This paper provides a simulation modeling methodology for an isolated data transmission pulse transformer used for a multi-point transmission topology. When an equipment manufacturer creates an equivalent circuit model from the basic information that can be obtained from a pulse transformer manufacturer, there is a problem where the frequency band is insufficient and the quality of the waveform cannot be judged.

Therefore, we focused on the internal structure of the pulse transformer for the purpose of improving the accuracy of the equivalent circuit, devised an equivalent circuit that added elements that assumed coupling between windings, and fitted it to the actually measured S-Parameter. We confirmed that the ringback shape, which is an important feature of the waveform, was able to produce the desired waveform analysis accurately.

In the design of the signal transmission line between the control panels that make up the control system for the industrial application, this result suggests that it is possible to study the specifications of the pulse transformer by simulation without using large-scale waveform measurement as the main means of study.

This study examined the atomic diffusion bonding (ADB) of wafers using 3at%Cu-Ag (20 nm) films in air and compared it to that using Ag films. The bonding wave was observed even with exposure times of the film surface to air t_exp up to 2.6 × 10⁵ s (72 h), which was significantly longer than using Ag films under the ADB process. A large bonding strength was obtained after post-bonding annealing at 100°C in the t_exp range up to 1 × 10⁴ s (3 h), and at 150°C up to 2.6 × 10⁵ s (72 h). Structural analyses revealed that Cu precipitated on the surface of Ag-Cu films during film deposition, which prevented the cohesion of the Ag-Cu film surfaces by film surface exposure to air and improved ADB performance.

In this paper, the compact thermal model of DPAK package, one of surface-mount type discrete power semiconductor packages, is targeted and is created with a conventional methodology and its temperature prediction accuracy is verified. It is found that the compact thermal model created with adequate region setting and evaluation function shows excellent junction temperature prediction accuracy with ~0.1°C/W prediction error compared to 3-dimensional detailed thermal model. Also, it becomes clear that temperature accuracy trend of the compact thermal model is different from that of the three-dimensional detailed thermal model. After that, boundary condition set, which is applied for the compact thermal model creation is discussed as a requirement which enables highly accurate temperature prediction.