MOSFETs

Figure Bl.22.4. Differential IR absorption spectra from a metal-oxide silicon field-effect transistor (MOSFET) as a fiinction of gate voltage (or inversion layer density, n, which is the parameter reported in the figure). Clear peaks are seen in these spectra for the 0-1, 0-2 and 0-3 inter-electric-field subband transitions that develop for charge carriers when confined to a narrow (<100 A) region near the oxide-semiconductor interface. The inset shows a schematic representation of the attenuated total reflection (ATR) arrangement used in these experiments. These data provide an example of the use of ATR IR spectroscopy for the probing of electronic states in semiconductor surfaces [44]-...

MOSFET. See Metal oxide semiconductor field-effect transistor. 

Semiconductor devices ate affected by three kinds of noise. Thermal or Johnson noise is a consequence of the equihbtium between a resistance and its surrounding radiation field. It results in a mean-square noise voltage which is proportional to resistance and temperature. Shot noise, which is the principal noise component in most semiconductor devices, is caused by the random passage of individual electrons through a semiconductor junction. Thermal and shot noise ate both called white noise since their noise power is frequency-independent at low and intermediate frequencies. This is unlike flicker or ///noise which is most troublesome at lower frequencies because its noise power is approximately proportional to /// In MOSFETs there is a strong correlation between ///noise and the charging and discharging of surface states or traps. Nevertheless, the universal nature of ///noise in various materials and at phase transitions is not well understood. 

The excellence of a properly formed Si02—Si interface and the difficulty of passivating other semiconductor surfaces has been one of the most important factors in the development of the worldwide market for siUcon-based semiconductors. MOSFETs are typically produced on (100) siUcon surfaces. Fewer surface states appear at this Si—Si02 interface, which has the fewest broken bonds. A widely used model for the thermal oxidation of sihcon has been developed (31). Nevertheless, despite many years of extensive research, the Si—Si02 interface is not yet fully understood. 

In addition to its use as a rectifier, the p—n junction (26) is the fundamental building block for bipolar, junction EFT (fFET), and MOSFET transistors. A thorough understanding of p—n junctions explains much of transistor behavior. The theory (5) of the p—n junction and its role in bipolar transistors was presented within a year of the discovery of the point-contact transistor. 

The MOSEET has three regions of operation. The cutoff region occurs for V g < Up. In this region, the drain-to-source current is the reverse saturation current of the back-to-back source and drain junctions. This leakage current is small but nonzero and allows charge to leak off capacitors which are isolated by cutoff MOSFETs. Because this is how bits are stored in dynamic memory (DRAM) ceUs, DRAMs must be regularly refreshed to retain their memory. 

Fig. 13. (a) The CMOS inverter circuit. The FET circuit symbols emphasize that MOSFETs are actually four-terminal devices in which the / -substrate is connected to for the PFET and the -substrate is connected to the ground for the NFET. Note the conventions on drain location for the PFET and NFET. (b) Corresponding cross-sectional view roughly to scale for a 2-p.m CMOS process, where Hrepresents siUcon, Si02, polysiUcon, and ... 

New Materials Technology. The unique chemical compatibihty of Si02 with sihcon and aluminum has been a significant factor in the dominance of siUcon-based semiconductor technology for MOSFETs in particular and integrated circuits in general. Two enhancements of conventional bipolar or MOS processes have been studied siUcon on insulator (SOI) and SiGe. An alternative material of importance is SiC. 

Diamond and Refractory Ceramic Semiconductors. Ceramic thin films of diamond, sihcon carbide, and other refractory semiconductors (qv), eg, cubic BN and BP and GaN and GaAlN, are of interest because of the special combination of thermal, mechanical, and electronic properties (see Refractories). The majority of the research effort has focused on SiC and diamond, because these materials have much greater figures of merit for transistor power and frequency performance than Si, GaAs, and InP (13). Compared to typical semiconductors such as Si and GaAs, these materials also offer the possibiUty of device operation at considerably higher temperatures. For example, operation of a siUcon carbide MOSFET at temperatures above 900 K has been demonstrated. These devices have not yet been commercialized, however. 

EXTENDING SURFACE ANALYSIS OF NEW MOSFETS BASED ON NANO AND MKM SCALED CDW PHOSPHATE AND OXIDE BRONZES... 

The development of this hybrid combination in a bipolar transistor has greatly enhanced the application of power transistors in the field of power conversion and variable-speed drives. It possesses the qualities of both the power bipolar transistor (BJT) and the power MOSFET. Like a power MOSFET, it is a voltagc-eontrollcd switching device... 

The latest in the field of static devices are MOS-controlled thyristors (MCTs), which are a hybrid of MOSFETs and thyristors. There is yet another device developed in this field, i.e. insulated gate-controlled thyristors (IGCTs). Implementation of these devices in the field of static drives is in the offing. 

Another eonsideration is the type of pass unit to be used. From a headroom loss standpoint, it makes absolutely no differenee whether a bipolar power transistor or a power MOSFET is used. The differenee eomes in the drive eireuitry. If the headroom voltage is high, the eontroller (usually a ground-oriented eireuit) must pull eurrent from the input or output voltage to ground. For a single bipolar pass transistor this eurrent is... 

This drive loss ean beeome signifieant. A driver transistor ean be added to the pass transistor to inerease the effeetive gain of the pass unit and thus deerease the drive eurrent, or a power MOSFET ean be used as a pass unit that uses magnitudes less de drive eurrent than the bipolar power transistor. Unfortunately, the MOSFET requires up to 10 VDC to drive the gate. This ean drasti-eally inerease the dropout voltage. In the vast majority of linear regulator applieations, there is little differenee in operation between a buffered pass unit and a MOSFET insofar as effieieney is eoneerned. Bipolar transistors are mueh less expensive than power MOSFET and have less propensity to oseillate. 

Bipolar Power Switch MOSFET Power Switch Rectifier(s) ... 

Figure 3-23 Examples of transformer-coupled base and gate drives (a) single MOSFET drive circuit (b) dual MOSEET drive.

For those applications where high efficiency is important, synchronous rectification may be used on the higher current (power) outputs. Synchronous rectifier circuits are much more complicated than the passive 2-leaded rectifier circuits. These are power MOSFE B, which are utilized in the reverse conduction direction where the anti-parallel intrinsic diode conducts. The MOSFET is turned on whenever the rectifier is required to conduct, thus reducing the forward voltage drop to less than O.f V. Synchronous rectifiers can be used only when the diode current flows in the forward direction, that is in continuousmode forward converters. 

A small Schottky rectifier with a current rating of about 20 to 30 percent of the MOSFET current rating (/d) is placed in parallel with the MOSFET s intrinsic P-N diode. The parallel schottky diode is used to prevent the MOSFET s intrinsic P-N diode from conducting. If it were allowed to conduct, it would exhibit both a higher forward voltage drop and its reverse recovery characteristic. Both can degrade its efficiency of the supply by one to two percent. 

---