Please note that JavaScript and style sheet are used in this website,
Due to unadaptability of the style sheet with the browser used in your computer, pages may not look as original.
Even in such a case, however, the contents can be used safely.
In the summer of 2002, NEC Electronics launched a project to break the on-resistance value of 2m ohm. As the result of enthusiastic persuasion by Kenya Kobayashi, team manager of the process development team, NEC Kansai Factory agreed to offer full-out cooperation for product development and trial production. Even at the starting line, there were real challenges ahead. At that time, the target value proposed by Takahashi, then general manager, was 1.9m ohm. The development team members hardly agreed with that target, because 1.9m ohm seemed exorbitant to them. "Say, 1.99, or we might be able to manage 1.95," said Manabu Yamada, a production developer who tried to change the target value, even though there were no grounds for believing that even a slight change would make the target any more achievable. It was just a proposal to bridge the team's hope of breaking the on-resistance barrier of 2m ohm with the pressure of working toward a seemingly insurmountable goal. Finally, after much discussion, Takahashi agreed to consider a target of 1.9m ohm.
To achieve on-resistance of less than 2m ohm, the team needed a finer development process. On-resistance of 2.4m ohm was the best could achieve using the conventional 0.35µm process. Adoption of a more sophisticated 0.25µm process was required. Although product development using the 0.25µm process had been used for consumer products, using it to produce automotive products seemed a distant possibility. Automotive products must be secure enough to withstand high voltages and robust enough to tolerate extreme environmental and operating conditions. Thus began an extraordinary project to develop power MOSFETs for both consumer- and automotive-type applications.
A power MOSFET, which switches the power of an electric system, looks like a simple device; however, it is developed only after much trial and error. For one thing, it's not possible to reduce its resistance only by adopting a fine process. Firstly, you need to improve a method of implantation of impurities at transistor formation. When elements such as arsenic, phosphorus, and boron are put in the right place with the right volume, the current flows more smoothly. As the result, the resistance value at the on-condition is reduced. Since a power MOSFET passes a three-dimensional vertical current, the key was diffusion layer formation. Working with the process development team, Kobayashi performed repeated computer simulations using his expert instincts to find the production method that would pass currents the smoothest and keep other important properties such as withstand voltage at its best. With his expertise and the right type of computer technology, the best of our production processes was gradually coming to fruition.
As a result of an innovative package connection method being adopted, connecting wires were increased from 2 to 4 (Figure 1). With increased wires, wire resistance at the connection part was reduced. Increasing the wires from two to four made the wire resistance one-half. The width of a wire normally was only 400陲卜, so that even with the limited area of connecting parts, it was easy to increase the number of wires from two to three by sliding their positions. However, the increase from three to four was not so easy.The center pin position of one of the packages was prescribed, although it was possible to use two types of packaging with the conventional method. Team members were stuck. One day the packaging design engineers who were busy with the four-wire connection issue stumbled on an idea. By using only one of the conventional packaging method that was capable of moving the center pin, four wires could be connected. This idea was good enough. Thus a package with a four-wire connection was realized.