Go to content


The fact that the carbon in a telephone transmitter had to be loose, that is it was only intended to fill portion of the cavity provided for it between the electrodes, was a factor that contributed to packing and poor transmitter performance.

The surfaces of the electrodes were flat in most transmitters and were situated vertically in that the surfaces were parallel with the diaphragm. Gravity kept the carbon properly located between the electrodes as long as the transmitter remained more or less vertical - a position that was assured when the instrument was a wall phone.

The candlestick phone was reasonably satisfactory so long as the user did not go in for gymnastics. Most users did as they were surposed to do however, if the phone was located in the bedroom and used while lying down, then a most severe strain was placed on the performance of the transmitter and many failed completely making speech impossible.

It is interesting to note that, for many years, the Bell System in the USA, under chief engineer John J. Carty, resisted moves to produce handset type phones for the apparent reason that the transmitters were not up to it.

Kellogg, in one of it's publicity bulletins, claimed that it was the first manufacturer to recognise the popular use of the handset had created a new problem. The transmitter is never stationary or even nearly so. It may be used at any angle, in any position. The only answer was to develop a transmitter that was fully non-positional and which was rugged enough to stand the rougher usage.

Kellogg, led by engineer George R. Eaton, examined the conventional transmitter and noted, (as per figure 1), that the carbon fell away from one or the other electrode when the transmitter was in a horizontal position.

Enter the Kellogg Non-Positional Transmitter!

They then tried partly filling a hollow ball with carbon and rolling it up, down or sideways and found that the carbon always contacts the same amount of the inner wall (figure 2).

The hollow ball idea was then taken a step further by placing a smaller ball within the larger. The side walls of each ball form the two electrodes and the space between the balls is partly filled with carbon. When the ball is rolled (figure 3) the carbon is consistantly in contact with the same amount of wall space of each electrode.

The ball within a ball concept is taken one step further as the balls are cut in half and the surface of the inner ball has been slightly increased (figure 4).

This concept was developed into the production transmitter and as can be seen (by figure 5), there is good carbon contact with both electrodes in any position. Following exhaustive "proving ground" tests, this transmitter was formally introduced into Kellogg telephones in 1933 with immediate success, being used first as standard in the "Masterphone" range on instruments.

Figure 6 shows a cut-away diagram of the transmitter showing all the main component parts.

From article by Charles A. Pleasance "Old Telephones" and from copies of Kellogg Bulletin No. 209, both reproduced in TCI newsletter "Singing Wires", June 15th, 1991. This article originally printed in the ATCS Newsletter, September, 1992.

Back to content | Back to main menu