HIGH FREQUENCY TRANSISTOR STRUCTURE

摘要

1,271,348. Semi-conductor devices. ITT INDUSTRIES Inc. 10 Sept., 1970 [16 Sept., 1969], No. 43281/70. Heading H1K. A transistor (Fig. 2) for high frequencies comprises a N-conductivity collector substrate 3 in which is formed a P-conductivity base 4 and an emitter of N-conductivity within the base; the PN junctions 3, 4 and 4, 5 extending to the surface; fabricated by conventional photolitho masking and diffusion methods. A thermally grown silicon oxide or R.F. glow discharge grown silicon nitride insulant layer 6 overlies the surface and is apertured to expose the emitter and base regions, in which electrodes 7, 8 are formed by conventional photolitho, masking and evaporation or sputtering methods. The effect of the lead inductance common to the base emitter circuit (Fig. 1, not shown) and external to the emitter bonding pad upon the H.F. gain of the transistor is reduced by an impedance transforming film transformer (Fig. 3) wherein primary 9 has n turns connected to substrate bonding pads for terminals A, B and a single turn secondary directly connected to base and emitter bonding pads. The input resistance R in of the transistor is reflected as n2R in to inputs A, B and the inductance 2 external to the emitter bonding pad is in series with output C, D and no longer common to the input and output circuits. The lower turns 12a of the primary 9 are evaporated from A1, Pt or Cu or printed through a mask over layer 6 between the emitter and base electrodes, and lower portion 13a of the core is formed by sputtering or sintering from high-permeability high-dielectric-constant highresistivity material, e.g. zinc manganese ferrite, and upper primary turns 12b are evaporated or printed over the lower core portion contiguous with the lower primary turns, whose ends extend to substrate bonding pads 14, 15 (Fig. 4). An insulant layer 30 is deposited conventionally from silicon oxide or nitride over the upper primary turns on which a single turn secondary 10 of A1 or Cu is evaporated between emitter 7 and base 8. The upper portion 13b of the core is deposited thereon by sputtering or sintering to surround the stack contiguously with the lower portion 13a to form a closed flux path. In a modification (Fig. 5, not shown) the secondary overlies the lower core and the primary is formed over the upper core. In a further modification a thin film capacitor is inserted between emitter and secondary (Figs. 6, 7, not shown) by evaporating a Pt or A1 electrode over the emitter electrode and insulant layer, over which a tantalum oxide, silicon oxide, or aluminium oxide layer is respectively sputtered and over which the secondary electrode extends. In a further modification impedance transformation is obtained by a LC network coupled to the transistor (Figs. 8 to 13, not shown) wherein a thin film capacitor having a first electrode, dielectric layer and second electrode is formed by conventional methods over the emitter electrode; the second electrode of Al, Cu or Pt extending over the insulating layer to overlap and attach the base electrode; the length of the extension defining the inductance which is adjusted by apertures of the extended portion of the electrode. A mathematical analysis is given.