Silicon carbide transistors are increasingly used in high-voltage power converters as they can meet the stringent requirements regarding size, weight, and/or efficiency of these applications. But why is this technology so fascinating to engineers? Our blog will provide some insights.
The outstanding material properties of silicon carbide (SiC) enable the design of fast-switching unipolar devices as opposed to IGBT (Insulated Gate Bipolar Transistor) switches. Thus, solutions which up to now have only been feasible in the low-voltage world with voltages of 600 V and below are now possible at higher voltages as well. The results are highest efficiency, higher switching frequencies, less heat dissipation, and space savings – benefits that, in turn, also reduce the overall system cost.
We identified this potential almost 30 years ago and established a team of experts in 1992 to develop SiC diodes and transistors for high-power industrial applications. Here is a short and incomplete list of milestones reached since:
MOSFETs (Metal-Oxide-Semiconductor Field-Effect Transistors) have been meanwhile commonly accepted to be the concept of choice when aiming at reliable SiC devices. Initially, JFET (Junction Field Effect Transistor) structures seemed to be the ultimate solution for merging performance and reliability in a SiC transistor. However, with the now established 150 mm wafer technology, trench-based SiC MOSFETs have become feasible. This way, the dilemma of DMOS (Double-Diffused Metal-Oxide Semiconductors) structures of having either performance or high reliability could be solved.
Wide bandgap-based power devices such as SiC diodes and transistors, or GaN HEMTs, (Gallium Nitride High Electron Mobility Transistors) are nowadays common elements in the library of power electronics designers. But why? What is so fascinating about silicon carbide in contrast to traditional silicon? And what makes SiC components so attractive to engineers that they use them so frequently in their designs despite their higher costs compared to silicon high-voltage devices?
In power conversion systems, designers continuously strive to reduce energy losses during the conversion. Modern systems are based on technologies in which solid-state transistors are switched ON and OFF in combination with passive elements. For the losses related to the transistors used, several aspects are relevant. First, engineers must consider losses in the conducting phase. In MOSFETs these are defined by a classical resistance. In IGBTs, there is a fixed conduction loss determinator in the form of a knee voltage (Vce_sat), and additionally a differential resistance of the output characteristic. The losses in the blocking phase can usually be neglected.
However, and this is a second point designers should consider, there is always a transition phase between the ON and OFF state during switching. The related losses are defined mostly by the device capacitances. In the case of IGBTs, further contributions are in place due to the minority carrier dynamics (turn-on peak, tail current). Based on these considerations you would expect that the device of choice is always a MOSFET, however – especially for high voltages – the resistance of silicon MOSFETs becomes so high that the total loss balance is inferior to that of the IGBTs, as these can use charge modulation by minority carriers to lower the resistance on conduction mode. Figure 1 shows a graphical comparison of the switching process and the static I-V behavior.
The situation changes when wide bandgap semiconductors are considered. Figure 2 summarizes the most important physical properties of SiC and GaN versus silicon. Significant is the fact that there is a direct correlation between the bandgap and the critical electric field of a semiconductor. With SiC, it is about ten times higher as compared to silicon.
With this feature, the design of high-voltage components is different. Figure 3 shows the impact, using the example of a 5 kV semiconductor device. In the case of silicon, semiconductor designers are forced to use a relatively thick active zone due to the moderate internal breakdown field. In addition, only a few dopants can be incorporated in the active area thus resulting in a high series resistance (as indicated in figure 1).
With the ten times higher breakdown field in SiC, the active zone can be made much thinner. At the same time, many more free carriers can be incorporated, and thus, there is a substantially higher conductivity. It can be said that in the case of silicon carbide, the transition between fast switching unipolar devices like MOSFETs or Schottky diodes, and the slower bipolar structures like IGBTs and p-n diodes, has now shifted to much higher blocking voltages (see figure 4).
Or vice versa: what was possible with silicon in the low-voltage range around 50 V is with SiC feasible for 1200 V devices.
The next piece about Silicon Carbide Technology will give you insights on how we control and assure the reliability of SiC-based semiconductor devices – stay tuned!
Links to explore further:
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.