A thermally optimizing method of thin film resistor trimming with machine learning

This paper proposes a novel trimming method of thin-film resistors dedicated to the power modules. Thin-film are utilized for applications including snubber circuits which avoids unwanted ringing appearing in output voltage transistors. Our previous works have revealed that applicable NiCr resistor is thermally limited, and then indispensable process affects degree temperature rise use. In this paper, we first formulate relationship between target resistance value trim dimensions using machine learning. With obtained equation, propose new method, enables trimmed pattern reduce variability maximum rise. The experimental results show proposed can suppress estimated range from 619K 179K.