches (intertoll selectors), to the toll train at the destination office. The example assumes that there was a direct trunk to the destination toll office. But if no such was available, the operator might have dialed through an intermediate toll office, again using arbitrary digit sequences. This scheme of setting up the “interior” part of the toll connection (that is, between toll offices) by dialing into switches was known as ”intertoll dialing”. The switching train used was called the “intertoll train”. Later, the trunks between toll offices came to be called “intertoll trunks” (and I will use that term here). The signaling and control protocols used in the intertoll train were quite different from those used in the toll completing trains at the destination local offices. The toll transmission selector, the “incoming selector” of the toll train (but in a sense the end of the intertoll train), among other things mediated between the two sets of protocols. Just as I had marked the switches that were part of the toll trains with a little triangle, I have marked the switches that were part of the intertoll train with little black squares in their upper right corners. I call attention that the toll transmission selector seen here (labeled “B”—my notation only) is different than that seen in Figure 3 (labeled “A”). The “front end” of the “A” kind is arranged for an intimate relationship with a toll switchboard. The front end of the “B” kind is arranged for more generalized access by way of an intertoll trunk. The step-by-step telephone switching system–handling of toll calls 4 4.1 Page 7 TALKING BATTERY FOR TOLL CALLS Introduction An important consideration in the architecture by which toll calls are completed with SXS equipment is the matter of “battery feed”. In this section, I will the give some background on that matter. 4.2 About “battery” In the field of telephone systems, the DC voltage that is used for many purposes (energizing telephone lines, operating relays in switching equipment, etc.) is referred to as “battery”. This flows naturally from the fact that traditionally that this voltage (typically nominally 24 V in manual switching systems, 48 V in “machine switching” systems) is provided by a storage battery “floated” across the output of a DC power supply. 4.3 The carbon transmitter From the introduction of the common battery mode of operating telephone lines through perhaps the 1980s, almost all telephone sets used a transmitter (microphone) of the variable-resistance carbon type. Here, the diaphragm, responding to the acoustic speech wave, flexes a conductive wall of a small chamber filled with carbon granules. The opposite wall is also conductive. As the pressure through the mass or carbon granules changes, the resistance through the mass (between the two conductive walls) varies. A nominally-constant DC current is passed through the transmitter. As its resistance changes so does the voltage across it, following the acoustic wave. The varying component of that voltage is an AC “audio”7 signal reflecting the speech. For a given acoustic input, the level of the electrical output signal varies significantly with the level of the DC current. 4.4 Battery feed In all but some specialized cases, under the common battery principle, the DC current that energizes the transmitter the DC current that flows in the telephone line, supplied from the serving central office. Given its importance to the operation of the transmitter, it is often spoken of as talking battery. 7 But we never say “audio” in this context, Rather, such signals are called “speech” or “voice frequency” signals. The step-by-step telephone switching system–handling of toll calls Page 8 In many cases, the circuit that feeds the DC current is closely integrated with the circuit that couples the speech signals between the two sides of the switch (or other circuit). In Figure 5, we see where the battery is applied in our example intraoffice local connection with SXS equipment once the connection is completed. Figure 5. Local intraoffice step by step connection–battery feed We see that the connector feeds the battery to both stations. In Figure 6, we see where the battery is applied in our example interoffice local connection with SXS equipment, again after the connection is completed. Figure 6. Local interoffice step by step connection–battery feed We note that, as in the intraoffice case, the connector provides talking battery both “forward” (to the called station) and “backward”, but here the latter does not reach the calling station. Rather, in this case, The step-by-step telephone switching system–handling of toll calls Page 9 the talking battery is feed to the calling station from the trunk circuit8 at the near end of the interoffice trunk. The reason it is done this way is that the combined resistance of the interoffice trunk and the calling station line might reduce the current below a desirable level were it in fact provided from the connector. Keep this structure in mind for later when I will discus departures from it in the operation of the toll network. 4.5 Loss in a toll connection “In the day”. trunks in the interior of the toll network were conveyed over physical pairs (either on open-wire lines or in “toll cables”). Eventually, repeaters (amplifiers) were used to overcome, in part, the attenuation that would be afforded by those lengthy pairs. But for reasons that are beyond the scope of this article, there was a lower limit on the loss of these “repeatered” pair circuits, based in part on their physical length. The result was that for a “long” toll connection, the loss of the interior portion might be quite substantial. 4.6 One mitigating factor One way to overcome this was to strive for the greatest possible electrical signal output from a telephone set for a given level of speech. One factor that directly influences the electrical output from a telephone set for a given speech volume is the DC current through the transmitter, which is in fact the DC current in the line. (See section 4.3.) Thus in a toll connection, it was sought to make that current consistently at the greatest value that was practical. This had a substantial influence on the siting and design of the transmission circuits that formed part of the toll completing trains. 4.7 4.7.1 Battery feed circuits In a local connection For reference and comparison, Figure 7 shows a typical example of the transmission circuit used in an interoffice local call in the SXS system, one of whose jobs is to provide talking battery feed to the two lines.. As we saw earlier, this occurs in the connector. 8 Actually called a “repeater”, since one of its purposes is to repeat the dial pulses over the interoffice trunk. The step-by-step telephone switching system–handling of toll calls Page 10 Figure 7. Illustrative local transmission circuit Battery and ground (collectively, “battery”) are fed to each line through balanced windings of two relays (here labeled with their typical designations in a local SXS connector). The real purpose of these relays is to determine whether the two lines are on hook or off hook. Thus they are actually supervisory relays, but because of one major purpose of this circuit, they are most often described as “battery feed” relays. The speech signals are carried between the two lines through coupling capacitors C1 and C2. The relay windings have fairly large inductances. Thus they do not “short circuit” the AC speech signals. This circuit does not regulate the line current, which varies with line resistance. Thus for a longer line, with a greater resistance, the current is less, and thus the transmitter output is less. But because of the increased loss of a longer station line, this is just the time we would rather not have a lower transmitter output. Still, in local service, even subscribers on longer lines received “acceptable” end-to-end transmission performance. I note also that in manual switching systems, the battery voltage normally used was 24 V, which made the situation here even “worse”. 4.7.2 On toll connections But especially in the earlier days of long distance service, where it was difficult to fully overcome the attenuation of the various lines, this shortcoming of the basic SXS battery feed circuits would be particularly adverse. This suggested that when the telephone set was involved in a long distance connection, it was desirable to use a different transmission circuit, one with a less-primitive battery feed circuit. Figure 8 shows a typical “toll” transmission circuit: The step-by-step telephone switching system–handling of toll calls Page 11 Figure 8. Illustrative toll transmission circuit Here, the line current is regulated by resistance lamp RL. As the current through it increases, the filaments get hotter, and their resistance increases. Thus the current through the lamp remains relatively constant with changes in the resistance of the “load” (in this case, the station sets as seen though their lines). This is spoken of as “toll grade battery”. The coupling of speech circuits across this circuit is done through a repeat coil 9 (audio transformer), RC. Relay SV (“supervisory”) has a fairly low resistance. Its purpose is to determine whether there is current flow in the circuit or not, and thus determine whether the station is on hook or off hook. Capacitor C2 provides a low impedance path for speech signals, preventing any significant fraction of the speech signal current from passing through SV and RL, where the power it represented would be dissipated and thus lost. A similar situation is seen at he “left side” of the repeat coil, where the two conductors handle DC signals that are part of the trunk signaling protocol. A secondary benefit of this circuit is that the transmission loss through it is typically a bit less than through the “local” transmission circuit see just earlier. The need to utilize this circuit leads to a couple of the differences between they way SXS equipment is configured and operated in local and toll connections. 4.8 Siting of the battery feed Figure 9. shows the same toll connection shown in Figure 4 except that the siting of the battery feed to the two stations is now indicated. 9 At first called “repeating coil”, since its job was to “repeat” the speech signal. The step-by-step telephone switching system–handling of toll calls Page 12 Figure 9. Toll connection showing battery feed to stations Firstly, for comparison, at the top of the terminating local office we see a reminder of where battery feed is sited in an intraoffice local connection. In the toll connection, we see that battery is fed to the called station from the toll transmission selector. And of course there is a transmission coupling path there as well, the main reason for the name of that selector. We also note that the calling line is fed battery from the recording-completing trunk circuit at the outward toll switchboard.10 This typically follows the same arrangement seen in Figure 8 (aimed in the opposite direction). There is DC on the conductors of the dial intertoll trunk and toll switching trunk, but it does not have any “transmitter current” implications. Its role is totally for signaling. We will be looking into that in detail in a little bit. Since in the local train battery is fed to the called line from the connector, why in the toll train is it fed from the toll transmission selector? Note that as a consequence of the relatively-few toll calls per station in this office, there were far fewer toll transmission selectors than toll connectors. The toll battery feed and transmission circuit is more costly than that in the local connector and takes a lot of space (the repeat coils used at the time were quite large). Thus it was more 10 The odd name of this trunk and its trunk circuit is an artifact of a specific mode of handling toll calls, with which we will not be bothered in this article. The step-by-step telephone switching system–handling of toll calls Page 13 beneficial to locate this more costly circuit in the toll transmission selector. 5 COMBINATION CONNECTORS It was actually quite common in SXS offices to use combination connectors. These could either operate with a local connector “personality” or a toll connector “personality”, depending on through which of their two “ports”11 they were seized by a selector. The concept is shown in Figure 10. Figure 10. Combination connector When the switch is seized through either port, it is marked busy at the other port. The arrangements by which local selectors choose a connector are made so that local selectors will make the subgroup of combination connectors their last choice, thus reserving them, so far as possible, for toll calls. 6 COMBINATION INTERTOLL AND TOLL TRANSMISSION SELECTOR In some network setups, it is desirable to have the same selector, for some calls, to act like an intertoll selector (connecting perhaps to a toll transmission selector at the destination office, in another building) and, for other calls, to itself act as a toll transmission selector, connecting to toll intermediate selectors in a destination office in the same building. This would allow a smaller number of connectors to be installed for any given combination of local and toll traffic. 11 The term “port” was not actually used in telephone industry practice “in the day”, but in a modern context it is well descriptive. The step-by-step telephone switching system–handling of toll calls Page 14 Of course, in these two tasks, the technical arrangements are quite different. For example, when acting as a toll transmission selector, the switch must include a battery feed and transmission coupling arrangement; when acting as an intertoll selector, it does not include that, but rather provides a through path for T and R. And there are differences in the signaling and control protocols on the “downstream” side. These two personalities can be selectively exhibited by a combination intertoll and toll transmission selector. The personality in effect depends on the level to which the selector is dialed, under control of a contact on a normal post contact assembly.12 This assembly provides for contacts that can be operated when the selector is at any of certain levels (this being done by a sheet metal “cam” having teeth for each level that can be bent out to operate a contact when the switch is at that level). The details of this creature are beyond the scope of this article. 7 7.1 STOP DIAL OPERATION Introduction We tend to think that, in the SXS system, as soon as one dialed digit is finished (that is, there have been no pulses for long enough that it is clear there will be no more for this digit), the next digit may be dialed. This is not exactly so. After the digit is dialed into a selector, the selector must hunt along the ten terminals on that level for an idle next stage switch and, finding one, must “cut through” to it. Then the A and then the B relay in that next switch must operate13 before that switch is prepared to receive that next digit. The basic general network planning specification requires that (when digit pulse trains are sent by a “sender” circuit rather than by a dial) there be a minimum 600 ms interval (the “interdigital time”) between the end of the last pulse of one digit and the beginning of the first pulse of the next digit. But the system design also relies on the assumption that when the pulse trains come from a dial, the human who is dialing (subscriber or operator) takes a while to “wind up” the dial for the next digit, which 12 The name comes from the fact that the contact assembly is mounted on the normal post of the switch, a vertical rod whose principal “regular” job is to be the backstop for a lever on the switch shaft so as to establish the “normal” (that is, not rotated) rotary position of the shaft regardless of its vertical position. 13 Actually, the B relay must be operated long enough (“soaked”) that its release time will then be as expected. The step-by-step telephone switching system–handling of toll calls Page 15 will (hopefully) in most cases give a sufficient minimum interdigital time. And if a subscriber consistently got wrong numbers, it might have been suggested that he dial at a bit less energetic pace. 7.2 Call to a panel dial office We have concentrated here on the assumption that the called station is served by an SXS office. But, during the era assumed for most of the article, many larger cities were served by the panel dial system, a very complex “common control” switching system. There, for an interoffice call that arrives from a “dial pulse” context (such an SXS office), when an incoming trunk is seized, an incoming sender must be attached to the trunk to receive the dial pluses. That sender “sends to the switches” the information needed to set up the connection14. The sender is then detached from the trunk and becomes available for assignment to another incoming call. The connection of an incoming sender to the trunk takes a little while (and that period can be prolonged if all the incoming senders happen to be already in use on other connections). Thus if, as soon as a trunk to a panel office is seized, the dialing of digits (in our case, by the outward toll operator) continued without pause, the first few pulses (or even an entire digit or more) might well be lost, and of course the connection will not be (properly) completed. There was a parallel situation in calls to other types of switching systems 7.3 Pulsing control protocols This is in fact a generalized issue in the telephone network as switching systems (both local and toll) more and more became of the “common control” type, in which incoming digits were always received by a sender or register, which had to be attached to a trunk when it was seized. Accordingly, there arose a set of protocols for what may be called “pulsing control”15. All of these protocols incorporated the mantra, “Never pulse into an off hook”16, where “off 14 The name “register-sender” would have been more apt for this unit, but the name “sender” was used because of the designers’ preoccupation with the unique way the information was “sent” to the switches 15 Actually, it is commonly called “outpulsing control”. That is because the sending out of digits from a switching machine (not directly from a dial) is typically said to be “outpulsing” (distinguishing it from the capture of arriving of digits, which however is not called “inpulsing”). 16 There is one obscure exception in a protocol associated with direct distance dialing. The step-by-step telephone switching system–handling of toll calls Page 16 hook” referred to one of the two syates of the (rearward) signaling channel. The protocol that pertains to our story here is the stop dial 17 protocol. It works this way. When the setup of the connection reaches a point where the system is not yet ready to receive a digit, the distant entity returns the off hook state on the rearward supervisory signaling channel. The near end knows this can’t mean that the called station has answered (its basic meaning): the entire number hasn’t yet been sent. So it interprets this off hook state as a stop dial indication, and it refrains from sending any further digits until the rearward supervisory state returns to on hook. 7.4 In an intertoll selector In the SXS intertoll train, though, there is a “race condition” problem. Suppose the digit dialed into the selector takes it to a level where there are trunks to a “register-based” system. Once the selector has cut through to the trunk, the panel system will promptly return the off hook stop dial signal. (In fact, the trunk will likely return the off hook even when it is not yet seized.) And this stop dial signal will propagate to the toll switchboard, to advise the operator not to dial any more until further notice. But, after the end of the dialed digit, there is a delay before cut through occurs (while the selector hunts on the dialed level for an idle trunk to the distant system). By this time, the toll operator may well already be winding her dial up for that digit. And some of the pulses of that digit (maybe all of them) will be lost, and the connection will not proceed properly. Thus, intertoll selectors that might access trunks into panel offices have a special provision, following what is called the delay dial protocol. When the switch goes to a level with trunks to a panel office, a contact on the normal post spring assembly is operated.18 As soon as the digit that had moved the switch to this level is finished, this contact assembly results in an off hook state being sent to the rear by the selector. 17 Although, as seen from the perspective of the entity that is being dialed into, it would be called the delay dial protocol. 18 A separate one from that used to control automatic ringing for ringdown trunks, a feature that the selector might also have. The step-by-step telephone switching system–handling of toll calls Page 17 This lights the cord supervisory lamp at the outward toll switchboard, advising the operator, as soon as practical, not to dial further digits until “further notice”. Then, when an idle trunk to the next stage has been found, and the switch cuts through to that trunk, that ground from this selector is removed. But immediately, there will be an off hook state (its delay dial indication) coming from the distant office, so the stop dial indication is continued. Then, when the distant office is ready for further digits (it having attached an incoming sender), the rearward supervisory state from it is returned to on hook. At the outward toll switchboard, the cord supervisory lamp goes dark, and the operator knows she can resume dialing. 8 8.1 CALLBACK AND RINGING CONTROL FUNCTIONS Introduction Various scenarios of handling the wide range of long distance in the varied situations that are encountered place special requirements on the toll completing train. 8.2 Callback There are several scenarios in which the outward toll switchboard must call the calling subscriber. One is to set up a suitable connection from the calling line to the outward switchboard. Another scenario is this. The outward toll operator cannot advance the connection because, say, all the trunks that might be the next link in the connection are busy. The outward toll operator advises the caller that the call cannot be completed now. She tells him that she will call him back when the call can be completed. When that “congestion” has been relieved, a “delayed call” toll operator calls the caller (whose number she has on the ticket made by the original toll operator at the start of the call). But if she did this through the local completing train of the office serving the caller, the battery feed and transmission circuit for the caller would be that used for local connections, As we learned earlier, this is not desirable for a toll call (which will presumably ensue). So instead this new outward toll operator calls the caller through the toll completing train at the office serving the caller. The toll transmission selector there of course provides a battery feed and transmission circuit worthy of inclusion in a toll connection. The toll transmission selector (in particular) must be prepared to properly work in this “retrograde” mode. The step-by-step telephone switching system–handling of toll calls 8.3 Page 18 Delayed ring Suppose that on a person-to-person call, the wanted party is not home at the time of the initial call, but her father (who answered) says that she will certainly be home by, say, 3:00 pm. The toll outward operator asks the caller if he would like to have the call placed then. He says yes. When a “delayed call” operator goes to set up that follow-on call, to expedite its eventual completion she may, before trying to call the wanted line, call back on the caller’s line, but in a way that ringing is not started. Among other things, this makes sure that the caller’s line is not busy as this scenario plays out. When the operator calls the called line, and it is answered, and the wanted party comes to the phone, the operator asks her to wait while the caller is brought into the connection. She then operates a key that makes ringing commence on the connection to the calling line. When the caller answers, the operator says, “We have your party on the line.” And the conversation ensues. 8.4 Rering Suppose that a toll call is placed on a “collect” basis, where (with the assent of the called party) the changes are billed to the called subscriber. Suppose that the caller accepts the operator’s invitation to learn, after the conversation is finished, how much the charge will be. Thus, after the conversation ends, the outward toll operator must again get in touch with the called party. She does this on the remnants of the original connection, pressing a key that makes the toll transmission selector again ring the called line. 8.5 In summary We can see that these “gymnastics” sometimes required in the handling of a toll call require special capabilities in the toll completing train. Most of these actually fall to the toll transmission selector, but the toll intermediate selector has to “play along”. When I discuss the detailed protocols (in Appendix B), that collaboration will be more apparent. 9 A-B TOLL OPERATION Imagine that for subscribers in Ourtown, calls to Nearville, which is not very far away, are nonetheless, from a rate standpoint, “toll calls”. And there are a lot of those “toll calls” made. The step-by-step telephone switching system–handling of toll calls Page 19 Most of these calls do not involve the kinds of complications that may arise in the broader field of toll calls (such as those discussed above in Section 8. To reduce the cost of handling such calls, and as well to make this some what more convenient for the subscriber, rather than handling these calls through the toll network (as described in great detail earlier), such calls are often handled by a simpler operating method most often technically known as “A-B toll”. This is discussed in detail in Appendix A. 10 ELIMINATION OF THE SEPARATE TOLL COMPLETING TRAIN Over time, changes in the transmission systems used led to the ability to have a lesser loss on the “interior” of the toll network. This in turn relieved the need to “squeeze the last bit of output” from the telephone set. In addition, the newer telephone sets (notably the 500-type) provided more consistent output with varying line current. These considerations reduced the value of an improved battery feed circuit in the toll competing train. Additionally, the decrease in toll call rates (on an adjusted for inflation basis) decreased the number of toll calls that were made on a “person-to-person” basis (which calls often required some of the special features of the toll completing train). Finally, the onset and eventual wide acceptance of Direct Distance Dialing further decreased the number of toll calls that were operator-handled and thus which might have required some of the special “gymnastics” that were supported by the separate toll completing train. As a consequence, starting in I think in the early 1960s, the Bell System moved toward the use of the “local” completing train for both local and toll calls (thus eliminating the separate toll completing train). This led to substantial cost savings in the local offices. One ramification of this was that toll operators would no longer be able to “re-ring” a party nor use “delayed ringing”. As suggested just above, the number of calls requiring either of these tools had greatly diminished. New operating practices were instituted to allow every call situation to be handled without these special tools. For some of these calls, this led to a more labor-intensive process than before, but the overall economic situation was greatly improved by this change. The step-by-step telephone switching system–handling of toll calls Page 20 11 SEIZURE AND PULSING In the local network selectors (all stages) and connector are “seized” by a DC closure (“bridge”) between the tip and ring conductors of the circuit, and digits are sent by a series of “pulses” that are open intervals in that bridge. But intertoll selectors, and the directly-related type of toll transmission selector, are seized with a what is called a simplex ground, and digits are sent by pulses of open on that ground. This matter is described in detail in Appendix B 12 INTERFACES AND PROTOCOLS In the local network selectors (all stages) and connectors are “seized” by a DC closure (“bridge”) between the tip and ring conductors of the circuit, and digits are sent by a series of “pulses” that are open intervals in that bridge. Typically the trunk from one switch stage to the next (if intrabuilding) comprises three conductors, the ring (R), tip (T), and sleeve). These trunks, if interbuilding, generally comprise only two conductors (R and T) But in the toll switching system intrabuilding links between successive switches may comprise four conductors (R, T, S, and a fourth conductor, designated SP or C depending on just where we are and what it does. And the protocols for the interaction of the switches differs considerably from that found in the local switch train. And intertoll selectors, and the directly-related type of toll transmission selector, are seized with a what is called a simplex ground, and digits are sent by pulses of open on that ground. The details of those signaling and control protocols and the uses of the four leads found at a particular interface are covered in considerable detail in Appendix B. -#- The step-by-step telephone switching system–handling of toll calls Page 21 Appendix A The A-B toll method of operation A.1 INTRODUCTION Imagine that for subscribers in Ourtown, calls to Nearville, which is not very far away, are nonetheless, from a rate standpoint, “toll calls”. And there