The functional testing of automotive electronic control modules (ECMs) presents many challenges, mainly due to the physical loads connected to ECMs: solenoids, motors, spark plugs, electromechanical transducers and so on. These represent an array of inductive and resistive loads that must be simulated in the production test environment to ensure proper testing of the ECM.

To accomplish these tests, a test engineer faces a “make vs. buy” decision: Design a load solution in-house or purchase a commercially available solution. Each approach has strengths and weaknesses. This article uses two key criteria—cost-of-test and time-to-production—to evaluate the load strategies that can be applied to functional testing in a production environment. It also presents key strategies to consider when planning for production testing.

Self-designed load solutions

For a typical test engineer, the creation of a load solution or “load box” doesn’t pose much of a technical challenge. The physical design is relatively straightforward. Most test engineers allocate ample physical space to house the various loads, the relays that switch them and a control interface to the relays. This may be implemented in a card cage large enough for the load cards or by building everything on a single plane such as a tray.

Independent of the physical package, other factors can complicate the design:

• Is the load box safe and reliable?
• Does it provide the required accuracy?
• How many load channels should be supported per card or tray?
• How much current is needed to support each channel?
• Is flyback protection required? (This is especially true for inductive loads.)
• Is access needed to measure the current flowing through a load?
• Must the load box mount on an instrument system?
• Is everything properly shielded from EMI?

Commercial load solutions

Many of the issues mentioned above are fully documented and specified in commercially available load solutions. That level of certainty is enhanced by the principal benefits of commercial solutions: lower cost-of-test (COT) and faster time-to-production (TTP). Let’s take a closer look at these benefits.

Cost-of-test

This evaluation criterion focuses on production and the test strategy. COT can be summarized with a simple formula.

The goal is to minimize COT. Inside the brackets, fixed cost and equipment lifetime are two important factors to consider when selecting a commercially available solution. The lowest COT is achieved through the lowest fixed cost and the longest lifetime (within practical limits). One consideration: The lifetime of a load solution can be extended if it is easily reusable in other projects.

Comparing fixed costs

Fixed cost is the initial investment in the load solution. The cost of fabricating a design will vary according to factors such as complexity and the number of load channels. As an example, it would cost approximately $1,500 to create a simple design with up to eight load channels, switching of 5-A currents and overvoltage protection. Comparing this to a commercial solution, a card alone may cost about $1,500; if a card cage must be purchased the total cost increases.

Figure 1. As the number of channels grows, the costs become comparable

Figure 1 compares the costs of the in-house and commercial approaches. As shown, the in-house approach costs less, but these solutions tend to be purpose-built for a single test. In comparison, a commercial design is more costly but it can cover a wider range of uses at that price point. One note: The chart suggests a crossover near 200 channels, which exceeds the typical need in ECM testing.

Continue to next page

Assessing lifetime and reusability

A custom in-house load solution can last as long as the product it is testing continues to be manufactured. For each new or different product, it’s likely that a new load solution must be fabricated—and this means incurring a new set of initial costs. The alternative is a commercial solution based on a flexible, modular architecture. With this approach, the possibility of retrofitting and reusing the commercial solution is much more likely for the testing of a new product.

The chart in Figure 2 compares the cumulative costs of in-house and commercial solutions over three new-product changes. The numbers are based on two assumptions: 1) the cost of changing the in-house design is the same as fabricating a new load solution; 2) the cost of retrofitting a commercial solution decreases with each change. In a high-mix/low-volume production environment, the number of changes can be even greater over a longer period of time, and this implies a bigger advantage for the commercial solution over time.

Figure 2. The cumulative solution cost over three new-product changes indicates that the commercial solution has an advantage.

This same concept can be leveraged for a shift in production capacity for different products. Figure 3a illustrates a production floor with four systems equipped with identical commercial load solutions. Each of these functional test systems includes a card cage and a corresponding selection of load cards.

Figure 3a. An example production line uses two sets of identical systems to support the testing of two different products.

Assume the following situation:

• Product A, which is being tested on Systems 1 and 2, is undergoing decreased demand, so the systems are running at 40 percent of capacity.
• Product B, which is being tested on Systems 3 and 4, is experiencing increased demand, so the systems are running at 80 percent of capacity.

Without further loading the capacity of Systems 3 and 4—and without building or purchasing a new system—a simple rearrangement of the load cards enables easy reassignment of either System 1 or 2 to the testing of Product B (Figure 3b). Comparing Figures 3a and 3b shows the effectiveness and life-extension made possible by the commercial solution.

Figure 3b. Refitting System 2 enables easy reassignment to the testing of Product B.

Maintaining a high utilization rate

Referring back to the COT formula, the ideal rate of utilization is always 100 percent—but this is not possible in practice. Fortunately, steps can be taken to sustain a high utilization rate by maximizing uptime and minimizing downtime. For any manufacturing line, three factors enable lower downtime:

• High reliability of the equipment
• Fast troubleshooting
• Readily available replacement parts

With an in-house load solution, troubleshooting time can be shorter due to deep familiarity with the design. It’s also possible to purchase additional parts to retain as spares, helping reduce downtime. Reliability is often the key shortcoming. If the focus is on creating the load box on a tight schedule, there is rarely time for any level of reliability testing prior to deployment.

For the commercial solution, reliability testing is a normal part of the development process. What’s more, manufacturers have statistical data that can be used to pinpoint which parts should be purchased as spares (for self-maintainers). If, as in Figures 3a and 3b, a common set of cards and card cages is used, then one set of spare parts can used interchangeably across all systems.

Continue to next page

Time-to-production

Demand forecasts are estimates and can therefore be unpredictable for new product introductions (NPIs) and existing products. Shorter-than-expected TTP or unexpected changes in a project schedule can set off a chain of cost increases.

Figure 4a illustrates the process of developing a new in-house load solution for an NPI. The cost profile is quite different from that of a commercially developed solution, which is shown in Figure 4b. Advantages of the latter include predictable delivery time; the elimination of design and fabrication time; and shorter TTP.

Figure 4a. Investment versus time for a test-development process that uses an in-house design

Figure 4b. Investment versus time for a test-development process that uses a commercial solution

As an additional benefit, the time savings enable the test engineer to focus on value-added areas such as debugging or qualification of the new product. Spending more time on these areas can result in better test coverage, which will have a positive effect on COT.

Summarizing: Side-by-side comparison

All of the foregoing can be condensed into the convenient comparison shown in Table 1. This summary will be useful when deciding to make or buy the required load solution.

Table 1. A side-by-side comparison of the two alternatives

Conclusion: In-house or commercial?

In a manufacturing environment, every element of cost flows to the bottom line. Selecting a suitable COT/TTP strategy for load solutions is a small part of the overall production strategy but it can provide significant savings in terms of both time and money.

Self-designed load solutions offer control over design and fabrication. In a low-mix/high-volume production environment, they also provide the benefit of minimal changes over a long lifetime.

Commercial solutions provide a higher rate of reuse, extending the overall lifetime of the equipment. This also leads to lower cumulative costs in a high-mix/low-volume environment that demands flexible capacity and frequent system changes. Further, standardizing on this approach allows it to be leveraged as part of a common test strategy. This creates commonality in multiple areas: rules and operating parameters; troubleshooting methodology; and support replenishment strategies. The combined effect is to provide additional benefits in time and cost.

As a starting point, evaluate each strategy as early as possible in the development of a test system. Look beyond the initial fixed-cost investments and consider the total cost envelope of each approach. With careful planning, you can select the best pathway to a lower cost-of-test in functional test.

References

• Implementing Loads for Automotive Functional Test, Bob Stasonis, International Test Conference, October 2003
• Cost of Test – The Components, Contributors and Ways to Reduce, Mike Clayton, Agilent Technologies, April 2002

About the author

Yeap Hock-Yew is a product manager in Agilent Technologies’ Measurement Systems Division. He has been involved with automotive functional test systems since 2005, concentrating on marketing, business development and the conceptualization of next-generation test systems. For any inquiries related to this article or Agilent's automotive functional test solutions, please contact [email protected].