This section focuses primarily on the developments in North America. Naturally, there was progress being made worldwide especially in the UK, Germany, Sweden and France. See Ref 1 (Kingsbury) for more on the state of international automation development circa 1887-1915.
The Road to the Automatic Telephone Exchange
“How is it possible, when visiting an [automatic] exchange, not to feel humbled by these monuments of ingenuity? One cannot but marvel at the evolution of these intricate mechanisms. I can still hear [from my youth] the clicking and rattling of the rotaries, humming louder than crickets in the noonday heat…”
--Robert J. Chapuis (Ref 2)
Two years after A.G. Bell patented the telephone, the first manual switchboard went live in New Haven, CT, 1878. From this start, inventors and engineers dreamed about a 24x7 automatic system to replace operators. In 1879, one year later, inventors began experimenting with electromechanical switching systems.
For nearly 80 years, both automatic and manual telephone exchanges coexisted, each driven by various factors. However, it wasn't until the 1960s that automatic exchanges effectively replaced manual ones. By 1965 in North America, there were only 95 manual Bell System offices with an impressive count of 13,027 automatic offices with many serving up to 10,000 customers. Independent phone companies likely had similar ratios.
Baby steps
Most authors give Almon Strowger credit for inventing the first automatic exchange, installed 1892 in La Porte, Indiana. His story is legend, “Undertaker invents first automatic telephone exchange.”
Great story even though the first exchange was removed after only two years. La Porte became the world’s first “Exchange Lab” with 5 different exchange types installed over a 6-year period. It took several more years before practical exchanges were developed that were super reliable, cost effective, and used in small and large communities. See Strowger’s first exchange for a brief story of this famous telephone system.
There were many interesting switching developments before the first LaPorte experiment in 1892. The figure below shows Strowger’s key patent in relation to 87 others, from a variety of inventors, from 1879 to 1900. Several of these are essential patents, like Strowger’s, while others were either modifications of existing inventions or impractical designs with no future [Hill].
Fig 1, US patents associated with automatic telephone exchanges
* Almon Strowger had at least 3 other patents issued after his famous 1891 patent, but these were not significant. Using the name “Strowger” alone usually refers to the Strowger switch or a Strowger-style switch.
Some other prolific inventors during this time include Connolly, McTighe, Westinghouse, Keith, Erickson, Lundquist, and Callender.
But why consider these patents and not the many others after 1900? Well, this was the era that defined what an automatic exchange could be. It was the first generation of the exchange in terms of architecture and devices. By 1900, some important patterns had emerged that became models for generations to follow.
From 1900 until the 1960’s there were also many important exchange switching innovations. Each advanced the state of the art. The inventions were used in Step-by-Step, Rotary, Panel and Crossbar offices. To explore 15 switch inventions that were pivotal in exchange history, go to the Switch page. The all-important first Strowger switch and the Keith upgraded switch are covered here too.
The next section reviews some progress milestones for automatic switching development.
Inching from switchboards to fully automatic
The manual switchboard was relatively easy to manufacture, had few moving parts, and could scale from 10's of subscribers to one that could reach 10K subscribers with one operator cord. Switchboard exchanges were cascaded and could reach many millions of long-distance subscribers without needing dial automation. (Ref 3)
Fig 2 shows an example of a 3-position board from 1925 with about 9,000 reachable subscribers by any of the three operators. See a high resolution view of this image and zoom in to see the individual subscriber identified jacks.
Long-distance calls were costly and time-consuming because multiple operators in different cities handled each call.
Fig 2, Section of CB1 manual telephone exchange switchboard from the Enfield telephone exchange, 1925-1960, London Science Museum
Initially, AT&T took a conservative approach and sat on the sidelines in the dial automation race. From about 1877 until ~1914 Bell’s companies (or affiliated/licensed companies) in North America only installed manual switchboards. Based on AT&T’s evaluations, automation in smaller cities was not financially justified and switchboards were the smart choice circa 1900. True, they did own a few automated systems by way of acquisitions. Ref 3, page 65
Western Electric became a pioneer in switchboards and in 1879 WE engineer C. E. Scribner became the father of the “telephone jack” (US patent No. 293,198, Telephone-Switch, granted 1884). It was used extensively and became standard equipment on many switchboards. (Ref 3)
The invention was similar in form to a pocket clasp knife (jackknife) so the concept came to be called a “jack”, the naming still common ~150 years later.
Fig 3, Scribner's "Jackknife" switch and plug (Ref 3, page 492)
As an example of the switchboard’s long life in the Bell System, in 1950, 62.5% of central office codes (5,213 exchanges) were automated in North America with 37.5% (3,257 exchanges) being switchboards. (Ref 4, page 432)
Automation moves forward in the Bell System
The leading independent equipment supplier was very optimistic about automation. This was the Automatic Electric Company (of Strowger fame). They built and installed Step-by-Step systems in North America and Europe. In 1897 AEC installed a 900-line Step exchange at Augusta, Geogia. At the time it was the largest automatic exchange in the world. The first ever 10K line, rotary dial, exchange went live in Chicago, Ill (the "Loop" district) in Feb, 1903 [Smith].
AT&T could not ignore the move to automation forever, especially viewing AEC’s success. In 1902 AT&T had a change of mind and decided to develop a 10K line exchange for metro areas. They needed to solve the “big city” problem and concluded that Strowger-style switching exchanges would not scale for metro and long-distance connectivity.
An earnest effort started in 1906 and two related exchange design architectures were developed, nearly in parallel. Frank Jewett lead the Panel switch exchange engineering team and Frank McBerty lead the Rotary switch exchange (7A Rotary Automatic Machine Switching). Both systems were motor driven, with revertive pulsing, common control, and used sequence switches to manage call progress state. Panel was mechanically more complex of the two and J. J. Carty (AT&T Chief Engineer) decided to develop Panel for North America and Rotary in Europe and elsewhere. See Appendix B for more on the productive life of inventor Frank McBerty.
Fig 4 provides a great perspective on the continuous climb of dial automation in the Bell System. Note that although the number of dial phones equals the number of subscribers in ~1962, operators were still needed for subscriber assistance and especially for long-distance calling. Naturally, the number of operators exceeds the number of switchboards since the boards were staffed 24x7. The reduction in operators 1950-1962 is due to the gradual introduction of direct-dial for long-distance. (Ref 4, page 433)
Fig 4, Population, operators and telephones over time
Fig 5 The Faraday Building, London, 1939 remodel (Wikimedia)
In 1935, an automatic exchange was opened in central London, England, (Fig 5) with more than 6,000 working lines. The complex task of switching subscribers over to the new exchange involved 60 engineers working for more than 15 months. (Ref 5)
Metro exchange automation trials
In 1915 the first metro scale Panel field-trial dial system was introduced in North America in Newark, New Jersey. More mature Panel exchanges were installed in Omaha, Nebraska, in 1921 (6 digits) and in NYC in 1922 (7 digits). The Panel exchange had integrated operator assistance that were occasionally needed for call completion.
The Rotary exchange was slightly ahead of its Panel cousin in the European field. By 1911 Rotary had received more testing than Panel. With Frank McBerty in the lead, the first commercial exchange (800 lines) was installed in Darlington, England in October 1914. In 1915 other installations were made in England, France and Sweden. At least 9 million lines were under Rotary control by 1973, all outside of North America. (Ref 3, page 608)
Step finds its place
In 1919 the first Step by Step switch (Strowger type) was installed in the Bell System (excluding those acquired via company acquisitions) and went into operation in Norfolk, VA. The exchange switches were manufactured by the Automatic Electric Company. In 1926 Western Electric began manufacturing Step-by-Step switches based on license agreements with AEC plus Western’s improvements. (Ref 3, page 554)
By 1912 there were 5.1 million Bell System phones and 3.6 million independent company telephones. The proportions of independents decreased progressively until about 1940 due to Bell acquiring the independents [Ref 10, Gabel]. The Automatic Electric Company (or as GTE) manufactured Strowger type switches and licensed designs until the early 1970’s.
Table 1 (Ref 4, page 434) shows the number of central office codes (the first 2 or 3 digits of a number) and the mix of exchange types from 1950. The four types co-existed with manual, SxS and some Crossbar (X-Bar) for small/medium size cities and with Panel and Crossbar for metro locations. One can imagine a similar table for Europe and elsewhere with the 7A Rotary Automatic Machine Switching type replacing Panel.
To enable wide area calling from offices using Step-by-Step switches, the Director System was invented. This was used worldwide in many exchanges for inter-office calling. In a small way the Director was "Step's answer" to what Rotary and Panel were designed to do from the ground up.
Final words
From the earliest days of telephony, dial automation was the goal of many forward-thinking engineers. The switchboard had a ~100-year run. During that time, dial automation became more mature and widespread. Dial automation (including Touch Tone) played a crucial role in transforming the telephone industry, making it more efficient, accessible, and reliable.
Appendix A
Progress towards the automatic exchange
one invention at a time
The table below provides some highlights where a significant improvement occurred in either the telephone set or the Central Office before ~1906. Of course, there were hundreds of incremental advancements each year (worldwide), and this table selects a few of importance. These steps inched us closer to a practical first-generation automatic exchange.
The Table has separate rows for manual switchboard (S) and automatic (A). The automatic exchange and the switchboard evolved in parallel over many years. Ideas from both influenced the other. Next, each row is reviewed discussing the important advancement.
Table 1—Telephone and exchange notable improvements over time
Early switchboards (1S/2S) were often passive devices with no or a small battery for auxiliary purposes. (Errata: row 4, right col, first automatic at La Porte was 1892)
Progress point 1S: The first manual Switchboard. Note that the telephone set required a (1) internal 3-volt talk-battery and (2) a hand cranked magneto (AC generator) to signal the operator. On the switchboard side, a “drop” occurred when it sensed the magneto current from the telephone – “I want to make a call.”
Wall telephone with batteries and magneto
The picture below shows five subscriber lines as they appear on a switchboard. Subscriber 4’s “drop” has fallen (the electromagnet’s finger moved up causing the hinged cover to fall) and this notified the operator of a new caller.
Strip of subscriber “drops” and jacks on a switchboard, from [Miller]
Progress point 2S: The Line relay replaces the drop-electromagnet. This relay can now operate a switchboard lamp (new) or ring a bell (new) for nighttime operation. While this might seem like a minor advancement, it was indeed a significant milestone. But the Line relay was essential for what was to come -- the automatic system. So, introducing the Line relay was a significant step. For switchboards, a cranked magneto operated the Line relay instead of the drop electromagnet.
Progress point 3S: The CO Common Battery approach was invented by H.V Hayes [Chapuis]. This method eliminated both the set battery and the magneto. Now the CO battery powered the telephone set microphone, the Line relay and switching equipment for all subscribers. This was a huge advance and reduced the cost of the telephone and the annoyance of changing batteries. For large automatic systems, some CO batteries might supply thousands of amperes. See Power Systems.
The simplified and improved telephone set circa 1892 [Miller]
The figure above shows a set and switchboard but would look very similar with an automatic exchange. Now the same telephone can work with a switchboard or automatic exchange (adding a dial). See Line Circuit for more information on the Line and Cut-off relays in the figure.
Progress point 1A: This entry is for Strowger’s first automatic exchange in La Porte, Ind. The set’s “dial” mechanism had multiple switch buttons for the subscriber to press to enter a called number. It was 6 years later before the rotary dial was invented. La Porte subscribers had five wires (plus a common earth ground) going to each home phone, not two as is common today. Strowger’s patent shows five wires per subscriber – very inconvenient for telephone companies to install and manage.
There were batteries in each telephone and at the CO to operate the switches. It was not a common battery design. So, the plan was not optimized for low cost or simplicity and the La Porte installation was essentially a prototype and never went into mass production. It did, however, start a revolution.
Progress point 2A: Alexander Keith and the Erickson brothers invented the rotary dial in 1898 (at last!) and significantly improved the Strowger-type switch in 1899. These were pivotal inventions that forever changed the automatic exchange landscape. Almon Strowger had a great idea but Keith and the Erickson’s invented the future.
Appendix B
Frank R. McBerty
Frank McBerty was a prolific engineer and worked for 18 years for the Bell System and in the telephone industry for ~50 years. He was the driving force behind many of the key ideas in the Panel and Rotary Machine Switching systems for Western Electric.
McBerty was granted 148 US patents for telephone technologies from 1899 to 1947 with 21 granted in 1915 alone. Saying he was productive is an understatement. Let's look at one of his significant patents.
The Motor Power Drive
McBerty was tasked to create a 10K-line switching system for big cities as mentioned in the main body of this article. He knew it would require many interconnected switches, of various types, to make the end-to-end talking path connection. Each switch would need some form of power to drive the movable contacts for making the desired connection with a fixed bank (or arc) of terminals.
He was likely aware of existing motor powered "machine shops" that used a motor drive(s) with shafts distributing the rotational energy to individual end stations. The figure below is an example from 1893 showing lathes connected to a distributed power system.
Machine shop at the US Naval Academy, 1893
McBerty may have imagined using a single motor, shafting, and coupling gears to distribute power to each one of many switches. For sure he was aware of Strowger's method using dedicated electromagnets to drive each switch mechanism. He decided against this method.
Regardless of what sparked his imagination, he designed a "distributed power" motor drive system for his Rotary exchange. The basic idea was also used in the Panel exchange system. Motor power not only operated talking path switches (including line finders), but also sequence switches (state machines) and mechanical digit recorders. It would not be uncommon for one ¼ HP motor to drive about 250 ganged switches of various types.
Below is a figure from a McBerty patent (1915) describing his motor drive system. He only shows a small array of "end point" line finder switches but other end devices could also be attached. The main shaft connects to a motor (not shown) and is geared to several vertical and horizontal shafts. Each end-point switch's carriage is rotated when an attached magnetic clutch engages the carriage with the power drive. Every end point device has a magnetic clutch to transfer rotational energy into switch motion.
Below is a short video showing a working example of the motor drive powering end-point switches of various kinds in a 7A Rotary exchange. This out-of-service system is 105 years old (2024) and is fully functional. The author and Brian Cameron discuss the operations.
After leaving Western Electric, Frank McBerty became instrumental in the “all relay” exchange design as president of the North Electric Company. For more on Mr. McBerty’s life see the PDF article below by Brian Cameron, an expert on Rotary Automatic Machine Switching.
References
1- J. E, Kingsbury, The Telephone and Telephone Exchanges: Their invention and development, Longmans and Green, 1915, London
2- Chapuis, Robert, 100 Years of Telephone Switching, Part 1, 1982.
3- Fagen, M.D., History of Engineering and Science Early Years, 1875-1925, Bell System 1975
4- Pferd, W., Bell System Technical Journal, Feb 1979, Page 433
5- https://en.wikipedia.org/wiki/Faraday_Building
6- Hill, R.B., Early Work on Dial Telephone Systems, Bell Laboratories Record, Jan 1953
7- Miller, Kempster, American Telephone Practice, 1905, New York, McGraw Hill.
8- Smith, Arthur Bessey, The Early History of the Automatic Telephone, circa 1907
9- Waterson, K.W., Change from Manual to Dial Operation, Bell Telephone Quarterly, July 1930
10- Gabel, Richard, The Early Competitive Era in Telephone Communication, 1893-1920, Law and Contemporary Problems, Vol. 34, No. 2, Communications, Part 1, Spring, 1969
11- Hershey, Harry E., Automatic Telephone Practice, 2nd edition, 1917