Electronic Products & Technology

Five Best practices for designing flexible test stations

May 1, 2013  EP&T Magazine

Can we afford to invest in a test station that only tests one product? This is the question on the minds of many original equipment manufacturers (OEMs) and other high-tech device-makers today. In the past, when most electronics products had small feature sets and long shelf lives, it made sense to have a dedicated test station when you were producing high volumes with few changes over your product’s lifespan.

On the other side of the test spectrum, such as in the military/aerospace (Mil/Aero) industry, low-production volumes, combined with high instrument costs, meant that test engineers had to get creative with the architecture and components in their test stations to maximize their usage. With hundreds of mission-critical elements to test in every aircraft or complex vehicle/system, it is prohibitively expensive to deploy a dedicated test station for each one.

Given current market realities for consumer electronics and other mid- to high-volume devices – where large feature sets and short shelf lives are becoming the norm – many OEMs have seen the benefits of deploying more flexible, multi-product test stations like those pioneered in the Mil/Aero industry. Such stations allow OEMs to cover a wider range of evolving products and to get the most use from expensive test instruments.

This article shares some of our test-station design insights and best practices gleaned from working with hundreds of OEMs in multiple industries over almost 15 years.

1. Decide on basic architecture

A good starting point for designing a highly modular and adaptable test station is to identify which components you want to be ‘generic’ (in the sense they could be used for multiple products) and those that are product-specific. Your ultimate goal is to design a station that can cover a wide range of your testing needs over the mid- to long term.

For example, identify which hardware and software components you need for all your testing – such as standard instruments, a rack, a PC, a test executive and flexible cabling – and which you need for your specific products – such as fixtures and test packages/test code – as shown in the illustration.

When designing your test station, we also recommend that you allow 30-35% extra capacity because your product features and testing needs will definitely evolve. This extra-capacity guideline applies to rack space, modular-instrumentation chassis slots, communication ports, switching channels, etc.

2. Maximize instrument usage

As mentioned, the Mil/Aero industry tries to get the most value out of each test instrument because instruments typically represent the greatest expenditure in any test station. Other industries can follow that example by carefully choosing instruments that will provide long-term value, such as open-architecture, modular PXI instruments. When selecting instruments, you need to ask important questions like, “Will they provide us with significant value over time for testing multiple products?” and “Will they be easy to swap or upgrade?”

For example, if production for Product A only occurs several days a week or several months a year, ensure that when you select your instruments, they are flexible enough to be used to test Product B, C, etc., when Product A is not in production. To help with the decision-making, consider moving any intensive product-specific testing requirements into your fixtures or software test packages so that your instruments can address wider and longer-term needs.

3. Implement swappable fixtures

Many benefits can be achieved through the use of easy-to-swap fixtures. Instead of connecting fixtures directly to your instruments – which is time-consuming and prone to wear and tear –

implement an interchangeable test adapter (ITA) such as those available from Virginia Panel, MAC Panel or others. While a good mass-interconnect ITA can be expensive, it will really pay off because it creates a rugged interface between your expensive instruments and your unit-under-test (UUT) specific fixtures, allowing you to quickly change testing from Product A to Product B, C, etc.

In the fixtures themselves, integrate all those product-specific elements – such as USB-JTAG downloaders and drivers – that you will only need for one product or even one iteration of the product. As well, you will probably have to consider adding some UUT-specific switching in the fixtures to avoid having to over-specify switching subsystems in the station; also plan to have a good switch controller in the station with all control lines routed to the fixture.

In effect, you will need to solve more of your product-specific test issues with your fixtures – which are considerably cheaper than instruments – to ensure your test station has the widest possible application.

4. Use a standard test executive

In a multi-product test station you will need a flexible test executive – such as National Instruments’ NI TestStand – for your software architecture. The test executive sequences your tests and handles all your inbound and outbound test data, etc. Think of it as your high-value software interface to help organize and execute your testing and, importantly, to easily integrate test packages for new or updated products. A further benefit of a standard test executive is easy integration of an off-the-shelf test-data management product to consolidate your test results and make them available in standard reports.

By standardizing on one test executive you will also ensure that each of your test stations can be used by the maximum number of personnel, facilitating administration and operation. The last thing you want is to have highly customized – even unique – software architecture on each test station, complicating updates and troubleshooting, and exposing you to undue risk should the station’s software architect leave the company.

5. Calculate your return on investment

As part of a next-generation test-station design process, it is of course important to do a detailed cost-benefit analysis to understand the upfront capital expenditures (CAPEX) as well as the longer-term financial savings and business advantages.

The following table presents a simple scenario where an OEM currently uses four test stations part time (two for Product A and two for Product B). Instead, they could implement just three stations with swappable fixtures, allowing each station to be used full time to test either product, thus better aligning CAPEX with market demand and the organization’s priorities.

*See Table 1   13MYD343_TM1

* Figures are for illustration purposes only. Actual test station and fixture costs can vary widely.

Summary

Since new technology and business pressures are forcing many OEMs to rethink the “one test station per product” paradigm, it’s a good time to consider implementing a next-generation test station – one that can easily be used across several product lines and lifecycle stages. As well, the more standardized your base stations, the more likely that multiple units like R&D, engineering, manufacturing, and repair centres will collaborate, sharing product and station intelligence to achieve common and complementary goals.

Finally, by designing a test station with multipurpose components such as modular PXI-type instruments, racks, cabling and test-executive software – and product-specific components such as swappable fixtures and test packages – you will derive the greatest value from your test equipment, ensuring that it covers rapidly evolving feature sets and multiple products in the short term and for the foreseeable future.

***For more information on designing flexible test stations from Averna Technologies, go to http:www.xxx…

PHOTO CA
PTION

13MYD343_TM1

* Figures are for illustration purposes only. Actual test station and fixture costs can vary widely.


Print this page

Related Stories

Leave a Reply

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

*