[ Pobierz całość w formacie PDF ]
.Three days later Brattain and Gibney had a third patentabledevice, a field effect amplifier using an electrolyte.On the 23rd,Bardeen produced still another field effect device, he credited to the Shockley effect. (It wouldn t work then either, but it set the stage forlater devices: a creative failure, Shockley called it).Now they enteredthe phase Shockley called Respect For The Scientific Aspects OfPractical Problems.On 21 November they produced a field effect in the liquid electrolyte. It was just a few cycles a second, not worth a damn, Brattain said.103 Ina seven-page logbook entry dated Saturday 22 November, Bardeennoted that the experiments show definitely that it is possible to intro-duce an electrode or grid to control the flow of current in asemiconductor.The next day, Sunday, Bardeen turned his attention to the possibilityof an amplifier.On Monday, Brattain witnessed new disclosures towardthat end.Shockley came into the lab the day before Thanksgiving towitness the entry as well.They had an amplifier that boosted currentabout 10%; but that was not enough to be useful and the frequency wastoo low.* An electrolyte is a conductor in which current is carried by ions (atoms that haveeither lost or gained an electron) rather than by free electrons. I THINK WE BETTER CALL SHOCKLEY 105On Thursday 4 December, Brattain performed three experiments,bolstering some earlier ideas.The following Tuesday they switched to n-type germanium.Accidentally washing away the electrolyte, they noted,revealed an anodized surface left behind that conducted better than theoriginal electrolyte.It meant that the gold plate deposited on the surfacewas directly touching the germanium; not insulated from it by a germa-nium oxide layer, which is what they were trying to produce anotherlucky accident.On 15 December, Brattain replaced the electrolyte with evaporatedgold in two points close together on the crystal.80 Bardeen believed thata small electric charge would inject holes (the minority carrier) intothe semiconductor surface, greatly increasing the capacity of the crystalto carry current.20 One point served as the grid, the other as themetal plate.They got a slight voltage amplification, but no poweramplification.On the afternoon of 16 December 1947, Bardeen, inspired by Lee deForest s triode audion, suggested that the two points be as close togetheras possible on the germanium crystal.Brattain devised a structure inwhich gold foil was spread over a plastic triangle, which touched a crys-tal of n-type germanium.Using a razor blade he cut a piece of gold foilinto two sections 0.04 centimeters (0.16 inches) apart: one was the grid,the other the plate.The electrodes were attached to those points andheld in place by the plastic triangle.All this required the dexterity of a brain surgeon: The whole devicewas less than half an inch long.The gold-tipped plastic arrowhead pointed down at the surface of thecrystal.Three sets of wires connected to the contraption.One of them aspring made from a paper clip to press the plastic down on the surface of thecrystal was bent like a coat hanger; another coiled wire, almost bulbshaped, plugged directly into the plastic; and the third, a thin wire connec-tion, went into the crystal.It was a crude, almost preposterous, setup, fartoo ugly to inspire a revolution, and just larger than a shoelace tip.When the germanium is simply sitting there, almost no current passesbetween the two pointed wires.If the third wire introduces a tiny cur-rent, the resistance between the two point contacts virtually disappearsand a much larger current can flow between the wires.The two men stood at the lab table and watched for the first timewhat would eventually be called the transistor effect.The power gainwas 4.5; the voltage went up by a factor of four. Current flowing in the106 I THINK WE BETTER CALL SHOCKLEYImage not availableFigure 13 The first point-contact transistor.forward direction from one contact influenced the current flowing inthe reverse direction in a neighboring contact in such a way as to pro-duce voltage amplification, Bardeen wrote laconically later.78 Theyrepeated the experiment and it worked every time.They could adjustthe power and alter the gain.Even then, Bardeen and Brattain knew what they had accomplishedwas important.That night, Brattain, unable to restrain himself, told hiscar pool buddies he had participated in the most important experimentof his life.(The next day he called them and swore them to secrecy.)Bardeen and Brattain talked on the phone later that evening.Theyhad no doubts they had succeeded in controlling the current in a semi-conductor in a useful, practical way.Brattain knew what they needed todo immediately. I think we better call Shockley.CHAPTER 6 There s enough gloryin this for everybodyBardeen and Brattain had one immediate minor but annoying problem:they didn t know what to call the uncomely contraption on their bench.Brattain, of the two the least able to keep calm, couldn t help blab whathe had done to his friends, despite the corporate secrecy that sur-rounded the work.John Pierce, his friend and colleague, was walking byBrattain s office soon after the invention, when the physicist invited himin, shutting the door behind them.Brattain explained the unnamed device he and Bardeen had con-structed.Pierce immediately guessed they had a replacement forvacuum tubes. I thought right there at the time, if not, within hours, Ithought vacuum tubes had transconductance, transistors would havetransresistance, Pierce recalled. There were resistors and inductors andother solid states, capacitors and tors seemed to occur in all sorts of elec-tronic devices.From transresistance I coined transistors. 12 That s it! said Brattain.22*Transistor it was a point-contact transistor to be specific, becausethe action end of the device was where the points of electrodes con-tacted the crystal
[ Pobierz całość w formacie PDF ]