PACKAGING AND SUBSTRATE TECHNOLOGIES

 

SUMMARY

The RF Front-End of a phased array sensor system is generally realised by a number of multi-chip modules (up to several thousands), which typically contain bare semiconductor dies and other electrical components mounted on a ceramic substrate. The substrate is mounted inside a hermetic enclosure including sealed glass feedthroughs. This conventional technique is an extension of the well-established hybrid microcircuits technology. The major drawback of the hybrid technology is the separate manufacturing and assembling of semiconductor dies, carriers, substrates, packages and interconnects. Because so many different materials have to be put together, their mechanical and electrical characteristics have to be matched, the manufacturing and assembly processes are different and the whole module production becomes very costly, low yielding and slow. To overcome these problems new comprehensive technologies have to be developed.

Isolating ceramic materials can be thermally matched to the semiconductors, have low dielectric losses, offer multi-layer structures and embedded passive components, can form a hermetic enclosure with integrated interconnects and could be manufactured in volume within a single process.

An excellent thermal conductivity is crucial for radar modules, especially for phased array radars where the modules are packed very densely and several kilowatts of heat have to be removed. Some ceramic materials are inherently good thermal conductors (e.g. AlN) while others have to be sintered to a thick metal conductor (e.g. LTCC on metal).

The proposed research work covers the evaluation and incremental development of the two most promising technologies (AlN and LTCC on metal) and will extend to the development of a multi-chip module prototype, which is compatible with manufacture on a high volume industrial assembly line.

MILITARY BENEFITS

Insulating ceramic technologies offer a significant size and mass reduction of the multi-chip modules, whilst maintaining the performance compared to traditional packaging technologies. In addition these technologies are compatible with automated, high volume production lines that provide high yield, excellent repeatability and run at low cost.

Every remote sensing application with an RF front-end will benefit from cost savings, miniaturisation and mass reduction. This is especially true of phased array radars that need thousands of transmit and receive modules (TRMs) with minimum variation between which must rely on volume production processes with good repeatability. Some of the more sophisticated radar systems would not be affordable at all without an advanced ceramic packaging technology.

The US government has already put considerable funding into several initiatives (e.g. DARPA ‘High Density Microwave Packaging Program’, 1993 – 1998). In fact most of the ceramic packaging technologies are originated and located in the US or in Japan and there is the growing concern that these technologies could fall under export restrictions, which then will leave UK defence companies in a difficult situation (very similar to the power semiconductor technologies).

Power semiconductors in SiC and GaN technology offer a much higher power density than existing GaAs devices. These new devices will have a significant impact on military applications. The packaging and substrate technologies that exist today can barely handle existing GaAs power devices thermally. Hence new materials, concepts and technologies have to be developed, so that the performance of these new semiconductor devices can be fully exploited.

RESEARCH OBJECTIVE

To generate test structures to evaluate and characterise the electrical (e.g. losses, etc) and the mechanical (e.g. thermal conductivity) parameters of selected multi-layer ceramic processes.
To perform thermal simulation and evaluate different materials and concepts of heat sinking.
To design a simplified module demonstrator, which is representative for a remote sensing application and evaluate the overall performance.
To address and analyse the reliability and long-term stability of the proposed ceramic materials and technologies.
To identify advantages and drawbacks of both suggested ceramic technologies (AlN and LTCC on metal).
To quantify the performance trade-offs associated with commercially available processes.
To perform incremental developments on the multi-layer ceramic processes so that they are ready to be transferred into a commercial, high volume low cost production line.

RESEARCH OUTLINE

Low Temperature Co-Fired Ceramic (LTCC) technology is a multi-layer ceramic process that can be used to fabricate low cost and high performance RF substrates and packages with embedded components and interconnects. Standard LTCC ceramic materials have a relatively poor thermal conductivity and are in particular not well suited for higher power applications.

But recent results of improved thermal management concepts, such as thermal vias, heat-pipes embedded within the LTCC substrates and laminating the LTCC substrate onto a metal plate (before firing) give confidence that LTCC is suitable for these applications. There is a strong thrust from the commercial industry (especially in the US) to develop power modules for wireless base-station amplifiers based on LTCC technology.

However the AlN based ceramic technology takes advantage of the inherent high thermal conductivity of the AlN material and hence no additional heat-sinking concept is required. But there are other material-related difficulties with AlN, which have so far prohibited a wider breakthrough of this technology.

The final manufacturing cost of a power module utilising these technologies is certainly the driving factor. Currently it is not obvious which ceramic material will be the winner for military applications. Therefore the planned research activity will address both technologies evaluating the trade-offs to be made to provide the required performance at an affordable cost.

Because the final application, the assembly process of the module and the ceramic substrate and packaging technology are closely linked, the proposed research activity should specifically address power related issues derived from typical sensor applications (e.g. low loss electrical components, low dielectric losses, high thermal conductivity, reliability, stability, ruggedness etc. ).

A representative but simplified demonstrator for a typical power module in a remote sensing system should be designed and manufactured. The overall performance will be evaluated. Throughout the entire project it is important that all development work on ceramic materials, heat-sinking concepts and power module designs are compatible with industrial automated assembling technology. It is crucial that the developed ceramic packaging process is a high yield, low cost, volume process.

CO-ORDINATION WITH EXISTING / PREVIOUS RESEARCH

These research activities will build on previous research work in the UK and in Europe when accessible and available.

 

Leave a Reply

Your email address will not be published. Required fields are marked *