摘要:

Pulsed and continuous operational curves for MOS power transistors are compared. The latter show pronounced negative differential output conductance of the drain, which is explained in terms of a mathematical model. The potential hazards of this phenomenon are discussed.

DOI:10.1049/ip-i-1.1986.0036

出版商:IEE    

年代:1986

数据来源: IET

摘要:

Results of experimental studies of open-circuit voltage decay (OCVD) in a high resistivity (3000 Ω cm) base (i.e. lightly doped)pn-junction diode are reported. For moderate values of applied junction voltage, a rapid initial decay followed by a slow decay is observed. As the applied junction voltage increases to higher values, the initial drop becomes more rapid, it is followed by a plateau and finally the slow decay. At very high applied voltages, a peak is observed in the decay curve after the initial rapid drop. The final slow decay rate remains the same in all cases. The initial rapid decay and the final slow decay can be explained with the help of the high-injection theory of Schlangenotto and Gerlach. According to this theory, the excess carriers recombine not only in the base (as in the low-injection theory of a base dominated diode) but they migrate to the emitter and recombine there also. Mathematically, this emitter base coupling is described by a parameterβ. The initial rapid decay and the final slow decay can be explained using a constant value ofβas assumed by Schlangenotto and Gerlach. It is, however, necessary to assume thatβdecreases with time to explain the observed peak. Possible physical reasons for the decrease inβare given. The final slow decay rate is used to obtain a value of 0.4 ms for the minority carrier lifetime in the base.

DOI:10.1049/ip-i-1.1986.0038

出版商:IEE    

年代:1986

数据来源: IET

摘要:

In the theory of microwave linear beam tubes, it has hitherto been assumed that the space charge potential for a cylindrical electron beam passing through a concentric, earthed, conducting tunnel broken by interaction gaps can be calculated by supposing that the tunnel is continuous. In this paper, the effects of the gaps are computed for both an electron beam and a disc of charge. It is shown that for a uniform electron beam these effects can be represented by replacing the actual boundary with an equivalent continuous tunnel, the radius of which is termed the effective tunnel radius. An empirical equation has been derived from which the effective tunnel radius can be calculated to an accuracy of better than ±3%. The effective tunnel radius is a function of the tunnel radius, the gap length and the gap separation but is independent of the beam radius. For a disc of charge, it is found that the effective tunnel radius depends on the position of the disc relative to the gaps but that the average value of the effective tunnel radius as the disc moves through the gapped tunnel is equal to the value computed for a uniform electron beam. This average value of the effective tunnel radius is independent of the disc thickness.

DOI:10.1049/ip-i-1.1986.0039

出版商:IEE    

年代:1986

数据来源: IET

摘要:

A new MISS switching device structure was designed and fabricated, which consists of Al/thin thermal nitride/pn+(buried)-p-Si layers and is isolated by diffusingn-well to the buriedn-layer. Furthermore, an+-shield ring which confines the carrier flow to the MIS interface and ap+-injector which injects carriers fromptonlayers were successfully implemented. The device reveals that switching and holding voltages (VsandVH) decrease with increasing area of the MIS junctionAjand with decreasingnp-junction areaAjand thin dielectric thicknessdi, respectively. The holding and switching currents (IHandIs) increase with increasingAJ/Aibut decrease with increasing dielectric thicknessdi. The fringing effect is minimised due to the confined structure.

DOI:10.1049/ip-i-1.1986.0040

出版商:IEE    

年代:1986

数据来源: IET

摘要:

MOTA: a new NMOS and CMOS timing simulator, is presented. Basically, it employs a one sweep nonlinear Gauss-Seidal relaxation technique to decouple node equations, and this results in a linear performance on the computation time over the number of the gates of the circuit. It has three distinct features: (a) it provides a ‘SUBCIRCUIT’ capability to simulate tightly-coupled circuit blocks. This solves the inaccuracy and the instability problems which are usually encountered in existing timing simulators, (b) it employs a physical table model for MOS devices with only 250 storage points, and (c) it utilises a simple variable time step control scheme and internal and external bypass schemes to increase the simulation speed. The run examples show that it is approximately 60 times faster than SPICE2G-5 while giving comparable precision.

DOI:10.1049/ip-i-1.1986.0041

出版商:IEE    

年代:1986

数据来源: IET