The Mold Thickness Analysis (MTA) system enables users to determine the thickness of a semiconductor package in a non-destructive way and at the speed and precision necessary in series production environments. At the same time the system offers innovative inspection options that can contribute to significantly improving the product quality. With its capabilities it supports the trend towards highly integrated but very small electronic components, a trend fed in the first place by the growing dissemination of smartphones and tablet computers.
The system - it bears the type designation TS9000 - is the first product in a new family of analysis systems for series production. According to Advantest, it is already in action at several customers. Target markets of future THz products will, besides the semiconductor industry, include pharmaceutical, automotive and ceramics production. The company announced that it intends to leverage its current leadership position in the terahertz technology to integrated non-destructive analysis option into industrial production lines.
Fig. 1: Measuring mold thickness by means of terahertz waves: The terahertz pulses are reflected by the surface as well as by the content of the package. The system receives the echoes and determines the thickness of the package by determining the difference of the signal's time-of-flight.
The market environment for the system is determined by an increasing trend towards ever-smaller semiconductor components, driven by the demand for smartphones. Chip packages for these markets must be extremely compact yet robust enough for mobile use. For these reasons, the semiconductor industry is very interested in finding the right balance of package thickness and ruggedness. Conventional measuring methods however are not suited for high-volume production. Therefore, chipmakers currently can measure the package thickness only relatively late in the production process, and they only can draw samples. Existing measurement methods are slow, complex, and due to their destructive nature, not compatible with the needs of large-series production quality control. This makes it difficult to assess