Abstract: In actual production, being affected by the manufacturing process, the reference voltage module inside the chip will have a certain deviation. During the chip mass production test, a kind of embedded memory unit trimming circuit is used to write trim code to control the chip’s internal circuits and fine-tune the parameters of the chip. As there are some factors that cause the test to fail during the test, retests are needed to improve the yield. However, if the traditional programming method is used for one time programmable (OTP) programming, the trim and program will be incorrectly programmed due to repeated programming during mass production testing, which will result in chip failure. Therefore, an improved OTP tuning algorithm is designed based on the automatic test equipment (ATE), which measures the initial voltage and frequency, and then tests and corrects the most appropriate trim code to avoid the repeated burning of OTP on the chip during retest. The proposed algorithm can improve the chip and reduce the production costs as well. The feasibility of the algorithm is verified by analyzing the test results of 8000 die on two wafers.

Key words: optoelectronics, semiconductor testing, improved test yields, OTP tuning algorithm, ATE automated test platform