Valdemar Poulsen: Inventor of Voice Mail

Subject: Jurassic Telecommunications, Part I
Date: Thu, 08 Jul 1999 19:22:13 -0400
From: Donald E. Kimberlin 
Organization: TELECOM Digest
Newsgroups: comp.dcom.telecom

Jurassic Telecommunications, Part I

Valdemar Poulsen - The Doctor Frankenstein of Telecommunications?

If the immortality we hope for really exists, then it follows there is likely a collegium of archangels or a pantheon of gods of man's higher accomplishments. Valdemar Poulsen rightly deserves a place in such a group for his contributions to man's shrinking of time and space; to man's increase of social intercourse, and thus, one would hope, the furtherance of peace and harmony in the world today.

However, one of Poulsen's major contributions has had its dark sides as well as its benefits.

The first notion of recording sound by magnetic means seems to have been stimulated rather early in Thomas Edison's spew of development around 1880.

Then-prominent American mechanical engineer Oberlin Smith, after a visit to Edison's Menlo Park, NJ laboratory, filed an 1878 patent caveat that was never followed up. It described the notion of recording electrical signals produced by a telephone onto a steel wire.

While investigating ways in which speech might be recorded, Edison;s assistant Sumner Tainter noted on March 20, 1881; "A fountain-pen is attached to a diaphragm so as to be vibrated in a plane parallel to the axis of a cylinder. The ink used in this pen to contain iron in a finely divided state, and the pen caused to trace a spiral line round the cylinder as it is turned. The cylinder to be covered with a sheet of paper upon which the record is made." (1)

It's interesting that had the Edison team followed this route and succeeded, the world may have had postally mailable recordings on paper sheets a hundred years ago. Rather, however, developments focused on purely mechanical means to record and play back sound. Not yet having any of the electronics necessary to amplify the weak magnetic signals or to prepare the magnetic medium by biasing it, mechanical recording certainly would have been seen as the only practical method of the era. From the Edison notebooks, it seems that idea lay fallow for almost two decades.

Oberlin Smith decided in 1888 that he would not pursue his idea. He "donated" it to the public by publishing his ideas about magnetic recording in the journal Electrical World. (2) This publication may have caught the interest of Poulsen, who after all, had attended the university at which earlier Danish physicist Hans Oersted made the connection between electricity and magnetism in 1820. By 1893, then 24- year-old Poulsen was working for the Copenhagen Telephone Company.

Poulsen attacked a point about magnetic recording that Edison had not addressed -- the matter of how to play back a magnetically recorded message. He found that, indeed, Faraday's principle of magnetic induction would operate to make a magnetic recording playable. Poulsen's first demonstration device was simply a steel chisel edge along which he moved a small pickup coil. He sidestepped a suggestion by Smith of using cotton thread impregnated with iron powder, advancing directly to a wire suspended across a room. He mounted the record/pickup coil on a moving trolley.

To achieve a compact and portable device for his patent application, Poulsen had by 1898 formed the wire into a drum-like vertical coil. This was rotated with a crank to cause the wire to pass under a fixed record/pickup coil assembly, as shown here. (3)

Poulsen's earliest patent papers showed he was aware that tape was a practical option to wire. It was not until later designers attempted to store steel wire on reels that wire twisting became an irritating source of high audio frequency loss. That change was not to ensue until around 1928, when Germans working for AEG and BASF addressed the Edisonian notion of applying iron power to a paper (by now paper tape) backing. This created the Magnetophon tape recorders used in German broadcasting until their discovery by American Jack Mullin at the end of WW2.

But, back to Poulsen and his first development. At the outset, his Telegraphone was intended to store either analog speech or digital Morse telegraph signals. Poulsen's original Danish patent application indicated his Telegraphone was intended for use to answer unattended telephone lines and record messages for later playback.

Thus, we see that Valdemar Poulsen's first plan for his development was to provide Copenhagen Telephone Company with central office based voice mail, which of course, has a parallel in the telephone answering machine and other forms of voice mail we now encounter daily. Much is made by persons in the recording industry of Poulsen inventing magnetic recording, but little or nothing is said of the often frustrating other outcome of his work!

It would appear, however, that the world little appreciated Poulsen's breakthrough at the outset. He took it to the Paris Exposition of 1900, there paralleling a promotional device used by Alexander Graham Bell a quarter-century earlier. Just as Bell managed to get the Emperor of Brazil to exclaim interestedly that a telephone worked (in Philadelphia in 1878), Poulsen snagged Emperor Franz Joseph of Austria into a demonstration of recorded voice on the Telegraphone. Based on that royal attention, the Telegraphone was described in glowing terms by the technical and scientific press as superior to the phonograph and a great advance in physics as well. It won Poulsen a gold medal, but not business success.

Poulsen obtained patents on his Telegraphone in a number of nations, and even founded an American Telegraphone Company in 1903, with a manufacturing plant in Wheeling, West Virginia. Efforts to market the Telegraphone as a business office dictation machine met with little success, but a number of Telegraphones were marketed to railroads through Western Union Telegraph as recording devices for Morse telegraph messages. Correspondence in the Lemuelson Collection of Western Union at the Smithsonian Institution attests to use of Telegraphones on the P. and R. railroad, the Northern Pacific railroad, the L. and N. and the D. and H. railroads.

One can surmise the Telegraphone drew AT&T's attention, as a version was offered that could answer an unattended telephone - even in 1903! American Telegraphone moved to Springfield, Massachusetts in 1910, then went into bankruptcy receivership in 1918, never to emerge; only to finally close in 1944 following Poulsen's 1942 death.

Other interests, however, benefited and prevailed from Poulsen's original concepts, even during his firm's bankruptcy. Not the least was AT&T, which for reasons not completely published, began delving into magnetic recording in 1930. Bell Telephone Laboratories initiated a major research effort in magnetic tape recording under the direction of Clarence N. Hickman. By 1931, prototypes designs were made for a steel tape telephone answering machine, a central-office message announcer, an endless loop voice-training machine, and a portable, reel-to-reel recorder for general purpose sound recording. None were said to enter production except for the voice trainer, which failed in the marketplace. AT&T's official policy on telephone recording devices was that they would not be allowed on public telephone lines. (4)

The steel tape ramification of magnetic recording seems to have been of particular interest to AT&T. Although their interest in magnetic recording was declared not an AT&T business objective, I personally saw steel tape playback units used in AT&T's overseas radio station for Miami at Fort Lauderdale, Florida. In that use, vertical steel tapes ran in a glass-enclosed cabinet about 6 feet high over flat brass rollers to endlessly play back the message heard by so many on HF radio over the years, This is a test transmission from a station of The American Telephone and Telegraph Company. This station is located near Miami, Florida.

Similar messages emanated from plants near New York and San Francisco for decades. ostensibly from those Telegraphone-like steel tapes. Obviously, by the 1960's, the later developments by Armour (since Marvin Camras' work in 1939), Brush and Ampex interests were mushrooming so as to overshadow any remembrance of the start Poulsen gave to the recording art.

Along the way, however, there was a heinous incident in which Poulsen's conception figured. At the Telefunken radio long wave radio stations built around 1910 at Tuckerton, New Jersey and Sayville, New York, Telegraphones were found useful for first recording Morse radio messages at normal speed, then transmitting them at high speed on the radio link so as to gain throughput on their expensive, gargantuan international radio links to Germany.

It just so happened that by 1915 Telegraphone-originated high speed transmissions raised the curiosity of radio experimenter Charles Adgar in New Jersey when WW1 was still a European war. Adgar, when one day playing back recordings of the US - German link, had the spring wind down on his Edison machine. Messages from Sayville became readable. One of them was a copy of the infamous 'Zimmerman letter', in which the German Foreign Minister encouraged Mexico to attack the United States, to divert attention from the European war. The final straw was the message on May 7, 1915 telling German submarine U-39 to 'get Lucy', ordering the sinking of the passenger ship Lusitania.

On intercepting that message, the US Navy immediately seized the Sayville and Tuckerton plants of Telefunken, ultimately expropriating them after the war. Finally, when GE and Westinghouse joint ventured the Radio Corporation of America, the new RCA was given them as part of reparations for the war. Poulsen, who obviously knew of his machine's involvement in that action, may indeed have felt like our tragic hero, Doctor Frankenstein.

Want to know more? Here are some references and websites with related information:

1. http://www.dmg.co.uk/ibex/museum/25years_a.htm
2. 'Some possible forms of phonograph' by Oberlin Smith, The Electrical World, September 8th 1888.
3. Danish Patent 1,260, Valdemar Poulsen, 1898.
   http://www.cinemedia.net/SFCV-RMIT-Annex/rnaughton/POULSEN_BIO.html
4. http://www.rci.rutgers.edu/~dmorton/mrchrono.html
   http://www.asb.com/usr/w2g3zfj/lusit.htm
   http://www.asb.com/usr/w2g3zfj/fliwh1/hiscom.htm

[TELECOM Digest Editor's Note: If you were reading this Digest several years ago, you will recall that Donald Kimberlin was a very frequent contributor; then he just sort of dropped out of sight. I hope that the above means we are going to be hearing from him again on a regular basis. Of course his article above will receive permanent placement in the Telecom Archives at http://telecom-digest.org/archives/history very shortly. Welcome back, Don! PAT]

$Log: vpoulsen.htm,v $
Revision 1.1  1999/07/16 01:31:36  lars
Added article about Valdemar Poulsen