Page 33 - A Practical Guide from Design Planning to Manufacturing
P. 33

The Evolution of the Microprocessor  9

          Under Shockley’s direction, Morgan Sparks built the first working
        junction transistor in 1949; by 1950, Sparks and Gordon Teal had vastly
        improved their techniques. For the previous 2 years, Teal had been advo-
        cating making semiconductor devices out of single crystals. The “crystals”
        used at the time were really just amalgams of many crystals all with dif-
        ferent orientations and many noncrystalline defects in between. Teal
        built an apparatus to grow single crystals semiconductors without these
        defects using the ideas of the Polish scientist J. Czochralski.
          Using Czochralski’s method, a very small seed crystal is dipped into a
        container of molten semiconductor and very slowly pulled out. As the
        crystal is pulled out, atoms from the melt freeze onto its surface, grad-
        ually growing the crystal. Each atom tends to freeze into proper place in
        the crystal lattice producing a large defect-free crystal. Also important
        is that impurities in the melt tend to stay in the melt. The crystal drawn
        out is purer than the starting material. By repeatedly drawing crystals
        and then melting them down to be drawn again, Teal achieved purity
        levels vastly better than anything Shockley had worked with thus far.
        In later years, the cylindrical ingots pulled from the melt would be sliced
        into the round silicon wafers used today. Using these new techniques,
        Sparks and Teal created the junction transistor Shockley had imagined.
          Rather than two diodes side by side, Shockley imagined a sandwich
        of three semiconductor layers alternating N-type, P-type, and N-type as
        shown in Fig. 1-4. The emitter at the bottom injects electrons into the
        base in the center. These electrons diffuse across the base to be captured
        by the collector at the top. The voltage on the base controls the injec-
        tion of electrons by the emitter. Because this device contains two P-N


                                    +                    +



                − − −             − − −                − − −
                                                        − −
                                   − −
          N-type  − −             − − −                − − −
        collector − − −           −  − −                − −
                 − −
                                     − −
                 +  +                                   +  +
          P-type            +      +  +         −
           base  +  +             − +  − −              +  +
                                      +
          N-type − − −            − − −                − − −
                                   − −
                 − −
         emitter  − − −           − − −   Current       − −   No
                                   − −
                 − −               −−     flows        − − −  current
                                                        − −
                                    −                    −
        Figure 1-4 Junction transistor.
   28   29   30   31   32   33   34   35   36   37   38