The Development of the Alternator

Alexanderson had made tests in strong magnet fields at high frequencies with a specially formed Swedish iron rotor and found the application satisfactory at frequencies of 100 kHz, and therefore wanted to replace the wooden core used in the alternator with this iron. Fessenden had asserted that the iron would melt in a strong magnetic field at the high frequency.

Alexanderson was encouraged by the vice president of General Electric to make an alternator in accordance with his own ideas. The result became an alternator with 2 kilowatt power at 100 kHz, run by a direct current motor at 2,000 r.p.m. In order to give the alternator the correct number of revolutions, he used a gearbox which geared up the alternator’s speed 10 times. The gear box replaced the previous primitive gearing which had used a belt and pulley.

The modesty Alexanderson showed during a long life came out during an interview when he discussed Fessenden’s and his own initiatives during the design of the alternator.

“Fessenden gradually was convinced and often expressed his appreciation of my work. How much were my ideas and his, it is difficult to say. It was a productive partnership. He asked me to do something, and I delivered. The patent on a specific solution of the problem is in my name, as it was my idea, but the overall idea, the objective for the work, was Fessenden’s.”

Fessenden’s role became, as the work with the alternator advanced, to encourage Alexanderson to experiment in accordance with those new proposals that Alexanderson had expressed in their discussions. The principal reason for the success was of course the financial support for the experiments which Fessenden could obtain through his company, The National Electric Signaling Co.

The 2 kW, 100 kHz alternator was sold in large numbers until 1910, when Fessenden lost all influence over the company NESCO. General Electric continued then to sell alternators to, among others, the US Army Signal Corps and the Japanese government.

The successful work with the alternator became the admission ticket to an engineering position at General Electric. On the first of May in 1906, Alexanderson moved from the Alternating Current Department to the Railway Department and to an independent engineering position with responsibilities for design of AC motors for railway purposes.

Besides his main task, Alexanderson continued to develop various parts of a system for long wave transmission, based on the alternator. The internal organization of General Electric did not require association with the GE Research Laboratory, yet his interactions with individual Laboratory researchers were lively.

In the year 1906, when Alexanderson began working at the Railway Department, it was generally considered in the USA that the electrification of the railway lines would occur. It was, however, argued within General Electric what type of current to choose for the locomotives. The argument for alternating current was the relatively efficient way of distributing and transforming the desired power to the motor. On the other hand, direct current had the advantage that the speed regulation was easier to perform and that reliable motors for underground railways and trams already were in operation.

It became a new challenge for the young engineer to develop an alternating current motor which had both the previously mentioned advantages, but also could compete in service reliability with direct current motors. Earlier, Alexanderson had become a contact to the Swedes, who were employed in America. On May 26, 1903 Robert Dahlander wrote, as an employee of the Swedish State Railway, a recommendation to the Director General for the Swedish railway in charge of experiments with electrical train systems, that he wished to contact Alexanderson during his USA visit. They had earlier corresponded concerning the performance of various types of motors for electrical operation. As mentioned previously, when Alexanderson visited Sweden in 1903 and 1906, he was invited to the ASEA engineer Ernst Danielson’s home. After the last visit, Danielson wrote to his friend in Schenectady and described his design of a so-called tandem motor for alternating current, in which the motor is constructed with two different pole numbers, thus providing three different speeds. He also discussed his idea of offering General Electric the opportunity to purchase the patent rights for the United States.

In the years 1905 and 1906 Alexanderson obtained patents on his earlier designs of alternating current motors. Therefore, he had gained experience from his own work and from the work of others in this special area, when he began to work for the Railway Department.

Alexanderson believed in the alternating current motor as the driving force for electric locomotives. He himself contributed some improvements and succeeded, among other things, in achieving motors with practically spark-free commutation.

One of the first railways where operation with AC motors was tested was the Washington-Baltimore-Indianapolis line. The experiment, however, became a failure because the trains were forced to halt at the Potomac River, as the bridges over the river to Washington, it seems, could not stand the heavy trains. The number of passengers decreased, and the company went bankrupt. The traffic was gradually resumed after a change to DC operation, and the new trains, probably considerably lighter than before, could cross the bridges over the river.

Alexanderson, who had not designed the above mentioned AC motors, now stepped in with his own design. This was used on the New Haven railway, and became a big success. The railway company in Washington wanted to blame failure entirely on the choice of current and claimed, with the majority agreeing, that alternating current would be abandoned for railway operation. Alexanderson, like Steinmetz, was sure that this was an incorrect conclusion. Twenty years later a group of experts agreed with the preference for alternating current over direct current. The locomotives subsequently built by General Electric for the Pennsylvania Railway were provided with motors of the same type as Alexanderson designed for New Haven, and this application also became a success for alternating current operation.

During his time at the Railway Department, Alexanderson could, of course, not work with the same dedication as previously on the development of the alternator. The economic difficulties that Fessenden’s company had gotten into contributed also to this slowdown.
GE, however, in 1906, had already delivered two alternators of 2 kilowatt output at 100 kHz which would be used for experimental purposes at John Hayes Hammond’s newly organized radio laboratory in Gloucester, Massachusetts. At the introduction of the first alternators, problems arose as a consequence of the fact that no electronic amplifiers were available at the time. In 1904, the Englishman John Ambrose Fleming had invented the diode that could serve as detector for modulated Grimeton radio waves, but it could not be used as an amplifier tube.This was the reason why GE delivered to Fessenden, along with the first alternator, a carbon microphone which had to be water cooled to withstand the load at high power level.

Alexanderson had, however, later made successful experiments with a so-called magnetic amplifier linked to the microphone and fed by the alternator. With this arrangement, a radio telephone line between Schenectady and New York was established.

During the years 1906 to 1910 Alexanderson was granted 52 patents especially concerning different types of motors and generators, security systems, voltage regulators, overload protectors and electric locomotives. Some patents were concerned however with development of the high frequency alternator. But, when Steinmetz in the year 1910 engaged himself in the Consulting Engineering Department, Alexanderson for the first time got a better chance to choose his own areas of activity.

Through acquaintance with John Hayes Hammond Jr. opportunities regarding development of radio were discussed, and Hammond told Alexanderson about de Forest’s audion (a triode), which at this time was considered to be the best detector for radio.

Alexanderson had long realized that an entirely electronic amplifier would replace the magnetic amplifier, if one only could design an amplifier tube suitable for radio frequencies. When he asked Hammond about the possibility of using de Forest’s audion as an amplifier, he immediately replied no. The reason was that the audion had too much residual gas and therefore could not be used for radio frequencies.

Fleming [see translators comments, p. 2] and de Forest had asserted that the remaining gas that was left after the audions had been evacuated was necessary for the tube’s correct function, and therefore stopped the evacuation before completion. Lee de Forest demonstrated, in October of 1912, his invention to The American Telephone and Telegraph Company (AT&T). The research manager H.P. Arnold and his team were amazed over the good function of the audion as a telephone line amplifier (low frequency amplifier). Successful experiments were made at AT&T Laboratories to achieve a good vacuum with new pumps, imported from Germany. Based on the experimental results, it was decided to buy the rights from de Forest to use the audion as a line amplifier.

Alexanderson could not give up the thought that the audion could be improved and thereby become applicable as an amplifier at high frequencies. He found it quite natural to discuss the idea with his colleague in the Research Laboratory at GE, Irving Langmuir, whose experiments within the new field of electronics he followed with interest. This was the beginning of a life-long cooperation and a permanent friendship between the two. Langmuir later, in the year 1932, received the Nobel Prize in chemistry for his research concerning surface chemical phenomena.

Langmuir, in 1913, immediately became interested in Alexanderson’s idea and started a number of experiments. He succeeded at once in producing a radio tube with sufficient vacuum for the amplification of radio frequency currents. Later he also produced a high frequency tube with many times more power, and this was used when linked to the antenna in order to modulate the alternator power for broadcasting.

During the experiments Alexanderson got the idea that using the audion as an amplifier could result in a much better selectivity* than so far achieved for radio reception. In 1913 he requested and in 1916 he was granted a patent on the idea. It was what he called “a geometric progression tuning.” “If a first tuning results in a selectivity of a hundred to one, the process can be done three times with radio frequency amplification, resulting in a total selectivity of a million to one.”
* Selectivity = ability to differentiate nearby radio stations from each other.