A glass telephone? a gas-powered phone? a cork phone? Why not?
Please note: this page contains many illustrations and will take a while to load. Please be patient.
Part 1: Carbon Pencil Transmitters
In the earliest days of telephony, Bell and Western Union challenged each other in the U.S. courts over just who had invented the telephone. The legal fight ended in victory for Bell, who received all Western Union's patents as part of the settlement deal. These included Edison's carbon transmitter, the basis for most future transmitters. Bell now had to defend these patents against the many small companies that had sprung up to service the new industry. To avoid the legal problems, inventors worked tirelessly to produce new devices to get around the Bell patents. This essay examines some of them, from the technically interesting to the downright silly.
Transmitters were covered by two main patents - Bell's, which described a permanent magnet; coil and diaphragm assembly, and Edison's , which used lamp black (fine carbon powder) or other substance whose resistance varied with pressure, and a metallic diaphragm to apply that pressure. Most of the alternate transmitters used variations of these, just different enough to avoid patent infringement.
Professor Hughes in Britain had been working on microphones through the 1870s. He may even have invented a workable microphone before either Bell or Edison, but did not publish or patent it. In May 1878 he finally released a practical design. This used a carbon pencil loosely held between two carbon blocks. These were glued to a thin wooden "sounding board". Vibrations of the sounding board varied the resistance between the blocks and the pencil. This altered a voltage passing through the unit. It gave a similar result to the Edison, but without the carbon powder - a small but significant point that allowed Hughes transmitters to be built in Britain without patent problems. It was a fairly efficient transmitter and produced clean speech. It had drawbacks - it hummed and buzzed, especially with new batteries. This was eventually traced to arcing between the carbon contacts at a high enough voltage. At loud voice levels the transmission broke up as the pencil vibrated off its contacts. Although the Hughes was a good microphone under controlled conditions, its use as a telephone transmitter was limited. In spite of this it became the basis for many other transmitters.
Hughes' basic design could be improved by simply adding more carbon pencils, and this path was taken by Frederic Gower in Britain. His designs used up to eight carbon pencils in a star pattern. It worked on the theory that at any time some of the pencils would be making contact, avoiding the breakup that the Hughes was prone to. The multiple carbon pencils lowered the overall resistance. The design was quite sensitive and fairly reliable, although it was expensive and needed careful setting up. Gower built a telephone around it, and the British Post Office adopted it as their standard phone in the 1890s.
A similar transmitter came from Crossley in 1879, who used four pencils in a diamond pattern. He called this a "compound microphone" to distinguish it from Hughes' "simple microphone" and patented it accordingly. There were two electrical paths through the transmitter, so again breakup was greatly reduced.
The drawback of the Crossley and Gower was their size. A large sounding board was needed, and generally it was mounted horizontally, so the phones were quite large. It must have been an odd feeling for the user, talking down to a flat piece of wood on top of a box. For this reason most inventors concentrated on building phones with a vertical diaphragm.
Mr A C Swinton devised an improved transmitter which did away with a diaphragm. A number of carbon pencils were strung on a taut platinum wire through holes at one end of the pencils.The other end of the pencils rested lightly against a horizontal carbon block, and the assembly was mounted in a lead frame. The pencils and wire vibrated with the voice. This device could be "tuned" for quite respectable performance by adjusting the angle of tilt of the frame, and therefore the pressure of the pencils on the block. It was rather fragile and susceptible to temperature change.
A novel design came from Professor Sylvanus Thompson. The mouthpiece led into a tube that bent upwards, somewhat like the Berliner. At the top of the tube, three carbon pencils in a tripod supported a carbon ball. The pressure of the sound waves on the ball provided the variable resistance needed. This style was christened a "Valve Microphone". There was no diaphragm as such so it was thought that this would work around the Hughes and Edison patents. The National Telephone Company, who held the British patents for the Bell, Edison, Crossley, Blake and Hunnings transmitters, disagreed. So did the British courts, They ruled that the carbon ball was a diaphragm, and so infringed the patents. They also ruled out the Swinton transmitter on the same grounds - the frame and wire formed a diaphragm in themselves. The legal situation was confused because other designs were also close to infringement, and some of the patents themselves were open to challenge. The situation in the U.S. was not so confused, as the patents for most practical transmitters were owned by the one company - Bell.
The Swinton design was revived in the late 1890s after some of the critical patents expired . Mr F C Allsop in Britain simplified the design into a compact rugged two-pencil model which his company used in their intercom phones. As can be seen from the illustration, it was well built and very compact . It must have been very sensitive for such a small diaphragm to be efficient.
Ader in France produced a similar design to the Gower and Crossley, with the carbon pencils arranged in two rows of five. It became a standard French transmitter of the period and was used for some years until the Bell patents expired. There are many good photos of Ader telephones on Frederic Niebart's website at http://perso.wanadoo.fr/fredouille/englishposte.htm
An interesting variation was the D'Arsonval transmitter, which had a thin iron sheath over each pencil. A horseshoe magnet behind the pencils could be adjusted by a screw to vary the pressure of the pencils against the blocks. By this means a very sensitive long distance transmitter could be obtained.
DeJongh produced a rather complicated arrangement where the pencils were mounted horizontally on brass pins set into the backboard. The pins sloped down at the front so the pencils rested against carbon blocks glued to the back of the diaphragm. The arrangement was very delicate and expensive and needed to be isolated from the phone case by rubber mounts. The expense of construction made it less practical than other models , although it was also refined by Allsop in the late 1890s.
It also turned up in the United States, where carbon balls were substituted for the pencils. They were mounted in cutouts in a carbon back block and rested lightly against the diaphragm. The design was possibly from Holzer-Cabot but this is unconfirmed. The illustration shows an "Acousticon" transmitter from the Dictograph company using this principle. Its construction is quite simple, but the machining would need to be rather precise.
Dictograph's transmitters achieved a high reputation for their sensitivity, and were built into some of the earliest desktop handsfree intercom systems. Excellent photos and technical details are available at Mike Schultz's website at http://www.uv201.com/Microphone_Pages/dictograph.htm
The Johnson Telephone Company in Britain improved the original Hughes model by using two carbon pencils and including a resistance coil in the circuit. This ensured that the circuit would not be broken even when the carbon pencils were quite active. It also eliminated the buzzing and humming that the Hughes was prone to. A thin sheet of mica acted as waterproofing on the diaphragm. It was reasonably sensitive on short lines, and inexpensive to build, but was not introduced in time to make large sales . It also solved one other problem. The coil allowed a standard resistance to be presented to the line, a worthwhile improvement in view of the wide range of transmitters now in use.
The German Post Office adopted a model designed by Mix and Genest that seemed to combine the best features of all of these. It used a carbon block glued to a wooden diaphragm, with three carbon pencils lightly held against it by spring pressure. The springs were damped by felt pads or hog-bristle brushes , and this gave a quiet, stable transmitter free of buzzes and rattles.
This was about the peak of carbon pencil technology. Once Bell's patents expired , manufacturers returned to the Edison carbon transmitter and worked on improving it into the carbon granule transmitter that was to last through most of the twentieth century.
Part 2 : Carbon Button Transmitters
Edison's patent on a carbon-powder transmitter effectively locked other manufacturers out of this style for seventeen years until his patent expired. It was the most efficient transmitter, but other types were explored in the interim. One alternate style was the single-carbon-contact transmitter.
In 1878, Francis Blake produced a new transmitter for the American Bell Telephone Company. His design mounted a small carbon button between an iron diaphragm and a spring-mounted platinum contact. The diaphragm was damped by springs, and an induction coil was used which also reduced dropouts. American Bell adopted the Blake because they desperately needed a better transmitter. They were short of cash due to their costly legal battle with Western Union. Blake offered his transmitter in exchange for shares in the company and his offer was gratefully accepted.
The transmitter was delicate and required careful adjustment, but it worked well on short lines. It was not particularly sensitive, because of the single contact, but at least the coil prevented total dropouts. Its size required that it be mounted in a separate box, and this gave rise to the large three-box phones of this period. In spite of these shortcomings it was produced in large numbers, simply because Bell could not afford to develop anything else.
Gent's transmitter was a much simpler idea. It had a wooden diaphragm with a carbon button glued to it. The rear contacts were carbon buttons mounted on weighted pivot arms. Adjustment of the weights varied the pressure on the buttons. The wooden diaphragm worked around Edison's patents. Although reported to give very good results, it would have been delicate and subject to dropouts. It quickly faded into history. It is interesting to note that Crossley and others used much the same arrangement with carbon pencils instead of carbon buttons, and achieved some commercial success.
Other inventors tried to improve on the basic principle. The obscure Locht-Labye transmitter used a similar weighted arm to make the contacts, but its diaphragm was a rectangular sheet of cork mounted on two flexible arms at the top. It is shown in the Pantelephone at left. It was otherwise unsecured. It would be sensitive to bumps and knocks, but was fairly sensitive to voice. It probably it deserved its obscurity. So did Freeman's transmitter. The diaphragm pressed on a two-arm lever, with carbon buttons at the ends. A complex induction coil sorted out the current so the output from the contacts at one end of the lever reinforced the output from the other end. Sometimes inventors just don't know when to stop.
f one diaphragm is good, two must be better. Burnley's transmitter split the mouthpiece into two tubes which fed sound to two diaphragms. Each had a carbon button glued to it and the two buttons rested against each other to provide the necessary carbon contacts. Technically it probably made sense, but it was bulky.
These last three transmitters showed the main weakness in the single-contact design - they were usually not very sensitive, and multiple diaphragms or a large single one were usually needed. Only the Blake made it into large scale production, for the reasons mentioned. Even it was soon replaced as better transmitters became available. L M Ericsson, for instance, produced a replacement carbon granule unit which neatly fitted into the mouthpiece cavity of a Blake box, and many Bell/Western Electric phones were upgraded this way. As a result, Blakes are now rare.
All these transmitters disappeared as the new HunningsCone carbon-granule transmitter established its superiority. They are now interesting, if rare, technical oddities.
Part 3: Carbon Granule Transmitters
Early research on carbon-pencil transmitters showed that an improvement was made by adding more carbon-to-carbon contacts. The size of the pencil was less important. If the pencil was replaced by carbon granules, more contacts could be fitted into the same space. Unfortunately Edison's patent on carbon powder transmitters covered this pretty well, so for some years any developments in this area could only be used by the Bell company who owned the patent.
In 1878 (the same year that Hughes and Blake patented their transmitters) the Reverend Henry Hunning, a Yorkshire clergyman, patented an improved Edison design. His patent described a transmitter with carbon held between two metal diaphragms. The gap in the middle was filled with fine hard coal dust rather than the compressed lamp black powder used by Edison. This made the transmitter more sensitive. The Hunning had one big problem - the carbon packed down to the bottom, and the transmitter lost sensitivity. This "packing" turned out to be the main problem to be overcome. Western Electric improved the design by using a carbon backplate and changing to small anthracite granules. This variant was known as the Delville transmitter. They built the Hunning varieties for Bell for many years, since it offered improvements on the Blake transmitter they currently used. Their solution to the packing problem was to mount it on the front of the case so it could be rotated occasionally to free up the granules.
Berliner's Universal Transmitter was similar to Hunning's, but it avoided the packing problem by mounting the transmitter case horizontally and feeding the sound into the base through a bent horn. The carbon block at the top could be adjusted by screw to "squeeze" the granules, reducing the packing problem. Three concentric grooves were machined into the inside of the carbon diaphragm. This gave a greater electrical contact area, and helped the diaphragm to shake up the granules and keep them loose. The Berliner patent was held by the Bell company, but the transmitter never went into production by them. It was still being developed since being patented in 1877, but was overtaken by the Hunning transmitter.
Eventually Berliner patented the transmitter in Europe and it went into production there.
A similar arrangement was used in Thornberry's transmitter, shown at left , with the addition of two brass plates inside the granule cavity. The greater contact area increased the sensitivity and this transmitter was used successfully in the U.S.A for long-distance work.
Berthon in France improved the Berliner by mounting spherical carbon granules between two carbon plates separated by a rubber ring. Adjustment, as in the Berliner, was by screwing the two plates closer together . One plate had concentric grooves machined into it. These and the carbon spheres greatly reduced packing and made the transmitter much less clumsy. Adjustment was critical and the carbon balls were expensive to make and rather fragile, so the design never went into wide use. The idea was good, though, and it was improved on later in the Solid Back transmitter. A number of similar British transmitters were marketed with small carbon balls instead of granules. They were known as "Manchester Shot" transmitters.
As an example of how to over-engineer a good idea, Boudet's transmitter is unsurpassed. A mouthpiece held a diaphragm with a small copper button glued to it. This pressed onto six hard carbon balls about one centimetre in diameter, enclosed in a glass tube. At the other end of the tube, another copper block was held by spring pressure against the balls. A screw adjusted the tension of the assembly. A glass telephone - what a great idea! Not surprisingly, none seem to have survived.
L M Ericssons by 1892 had developed a very efficient carbon granule microphone which was in direct conflict with the Edison patents. They got away with it because Bell and Edison had not patented their transmitters in Sweden. The Ericsson model filled the gap between the diaphragm and the back plate with a star-shaped felt pad. Star-shaped slots machined into the back carbon block increased the contact area, making the transmitter very sensitive for its size. This design both retained the carbon granules and damped the diaphragm. The gap was adjustable by a screw. This stopped the granules packing down and produced a well-damped, compact, rugged transmitter. It was so good that they were able to use it in the first handset phones. It was a very advanced design for its time, but its use was limited to the Scandinavian countries until the Bell patents expired late in the 1890s.
The final development was produced by Western Electric. It was called the Hunning'sCone-Deckert Transmitter. The back carbon block was machined with criss-cross grooves, leaving the surface covered with tiny pyramids or cones. Each cone had a tuft of silk attached, which pressed lightly against the front diaphragm. This damped the diaphragm, and reduced packing as the carbon grains were trapped between the fibres and cones. Although more expensive, it was solid and reliable, and quickly replaced most of the other transmitters in use. It survived until the introduction of the solid back transmitter.
It is particularly interesting to note that all these developments took only about ten years to the mid 1890s.
Part 4: The Strange Ones
The principle of varying an electric current by passing it through some compressible carbon medium was well understood, and most alternative transmitters to the Bell designs used some variation of this principle. But there is always an inventor somewhere who wants to do it the hard way. Some of these people should have known better.
Amos Dolbear almost got it right in 1880 with a Condenser Receiver. It consisted of two thin metal plates, one fixed, isolated by a narrow air gap, with a high voltage applied between them. By varying the voltage, the distance between the plates varied and sound could be produced. In a neat bit of reverse engineering, this principle was used to produce a highly workable condenser microphone in the 1890s. The movable plate became the diaphragm, and the voltage varied with the distance between the plates. It turned out to be impractical for telephone use because of the high voltages needed. It was mainly used for early radio work, where high voltages were readily available and were needed for the early unamplified radio transmitters. In the 1970s the idea was revived to produce high fidelity ribbon loudspeakers by Wharfedale in Britain, among others.
Bell and Tainter's Photophone from 1879 was a scientific curiosity more than a practical transmitter, but it worked. A beam of light was directed onto a polished diaphragm, which reflected the beam back to a selenium button. Selenium varies its resistance in the presence of light. The effect of sound waves on the diaphragm caused the strength of the light beam falling on the selenium to vary, giving a varying electrical output. It was practical, but would have been too expensive and delicate to be a commercial proposition.
Tainter, not a man to leave a bad idea alone, then went on to produce the Radiophone. In construction it was similar to the Photophone, but used focussed heat from a gas flame instead of light, and lamp-black to absorb the heat instead of selenium. Carbon's resistance decreases as it heats up, generating a variable resistance in much the same way as the Photophone. It was thoroughly ignored, as a gas-powered telephone deserved to be.
A simple solution to making a more powerful carbon transmitter was to increase the voltage through it. A number of experimenters tried this, but at more than three volts the carbon granules tended to heat up and fuse together or jerk apart violently as they expanded. The arcing and burning generated a lot of noise on the line, and eventually would render the transmitter useless. Some experimenters tried air cooling in the carbon chamber, but a more successful version was Prof. R A Fessenden's Trough Carbon Transmitter. This circulated water through a jacket around the carbon, and allowed up to 15 amps to be carried through the transmitter. Although impractical for telephony it also found its niche in the early radio transmitters where such high currents could be used.
Majorana's Hydraulic Transmitter combined the condenser microphone with water cooling to produce what sounds like the most dangerous transmitter ever built. A jet of pressurised acid water was passed between two platinum plates, one of which acted as the diaphragm. The varying distance between the plates varied the electrical current carried through the water. The acid water acted as a conductor , coolant, and a way of removing air bubbles caused by electrolysis. Again, it could carry massive currents, so it also finally found its application in radio telephony. Unamplified it could reputedly carry a voice transmission over 500 kilometres. I cannot imagine how a microphone carrying high voltage and current and filled with acid under pressure would be regarded in Occupational Safety circles these days. It was finally made redundant by De Forrest's invention of the amplifying valve. This allowed the signal to be built up at the receiving end, removing the need for such high power at the transmitter.
These transmitters were simply impractical compared with the cheaper, more reliable carbon transmitters. In spite of this, they are noteworthy for the way they exploited obscure scientific principles and sometimes found their own niche.
Telephony - McMeen & Miller, 1923
The Practical Telephone Handbook- J Poole, 1912
Telephones - Their Construction & Fitting , F C Allsop 1917
A Manual Of Telephony - Preece & Stubbs , 1893
Telephone - The First 100 Years -
John Brooks 1975
The Practical Telephone Handbook - J Poole 1891
Electric Bells and Telephones - B E Jones, Amateur Mechanic & Work Handbook, 1926
Popular Guide to Commercial and Domestic Telephony - M Byng andF G Bell1898