Advanced High Voltage Power Device Concepts

Power devices are required for applications that operate over a broad spectrum of power levels as shown in Fig. 1.1 [1]. Based upon this figure, the applications can be broken down into several categories. The first category is applications that require low operating current (typically less than 1 A) levels. These applications, such as display drives, usually require a large number of transistors that must be capable of blocking up to 300 V. The small size of the low-current transistors allows their integration on a single chip with control circuits to provide a cost-effective solution.

As discussed in the textbook [1], the power thyristor was developed as a replacement for the thyratron, a vacuum tube used for power applications prior to the advent of solid-state devices. The simple construction of these structures using P-N junctions enabled commercialization of devices in the 1950s. These devices were found to be attractive from an applications viewpoint because they eliminated the need for the cumbersome filaments required in vacuum tubes and were much more rugged and smaller in size. The power thyristor provides both forward and reverse voltage blocking capability, making it well suited for AC power circuit applications. The device can be triggered from the forward-blocking off-state to the on-state by using a relatively small gate control current. Once triggered into the on-state, the thyristor remains stable in the on-state even without the gate drive current. In addition, the device automatically switches to the reverse-blocking off-state upon reversal of the voltage in an AC circuit. These features greatly simplify the gate control circuit, relative to that required for the power transistor, reducing its cost and size.

The basic structure and operation of the thyristor are discussed briefly in Chap. 2 and in more detail in the textbook [1]. The thyristor contains two coupled bipolar transistors that provide an internal positive feedback mechanism that allows the device to sustain itself in the on-state. Analytical models were provided in Chap. 2 for all the operating modes of the thyristor. These models are applicable to the silicon carbide devices discussed in this chapter. The motivation for the development of thyristors from silicon carbide originates from the high on-state voltage drop and slow switching speed of the high-voltage silicon devices. In Chap. 2, it was demonstrated that even a 10-kV silicon thyristor structure has a relatively high on-state voltage drop (close to 3 V) even when the high-level lifetime in its drift region is 100 μs. In the case of silicon carbide devices, the width of the drift region can be greatly reduced (about ten times) when compared with a silicon device with the same voltage rating. This allows obtaining devices with much faster switching speed.

As discussed in the previous chapter, the thyristor structure contains a set of coupled transistors that provide a regenerative action during the conduction of current in the on-state. These devices are designed for operation in AC circuits where the anode voltage cycles between positive and negative values. The regenerative action is disrupted whenever the anode voltage reverses from positive to negative. The turn-off of the device then occurs with a reverse recovery process to establish blocking voltage capability. Such device structures are not suitable for applications in DC circuits unless expensive commutation circuits [1] are added to reverse the anode voltage polarity. The development of a thyristor structure that can be designed to turn on and turn off current flow under control by a gate signal in a DC circuit was motivated by this need. Such thyristors have been named gate turn-off (GTO) thyristors. The GTO is turned on in the same manner as the thyristor structures described in the previous chapter, while the turn-off for the GTO is accomplished by the application of a large reverse gate current. The gate current must be sufficient to remove stored charge from the P-base region and disrupt the regenerative action of the internal coupled transistors.

The silicon IGBT is arguably the most successful innovation in power semiconductor devices during the past three-decades. By using a combination of bipolar current flow controlled using an MOS-gate structure, the power gain was increased a million fold when compared with existing power bipolar junction transistors and power MOSFET structures with high blocking voltages [1]. The widespread applications for the device in consumer, industrial, transportation, lighting, and even medical applications is a testimonial to its highly desirable characteristics. The IGBT offers a unique combination of ease of control due to its MOS-gate structure, low chip cost due to its relatively high on-state current density, and exception ruggedness. Silicon IGBT modules are now available with blocking voltage capability up to 6.5 kV and current handling capability of 1,000 A. Any new high voltage power device technology must offer significant improvements in performance relative to the silicon IGBT to be considered attractive for applications.

In the previous chapters, it was demonstrated that the maximum operating frequency of high voltage bipolar silicon power devices is limited by the power dissipation due to their slow switching transients. The rate of rise of the voltage and rate of fall of the current during the turn-off process in these devices is slowed down by the presence of the large amount of stored charge in the drift region. Consequently, high voltage silicon carbide unipolar power MOSFET devices are a very attractive alternative to silicon bipolar power devices [1, 2]. Silicon carbide power device structures have been discussed in detail in a previous book [3]. In that book, it was shown that the conventional planar power D-MOSFET structure, developed and widely utilized for silicon, is not suitable for the development of silicon carbide devices. Two problems are encountered when utilizing the conventional power D-MOSFET structure for silicon carbide. The first problem is the much larger threshold voltage required to create an inversion layer in silicon carbide due to its much greater band gap. The doping concentration required in the P-base region to achieve a typical threshold voltage of 2 V is so low that the device cannot sustain a high blocking voltage due to reach-through of the depletion layer in the base region. The second problem is the very high electric field generated in the gate oxide because the electric field in the silicon carbide drift region under the gate is an order to magnitude larger than for silicon devices. This leads to rupture of the gate oxide at large blocking voltages.

In the previous chapter, it has been demonstrated that the silicon carbide planar-shielded inversion-mode power MOSFET structure has excellent on-state resistance for devices with breakdown voltage of up to 10,000 V. However, the specific on-resistance for these devices becomes relatively large when their blocking voltage is scaled to 20,000 V. Consequently, there has been interest in the development of silicon carbide-based high voltage IGBT structures. Due to the high resistivity of P-type substrates in silicon carbide, most of the development work has been focused on p-channel silicon carbide IGBT structures that can utilize heavily doped N-type substrates.

As discussed in Chap. 5, the silicon insulated gate bipolar transistor (IGBT) has been a highly successful innovation that has been widely accepted by the industry for power control applications with supply voltages ranging from 300 to 6,000 V. As shown in that chapter, the optimization of the IGBT structure from an applications standpoint requires reduction of the lifetime in the drift region to enhance its switching speed. This is accompanied by a significant increase in the on-state voltage drop for the IGBT structure. The large on-state voltage drop in the IGBT structure for smaller lifetime values in the drift region can be traced to poor conductivity modulation of the drift region near the emitter. A superior on-state carrier distribution can be obtained by utilizing thyristor-based on-state current flow as shown in Chap. 2. The gate-turn-off thyristor (GTO) was developed to take advantage of the low on-state voltage drop. However, the gate drive current for the GTO is very large as demonstrated in Chap. 4.

As discussed in Chap. 8, there was a flurry of activity in the 1990s to explore the development of MOS-gated thyristor structures due to their reduced on-state voltage drop when compared with the IGBT structure. The base-resistance-controlled thyristor (BRT) structure was proposed [1, 2] to take advantage of thyristor-based on-state current flow under MOS gate control to reduce the gate drive requirements. In comparison with the MCT structure discussed in the previous chapter, the BRT structure had the advantage of using a double-diffusion process similar to that used to manufacture IGBT structures. A rigorous study to understand the physics of BRT operation and evaluate the performance of experimental devices with blocking voltages ranging from 600 to 5,000 V was conducted in the 1990s [3–9].

As discussed in Chap. 8, there was a flurry of activity in the 1990s to explore the development of MOS-gated thyristor structures due to their reduced on-state voltage drop when compared with the IGBT structure. The base-resistance-controlled thyristor (BRT) structure was proposed [1, 2] to take advantage of thyristor-based on-state current flow under MOS gate control to reduce the gate drive requirements. In comparison with the MCT structure discussed in chapter 8, the BRT structure had the advantage of using a double-diffusion process similar to that used to manufacture IGBT structures.

The previous chapters have discussed a variety of high-voltage power device structures, based upon both silicon and silicon carbide, for use in high power applications such as mass transportation and power distribution. In this concluding chapter, the performance of these devices is compared and contrasted to provide an overall perspective of the available technologies. The comparison is performed using two categories of voltage ratings, namely 5- and 10-kV blocking voltage capability.