For the PDF version of this article, click here.

Because 150-mm wafers tend to be the technology arena used to accommodate research and development of non-silicon processes, such as III-V compound semiconductors, as well as advanced micro electromechanical systems (MEMS), bio-MEMS or other nano-engineered structures, 150-mm probe platforms must offer the flexibility for quick-change setup in laboratory situations. In many such environments, the same measurement platforms need to be able to adapt for other research and development (R&D) requirements, such as packaged chips, multichip modules (MCMs), failure analysis of shards from large-format silicon wafers, motherboards and bio-structures or lab-on-a-chip designs. From a user perspective, the probe measurement solutions also have to be simple enough to efficiently achieve the desired results in the hands of non-electrical/electronic engineers, such as chemists, bio-engineers, grad students, etc., who may not be familiar with traditional wafer probing and measurement technologies.

This article provides an overview of the application and technology trends driving diversity in the 150-mm wafer segment along with a discussion of the new-generation, configurable measurement platforms that are addressing these constantly evolving requirements.

The continuing demands of high-performance, bandwidth-hungry applications are constantly pushing semiconductor designs into new arenas requiring advanced wafer level processes. As speeds increase and the dividing line between digital processing and analog circuit theory becomes blurred, the challenge of maintaining signal integrity is a critical factor for success. Signal integrity analysis of high-speed electronic designs requires that the interconnect models be valid over a wide bandwidth with accurate characterization of key factors such as rise time, pulse width, jitter, timing, noise content, insertion loss, return loss and crosstalk. Precision RF microwave and millimeter-wave measurement capabilities are vital when dealing with signal integrity analysis of high-speed and differential interconnects.

Higher data rates are also outstripping the performance capabilities of silicon and are driving the widening usage of III-V compounds, such as gallium arsenide (GaAs), silicon germanium (SiGe) and indium phosphide (InP). While offering significant performance advantages, the high-cost substrates and process challenges associated with these processes generally require the use of smaller form factors, such as 150-mm wafers.

In addition, 150-mm wafer technologies are being increasingly used for multidisciplinary applications, such as MEMS, bio-MEMS carbon nano-tubes and other non-silicon substrate structures. As shown in Figure 1, the major share of 150-mm wafer technologies are being used in academic environments, such as university labs, as well as in R&D departments of semiconductor and systems manufacturers, government research labs, and fabless design houses.

These users require a high degree of adaptability from their test and measurement investments. In academia especially, the required measurement techniques can encompass a wide range of disciplines. Today's grad student might be researching silicon carbide power transistors in one clase looking at millimeter-wave noise figures in another class. In such dynamically changing environments, it is beneficial to be able to tailor the platform so that it can be conveniently re-optimized for the specific challenges at hand.

In the past, probing and measurement platforms have not typically been designed for multipurpose operation. In order to optimize precision measurement capabilities, previous generation solutions have tended to be application-specific and monolithic — essentially designed to perform a certain set of measurement tasks with a high degree of repeatability and accuracy. When multiple measurements required, the typical approach has been to deploy multiple platforms to handle the full suite of tasks. Although this practice certainly can be cost-effective for higher volume production-oriented requirements, it forces a number of untenable compromises in R&D and laboratory situations — from a budget and operational perspective.

Previous attempts at designing multifunction probing platforms to meet the budget and diversity requirements of lab situations have tended to fall short on the critical issue of delivering precision measurement capabilities. Any such compromises on performance are completely unacceptable, especially since R&D and university labs are pushing into new technology arenas that require the best measurement and characterization methodologies available.

The key to success in the newest generation of measurement platforms is the use of a modular design from the ground up, which enables the core platform to be optimized for precision motion, probe placement and ease of use, while the specialized probe modules are optimized for specific precision measurement challenges and quick-change setup to accommodate different requirements.

The modular system design approach overcomes a number of challenges that made it hard to smoothly adapt previous-generation monolithic platform architectures for multiple functions. For example, it is quite difficult to integrate and optimize conventional wafer probers to handle large tuning systems. In contrast, the flexibility of quick-change device under test (DUT) holders and probe assemblies in a modular system allow the devices to brought in much closer proximity to the tuners, thereby improving measurement efficiency.

The ability to quickly change over between measurement tasks is also a key advantage of using a modular system approach. For example, DUT holders for different devices typically can be changed out in less than a minute. In addition, the modular design allows probes to be quickly changed out and/or repositioned at required locations throughout the test platen.

In the RF world, probe management issues, such as quick positioning, calibration, standards, and overall range of movement can also be critical factors for efficiently achieving desired test objectives. The positioning mechanics need to allow the RF probes to be quickly and easily placed anywhere on the platen. For instance, probes frequently need to be brought together for calibration. Any RF testing process using a vector analyzer will require the probes to be regularly positioned within a two to five mils of distance in order to accurately perform calibration functions.

Then, when making RF measurements, the individual probes may need to be placed at various distances around the test platen, depending on the test parameters and the device under test. Testing of a TR module or power amplifier module can require two to three inches of distance between probes and the testing of traces in a printed circuit board (PCB) assembly might require interprobe distances of up to six inches. Previous-generation monolithic probing systems could not deliver the flexibility to accommodate this range of motion quickly, efficiently or with the required level of positioning accuracy.

Previous designs often required test probes to be either bolted to the platen, which limited positioning range to only one or two inches, or they were magnetically attached, which could create problems with maintaining proper orthogonality. In contrast, the combination of quick-release probe holders and a sliding-clamp mechanism allow users to rapidly position RF probes at any required test points, with consistent orthogonal alignment.

Fully integrated microscopy is another key aspect of new-generation modular platforms. As contact interface consistency becomes increasingly important, it is vital for users to be able to quickly reposition probes into proximity with the test points and then bring them into exact contact alignment. An integrated microscopy bridge allows the user to easily position the high-power microscope anywhere within the probing plane, thereby enabling the optics to be smoothly translated throughout the entire test envelope. As with the probe positioners, the microscopy unit is designed to maintain uniform orthogonality during repositioning to any new location. Unlike traditional boom-stand optics used in many previous-generation platforms, integrated orthogonally consistent microscopy doesn't require any tweaking of alignment at the new position, making it immediately usable for optimizing contact accuracy.

In addition, the tight integration and consistent orthogonal relationship between the optics and the probes can help to avoid costly accidents. With conventional systems, the number one cause of RF probe failure is from damage during translation or repositioning because magnet or vacuum-based positioners offer no way to keep the probes on a consistent plane. With the least-expensive RF probes costing $600 and high-end probes costing as much as $8000, inadvertent collisions between probes and the microscope units or other obstacles can quickly become a significant factor in total cost of ownership. By automatically controlling the orthogonal movement of both probes and optics with uniform clamping mechanisms, new-generation measurement platforms can protect even novice users from costly mistakes and accidental collision damage.

When conducting precision RF measurements on today's high-performance devices, users are constantly moving the probes between the device under test and the standards to calibrate the system, or other functions such as cleaning or planarity checking. The availability of multiple auxiliary chucks in a modular system coupled with a full range of motion can be helpful by allowing easy movement of the probes between all of these stations without having to detach or manually reposition the probes.

Another important feature in modular platforms is the use of a “rollout stage” that allows all of the chuck tools to be brought forward out from under the positioners in order to provide easy access for loading or changeover of DUT, standards, etc.

A good example of using the new modular test platform in the RF world would be for optimizing load-pull testing, where system-induced impedance can be a major obstacle to achieving desired measurement results for low-power operation.

With older systems, the required cable lengths could be 100 mm or longer, translating into as much as 2 dB to 3 dB of insertion loss at 18 GHz. By optimizing the DUT holder to bring the device under test up close to the tuners, a new-generation platform can eliminate the need for any extra cabling, which significantly reduces the impedance associated with the test mechanism.

Such a “zero-length cable” configuration can bring the total insertion loss attributable to the test system into the 0.2 dB range. Whereas previous systems had difficulty getting below 5 Ω for power amplifier tests, new lower impedance systems can allow for precision testing at 1 Ω or less. This goes a long way toward the ultimate objective of making the test system “invisible” to the tuner and vector analysis functions.

Having access to new-generation flexible measurement platforms, which can be quickly and cost-effectively re-optimized across a range of precision measurement tasks, enables R&D departments, university labs and fabless design houses to achieve their budget and operational objectives over the short and long term.

Instead of needing to invest in multiple platforms for different measurement tasks, this “no compromise” approach allows users to leverage a single-system solution with quick-change probe technologies that are optimized for specific measurement requirements. In addition, modular measurement platforms are laying a foundation of ongoing adaptability that can be optimized for future needs without a fork-lift swap-out of the core system or continually paying a price premium to meet emerging measurement requirements.

Diana Laboy-Rush is a product manager for the M150 measurement platform for Cascade Microtech. She has more than eight years of semiconductor applications experience ranging from wireless remote control design to power management system verification. Laboy-Rush has a Bachelor of Science in Electrical Engineering from California State University, Long Beach and a Master of Science in Engineering Management from Portland State University. She has been with Cascade Microtech since May 2005.

Terry Burcham entered the microwave industry in 1968, responsible for amplifier R&D until 1974. From 1974 to 1982, he was co-founder and vice president of Aercom Industries, producer of GaAs FET amplifiers and microwave components. Burcham joined Tektronix in 1982 as marketing manager for the Frequency Domain Instruments Division. At Cascade Microtech, from 1987 to 1997, he held positions of marketing manager and sales vice president. In 2002, Burcham came out of retirement to be worldwide applications development manager.

RSS    Save to Del.icio.us  Digg This