Tip transistor

from Wikipedia, the free encyclopedia
Replica of the first transistor by Shockley, Bardeen and Brattain from 1947/48 in the Nixdorf Museum

The tip transistor was considered to be the first practically implemented bipolar transistor and is regarded as a pioneer of modern semiconductor technology . However, later research showed that it was actually a thyristor . The basics of modern semiconductor technology had not yet been worked out at the time of its development and all work was largely based on trial and error. The circuit symbol of the bipolar transistor is still reminiscent of the structure of the tip transistor.

development

In March 1942 Karl Lark-Horovitz's research team at the University of Purdue received an order from the military research authority OSRD to develop improved crystal detectors for use in the radar devices that were currently being developed. The company Bell Laboratories was involved in military radar research. To prepare for the competition in the news sector after the end of the war, Mervin Joe Kelly ( Head of Research at Bell) founded, among other things, a solid-state physics research group in July 1945 , in which the semiconductor group was headed by William Shockley . The scientists in this group were: Walter Brattain , Gerald Pearson , Hilbert Moore, Robert Gibney. Later joined John Bardeen added. This group tried to work out a theory for the effects in semiconductors with the aim of developing a field effect transistor based on the model of the patent by Julius Lilienfeld (1925). In doing so, they built on the experience of the research team of the Purdue Group, whose tip diodes were already mass-produced.

Typical data of a BTL (Bell Telephone Laboratory) tip transistor "M1734" from 1954
parameter value
lifespan 70,000 hours
vibration 100  g max.
Noise 54 dB at 1 kHz
Current amplification in basic circuit 0.2 ... 3
Power amplification 18 dB
Collector-emitter operating voltage U CE 100 V
Collector current I C 40 mA
Power loss P v 120 mW
Cutoff frequency f gα 50 MHz

In lengthy experiments with electrodes in the immediate vicinity of the point contact of these diodes , Brattain and Bardeen found that the current through the diode could be influenced and that it was amplified at the same time . On December 16, 1947, a demonstrable functional model, with a power gain of 100 times, was put into operation and presented to the company management on December 23, 1947 in the form of a microphone amplifier . The discovery was kept secret until June 22, 1948, after which the technical staff and the military were inaugurated. Surprisingly, the invention was released from secrecy by the military, so that it could be published in the Physical Review on June 25, 1948 . The invention was reported in the daily press on June 30, 1948.

Large-scale production began in 1951 at the Bell company Western Electric. In 1950, also at Bell, the planar transistor was developed which, thanks to its significantly higher reliability, displaced the tip transistor from the lower frequency range up to approx. 5 MHz. Until about 1956 the tip transistor defended the frequency range up to about f T  = 100 MHz, after which it was quickly replaced by the more modern drift transistor.

The table shows the technical data of a typical tip transistor from the 1950s. In view of the technical development in the following decades, the characteristics of these transistors are of course not comparable with modern components. Because of the new possibilities, especially with regard to miniaturization, the entire production was bought up by the military in the early years.

Structure and properties

Sectional drawing of pnp tip transistor

The doped crystal is soldered without a barrier layer on a metallic base plate, which serves as the base connection . Two 15 μm thick phosphor bronze wires are pressed onto the crystal . The distance between these tips is approx. 30 μm. The system is then cast in the metal assembly tube through the assembly peephole and the wire tip, which is to form the collector connection, is welded to the crystal with a short current surge. The resulting layer structure forms a four-layer diode with a control connection ( thyristor ), which results in a current gain greater than one in the basic circuit . This was not understood at the time and repeatedly caused problems during amplifier development, as it could lead to unwanted switching of the transistor.

Furthermore, the Schottky contact of the emitter connection was extremely sensitive to overloads and often led to failure in high-frequency pre-stages due to electrostatic discharges. Moisture problems were also a major cause of failure due to the plastic potting compound caused by the structure. For example, the hearing aid manufacturer Zenith Radio Corporation had to recall all hearing aids in April 1953 and re-equip them with miniature tubes instead of transistors.

Notes on the collector zone

Since the tip transistor was quickly being replaced by improved transistor types (primarily by junction transistors such as the drawn , alloy and diffusion transistor ), its research has never been satisfactorily completed. The technology that has become known for building layers on the collector is as follows: The single crystal is melted under the tip by strong current surges and then recrystallized again. However, there are defects in the crystal lattice that act like p-doping and form a barrier layer to the single crystal. This also reduces the thickness of the base zone. As a result of the heating, impurities diffuse from the wire tip at the same time and there dop a new, very thin n-zone. This explains the current gain> 1 in the base circuit, which a 'normal' transistor does not have. The process is difficult to control, which explains the strong scatter and sometimes even leads to the lack of effect (see data above). The emitter does not require any treatment and forms a Schottky contact.

Web links

Commons : Top Transistors  - Collection of Images, Videos, and Audio Files

Footnotes and individual references

  1. ↑ Patent CA272437 : Electric Current Control Mechanism. Inventor: Julius Edgar Lilienfeld (Published July 19, 1927; Registered with the Canadian Patent Office ). ‌
  2. ↑ Patent US1745175 : Method and Apparatus For Controlling Electric Currents. Published January 28, 1930 , inventor: Julius Edgar Lilienfeld. ‌
  3. ^ Crystal Rectifiers . In: Henry C. Torrey, Charles A. Whitmer (Eds.): MIT Radiation Laboratory Series . tape 15 . McGraw-Hill Book Company, New York 1948, pp. 443 .
  4. ↑ Michael Eckert, Helmut Schubert: Crystals, Electrons, Transistors. From the study room to industrial research. Rowohlt Verlag, Hamburg 1986, ISBN 3-499-17725-0 , pp. 174ff.
  5. ↑ Bell's comment on the original data sheet from 1951: The M1734 units average about 20 megacycles. and These frequency figures are quite variable at present.
  6. ^ J. Bardeen, WH Brattain: The Transistor . A semi-conductor triode. In: Physical Review . tape 74 , 1948, pp. 230-231 .
  7. ^ News of the Radio. In: New York Times. July 1, 1948, p. 46.
  8. ^ Rudolf Rost, Hans Martin Ernst: Kristalloden-Technik. 2nd Edition. Verlag Wilhelm Ernst & Sohn, Berlin 1956, p. 333 (section: GTA2 / 100MHz ).
  9. ↑ Heinz Richter : Transistor Practice. Franckh'sche Verlagsbuchhandlung, Stuttgart 1959, supplementary sheet to p. 72, OC170 / 171 u. OC614 / 615.
  10. ↑ Karl Otto, Horst Müller: Flat Transistors. VEB Verlag Technik, Berlin 1960, p. 32.
  11. ^ JA Becker, JN Shive: The Transistor-A New Semiconductor Amplifier. In: The Transistor-Selected Reference Material. Bell Telephone Laboratories, New York NY, Nov. 15, 1951 (Supplements from Bell Symposium Sept. 17, 1951), p. 104.
  12. ^ The formation of germanium surfaces In: Rudolf Rost, Hans Martin Ernst: Kristalloden-Technik. 2nd Edition. Verlag Wilhelm Ernst & Sohn, Berlin 1956, p. 176.
  13. ↑ Joachim Dosse: The transistor. Verlag R. Oldenbourg, Munich 1957, p. 34.