An Interview with
David Wilkie
Everett Charles Technologies

by: The Board Authority
March 1999

The corporate technologies liaison of Everett Charles Technologies discusses how the world's largest test equipment supplier is addressing electrical test issues.

The Board Authority: What is your opinion of the increasing demand for reliable and affordable electrical test?

DW: There are a variety of markets for tested interconnect substrates, and the demands vary amongst these markets. In general, we see several trends, and we are pursuing solutions that address them effectively without requiring obsolescence of existing test vehicles.

First, there is the continuing trend towards finer pitch and higher density. Ten to fifteen years ago, this was regarded as a result of the transition to surface mount technologies. Now, we see multilayer laminate structures for IC packaging which extends this demand towards finer pitch and higher peak local density.

A second and equally important trend is the demand for improved measurement parameters and measurement modes in electrical test. Customers (and test engineers) demand increased low-resistance continuity measurement accuracy, embedded component measurement, high-resolution measurement of embedded resistance, and on-product verification of RF impedance performance of specific signal traces in substantial volume.

It is also worth noting that the number of electronic interconnect substrates requiring verification continues to increase, as does the average complexity of these products. Put simply, the amount of test measurement we need to perform is increasing and, as noted above, the quality of these measurements must improve accordingly.

Finally, it is important to note that the cost of conducting no testing has increased. Many interconnect customers are now requiring bare board manufacturers to shoulder the responsibility for consequential damages resulting from defective substrates.

We do not believe that any single technology offers a "total solution" for electrical test. The successful test manager will be the one who is able to integrate combinations of functions, deriving the best benefits of each. It is our responsibility as a test vendor to support these customers with integrated solutions.

We believe that visions of a future which assume a vast decrease in the total quantity of substrate testing, or a substantial degradation of test performance, are unrealistic. This isn't going to happen soon. The market wants a better test which is faster-not a limited test which is slower-and addresses gigahertz logicdemands with features such as RF impedance test. Test engineers must become test managers.

The Board Authority: What are typical performance parameters for existing technology, and how does this compare to possible near-term future technologies?

DW: Mechanical probing systems exist which can often meet the mechanical accuracy and electrical measurement requirements, but they remain somewhat slow when compared to fixtured test systems. Most probes are limited to roughly twenty test points per second and, therefore, require five to sixty minutes to test a single board. Proposals for a variety of "fixtureless" electron beam, plasma, or laser test systems promise the possibility of a 10:1 improvement over this behavior (at the expense of measurement parameter range). However, none of these promise the throughput of existing high-volume board testers-electronic test speeds up to 25,000 test points per second.

With appropriate fixturing, measurement resolution down to fractions of one Ohm is available with existing technologies, but the high impedance nature of most of the alternative technologies results in a measurement performance with one or two orders of magnitude poorer.

The Board Authority: "New test technology" is become a popular phrase among electrical test people. What is your position?

DW: We've been blessed with opportunities to review a variety of proposed measurement and probing alternatives, most of which are rumored in the market. It is ECT's view that no single technology offers a complete solution to the primary needs of the users, and that combined technologies are the reasonable and likely answer. We are excited by the opportunity to integrate the best new technology with proven features, and our R&D investment has continued unabated.

Because we think that the answer will be a selection and integration of different technologies, we believe that a critical planning activity includes moving the test area towards seamless integration of equipment types, whether that equipment involves technologies which are new or old.

The Board Authority: It seems you do not believe in the point that off-contact test technologies will become alternative test solutions in the coming months. Can you explain why?

DW: On the contrary, we are active participants in a variety of emerging technologies. However, we believe that it is misleading our customers to imply the emergence of a fixtureless, painless "total" solution to all of the test demands that we discussed earlier: speed, density, pitch, measurement accuracy and fault coverage, and RF characterization-all within the next 12-24 months. That would be ideal-but it isn't going to be that simple.

A number of electron beam, ion beam, laser plasma, and optical alternatives have been suggested. These are generally noncontact methods or, in some cases, use a limited number of contacts.

While all of these technologies are new, and subject to future enhancement, they all appear to share the following major shortcomings, to greater or lesser degrees.

1. Speed. These methods generally energize the board via a fairly highimpedance source. This means that any voltage changes to be imposed on the board are supported only by limited electron currents, and voltage is sensed by limited means. As circuit boards are essentially capacitive loads, these methods tend to remain slow compared to full contact test. If the benchmark of comparison is an older mechanical flying probe system-and if you presume no parallel progress in mechanical probe systems-then the numbers can be made to look good. But the reality is that most of the tested production of substrates runs through fixtured systems.

The best of present "fixtured" test systems reach speeds of up to 25,000 measurements per second, and this is not a theoretical limit. The new technologies are often described as limited to tens or hundreds of tests per second, assuming successful advanced development. This is a speed performance difference of 100:1. It is difficult to imagine substrate manufacturers investing in 100 times-or even ten times-as much test equipment. The total number of electrical interconnects requiring test is not decreasing.

2. Indirect Measurement. None of the disclosed methods have good resolution for low-resistance requirements. A few reach below 100 Ohms, but poorly. As most metallic contacts on the board are supposed to be very low in resistance (less than 1 Ohm), this results in a loss of test coverage. In most cases, these methods do not validate continuity by directly passing a current through the DUT conductors and thereby "operate" the conductor being tested. Similarly, they do not typically measure isolation by imposing a voltage difference between supposedly isolated conductors, verifying the absence of current flow. Instead, they tend to rely upon various "indirect" methods of measurement. Most of these reduce to sensing the capacitive effects of a network of given size, as described above. Such methods are particularly poor in detecting marginal failures, such as a circuit almost cut in two, but joined by a short resistive segment.

Another example generally undetectable with the above methods would be an excess resistance caused by substantial over-etching of a conductor. A third undetectable example would be a circuit including an embedded resistor element within the board, where such resistor is low in value and slightly out of tolerance. High-speed signals, including RAMBUS and other emerging technologies, promise to increase the usage of embedded passives.

3. No support for RF test. Interest in RF impedance performance continues to increase, and contact is required for this test.

4. Operating Costs. Many of these systems are complex in construction, and some require elaborate vacuum systems with substantial maintenance commitments. While the idea that they are "fixture-free" is appealing to an engineer who just lost his evening to a defiant fixture, the same loss of time can occur to a balky vacuum seal or unstable optics system.

The Board Authority: What do you suggest?

DW: We believe that 21st century test will involve integration of specialized technologies. To prepare for this, test departments need to streamline software systems to deliver accurate test data for a range of technologies. To minimize the pain of this integration, it is worthwhile to consider reasonable limits on the number of different equipment models and brands to be intermixed. Whichever supplier one chooses, it will be helpful to the engineer to achieve some degree of uniformity of equipment and software. This ability to channel data will be important to any combined-technology approach, no matter what the technique. A test manager who must support an unplanned equipment mix will begin with a severe disadvantage.

Unfortunately, the simple sophistication of uniformity is slow in coming to final test areas. Too many shops try to support a broad mix of equipment, fixture, and data types. The good news is that we see this changing as board shops emerge as global corporations-and test engineers emerge as test managers.

As far as endorsing specific measurement techniques is concerned, the R&D engineer in me doesn't think the available information supports more detailed analysis than I've given above.

The Board Authorlty: Basically, you recommend more management instead of engineering in test departments. Will it be enough?

DW: The goals of solid engineering and good business management are complementary. The globalization of the PCB industry, combined with the technical pressures we've noted, result in an opportunity to increase the visibility and focus in test-in a desirable way. In the past, test has often suffered as the "no-value added" poor uncle within the board shop. Today, the cost of shipping defective product is too high. Test managers have finally found a voice in corporate councils, and have new opportunities to streamline dataflows and equipment selection. In the past, there was one family of drills in the drill room, but three to five kinds of test systems in the test room. This doesn't have to continue. So it's not an issue of reducing engineering attention; rather, it's an opportunity to leverage this increased visibility towards a better organized suite of tools in test. I'm not suggesting planning as an alternative to new technology, but it is a necessary prerequisite to integrating and capitalizing upon new technologies, whatever the source.

The Board Authority: Do you have ongoing developments?

DW: Always. We're occasionally short of sleep, but never R&D opportunities. We're fortunate in having many talented engineers involved in coordinated R&D activities at several of our divisions, including the original test equipment division, Luther & Maelzer, BSL" and, most recently, atg® Test Systems. As far as advanced proprietary development is concerned, we try to avoid announcing products before they are ready for general introduction. However, I'll try to give you a few examples of test system products that were recently released, or are being released as shipable product at the IPC Printed Circuits Expo.

I expect to see a new test system for extremely fine-pitch high-density product types, providing nonetheless a full contact test, yet making minimal witness marks upon the product. For appropriate product types, this product is capable of production throughputs. No compromise of long-standing test specifications is required. DC measurement is made of both continuity and isolation, with excellent accuracy and resolution. The approach taken in the product is completely different than anything seen before, and relies upon innovative combination of advanced mechanics, software, and electronic hardware. Verification and validation processes have been underway for many months, and the product unveiled at the IPC Expo will be immediately available and ready for market.

A recently introduced product is the world's first production ready on-product (not limited to coupons), fully robotic TDR with complete pass/fail automation, and data-driven programming. Another example is the newly introduced ECT Oz" system. This is the first automatic fault verification system. Not only does it improve labor and material efficiency in the rework of defective boards, it also increases, by an order of magnitude, the following characteristics: quantity, quality, and immediacy of process specific failure analysis data available from final electrical test to the fabrication side of the bare board shop.

We are also working very hard on improving the service and support infrastructure available to ECT, Luther & Maelzer, atg® Test Systems, and BSL" customers worldwide.

The Board Authority: Do you have any suggestions for test users?

DW: While some of the proposed new methods hold promise in specific areas, at this moment they do not truly exist as deliverable products. While we all wait for this situation to change, an immediate issue is to prepare for the future by managing a program of investment in interoperable and upgradeable systems. When investigating new technologies, don't forget the basics: test coverage, measurement, speed, cost, etc. Understand whether you are funding R&D risk, or buying a proven product. Finally, evaluate test technology partners as business partners.