Honda R&D Co Ltd announced the development status of the silicon-carbide bipolar junction transistor (SiC BJT) that the company is developing in collaboration with Shindengen Electric Manufacturing Co Ltd.

The two companies, which are aiming to realize a high-current BJT for use in vehicles, made an announcement about 50A and 100A products. The BJT was first revealed at the 7th European Conference on Silicon Carbide and Related Materials (ECSCRM), which took place in September 2008 (See related article).

The BJT features a high current amplification factor. As a current amplification factor increases, a BJT can be switched with a smaller current, enabling to reduce the size of the control circuit of the BJT. Though BJTs have an advantage of having a small on-resistance, their control circuits tend to be large for current control.

The 50A product has a chip size of 0.54cm2 and an active area of 0.25cm2. Its current amplification factor is 145 at room temperature and 50 at 250°C. The withstand voltage is 1,100V, and the on-resistance is 1.7mΩcm2 at room temperature.

On the other hand, the chip size and the active area of the 100A product are 0.79 x 0.73cm2 and 0.5cm2, respectively. Its current amplification factor is 135 at room temperature and 72 at 250°C. The withstand voltage is 1,200V, and the on-resistance is 3.5mΩcm2 at room temperature and 6.6mΩcm2 at 250°C.

This time, Honda R&D and Shindengen Electric formed a DC-DC converter circuit using a prototyped BJT and confirmed that it is possible to reduce the switching power loss at 15kHz by about 75% compared with the case where a silicon insulated gate bipolar transistor (Si IGBT) is used. As a result, the switching frequency can be enhanced by six times to 90kHz, according to the companies.

As for heat resistance, while Si BJTs start thermal runaway at 150°C, the SiC BJT does not do that even though its leak current increases.

Also, after a collector current of 4A (current density: 100A/cm2) was applied to an SiC BJT chip, the chip kept degrading little by little for about 100 hours. But it became stable thereafter and operated for 1,000 hours. The two companies consider that the degradation is caused by crystal defects called "basal surface transition" in the epitaxial layer.