By Peter Ostrow
The United States military, agencies of the federal government, and prime contractors that support these organizations have purchased previously owned test and measurement equipment for more than 50 years. Yet despite the significant amount of previously owned equipment that is purchased regularly, there are few guidelines — either official or understood — for the research and procurement of this necessary equipment.
There are many reasons that an engineer or organization may choose to procure previously owned test equipment, yet these reasons generally fall within one or more of three categories: to save money, to save time on delivery of a new asset, or to replace an exact model/configuration that is no longer available as new.
Electronic test equipment — commonly referred to in the military as TMDE (Test, Measurement and Diagnostic Equipment) or GPETE (General Purpose Electronic Test Equipment) — has been a key component in the life-cycle management of electronic systems since the earliest communications systems were deployed and it continues to play a role in maintaining operational readiness for military services.
Test equipment ranges from a simple hand-held voltmeter for general troubleshooting to a complex signal analyzer for radio spectrum testing. This equipment — portable, benchtop, or part of automatic test systems — must support all electronic operations, from the field to the depot.
When a new system is being deployed or a repair to an existing system is necessary, the engineer or procurement official often must decide whether to purchase new test equipment or previously owned equipment. The vast majority of federal and military projects typically involves manufacturing, maintenance, or production that relies on "proven" previously deployed and tested technology. Hence, there is usually an adequate supply of previously owned equipment.
If this is the case, budget or urgency may influence the decision to purchase new or used. Purchasing previously owned equipment may end up costing less than half the list price of the new asset and may be delivered as quickly as a day or two. However, there are several factors that would-be purchasers of previously owned equipment should keep in mind.
First, search for the model internally within your own company/organization. Second, if it is not available inside the department or organization and you choose to procure a previously owned piece of equipment, always know where you are getting the asset. Specifically, who is guaranteeing quality, warranty, and right of return? Typical bidding processes, therefore, do not work, as one cannot discriminate between bidders based on "reputation" or hearsay. Therefore, the best way to procure this asset is through an informal bidding process in which only bona-fide test-equipment dealers, such as those that inventory the products, are invited to compete. Examples of dealers include organizations such as Telogy, DoveBid, and ElectroRent. Our service, which can be reached at www.navicpmart.com, does the work for you by aggregating the inventory of the three largest dealers, standardizing terms and conditions, and offering the products at a reasonable price. Regardless of whom you choose, know whom you are doing business with and how they stand behind their products.
Third, do not allow substitutions unless it is clearly understood that the asset will work as configured.
Fourth, if delivery is the main issue, make sure the asset is truly available from the dealer and that the manufacturer/distributor can't expedite delivery. In tougher economic times, vendors of new equipment are much more willing to work for your business.
Finally, avoid buying from end users directly, as there are hidden costs that make this an unacceptable choice for anyone other than organizations with complete evaluation, calibration, and repair facilities.
When a new unit is not appropriate
Because military systems evolve and often have their operational lives extended, the original test systems and equipment must still be available and maintainable to fill the support role for these systems. Today, two factors are combining to make system support for all types of systems more difficult.
The life span of military systems is extending as the government is squeezed by rising personnel costs and massive increases in the acquisition costs of new systems. Programs like the B-52 bomber, the F-16 jet fighter, and the M1A2 Abrams main battle tank have life cycles extended far beyond their original program goals. Therefore, procedures to test these older systems often are based on test equipment that, if removed, must be replaced with product that is a form, fit, and function substitution, or proper system functionality will be at serious risk — requiring a software rewrite and a complete system revalidation.
At the same time, product life cycles are shortening even as the move to commercial off-the-shelf (COTS) equipment accelerates. From an industry norm of five to 10 years, product life cycles have decreased to today's three- to five-year range. Further, manufacturers are not stocking spares for long-term support to the extent they once did, which makes test equipment harder to support and maintain. Also, manufacturers are adopting a "planned-obsolescence" marketing approach. By introducing newer and better features in new models, the manufacturer can leapfrog his competitor with a unique product. This feature shift makes it difficult to identify an appropriate replacement unit and makes it significantly more costly to repair or directly replace with an item from the original equipment manufacturer.
As the test equipment and systems that house integrated GPETE instruments age, test-system maintenance becomes increasingly problematic. Mechanical switches wear out and critical components become unavailable. As the military logistics systems run out of spares, the support commands have ever-increasing difficulty in maintaining readiness. Once their own support chain is unable to maintain an instrument and the manufacturer either has no more replacement parts or makes the cost of repairing prohibitively expensive and time-consuming, support commands have serious problems.
Further, there are relatively few government-operated repair and calibration facilities available today, and technicians with less training staff those remaining labs. This is due in part to the reliance on subassembly-level diagnostics and repair and to a pullback from traditional component-level troubleshooting. As a result, it is far more difficult to diagnose and repair a problem in relatively old systems because the equipment was not designed with this support rationale. Also, there is less likelihood that spare parts are available — especially original parts — once a problem has been diagnosed. The manufacturers' solution is to supply a next-generation "replacement" instrument, which is usually neither form-, fit-, nor function-compatible with the original.
Associated costs involved
Before integrating such a replacement instrument, the support facility must consider the "switching" costs carefully. Among the factors to consider are physical size and weight, power consumption, cooling requirements, electrical specifications, programming language and interface, and software drivers.
While the integration of new products to replace old units can sometimes be relatively straightforward, the real-world complexities of integrating a new product can be significant. The real cost drivers in the integration of replacement instruments lie in the areas of functionality, documentation, and software compatibility. Each time an older product is replaced, there is a resultant cost in the recrafting of test procedures, the software routines, and the accompanying required documentation for next-generation instruments.
This is especially true in Automated Test Systems (ATE). Historically, the investment in procedures, software routines, and documentation is the most costly part of test system's design and development. Therefore, in most cases modifying these items to accommodate a new replacement unit can be significantly more expensive and time consuming than the alternative of exact replacement. Although new instruments usually have better capability than their predecessors, the capability is usually not necessary. As an example, the following differences are noted between a popular (but no longer manufactured nor supported) signal analyzer and its current replacement:
- Interface timing different (software compatibility)
- Measurement timings different (software compatibility)
- Physically different shape (may require ATE rack modifications)
- Cooling intake/exhaust in different location (may require ATE rack modifications)
- Front panel controls/readouts different (a user documentation change is required)
- Front-panel connectors different and/or are in different locations (cabling changes)
- The front-panel nomenclature different (a user documentation change is required)
- The rear panel layout and connector positions different (requires changes in cable routing)
Replacing an instrument in a test system with anything other than an original piece of equipment requires engineering expertise — not only understanding the original test procedure, but also the characteristics of the original device under test (DUT) and the specifications and capabilities of the original instrument.
In the case of ATE, expertise in the original programming languages is necessary. Complicating the migration procedure are the subtleties in the software that depend on unique characteristics of the original instrument such as RF switching speed, input switching timing, and function command timing. In many cases, new instrumentation may not be able to emulate the required features/measurement of the original instrument.
Beyond these issues, there are broad sets of support requirements that need to be addressed when inserting a new item into a test bench or system. These issues include different technical specifications, requiring a rewrite of user manuals, maintenance procedures, and calibration procedures; operators must be retrained; maintenance and calibration technicians must be retrained; logistics support and provisioning must be developed; and new tools and support equipment may need to be purchased.
Further complicating the replacement scenario is the dwindling supply of competent hardware and software engineers who can recreate identical test-system functionality with a new GPETE instrument. This is primarily due to a loss of expertise through retirements and the switch to outsourcing.
"Exact replacement" reconditioned instruments
So what are the alternatives when a critical system needs to be replicated, repaired, or made more reliable and maintainable? The answer lies in the use of "exact replacement" instruments from a broad pool of existing products in the commercial and government sectors. By applying a rigorous inspection, evaluation, and reconditioning process, technicians can bring existing instruments up to "new" standards with a useful life equal to the original. This essentially makes an instrument "zero time" — or refurbished or rebuilt to the same levels of functionality and quality as when originally produced with the reliability and maintainability of the original "new" product.
Given the serious implications for substituting similar-but-not-exact replacements to support old benches and systems, a viable support technique is necessary to acquire reconditioned equipment through the secondary marketplace.
In most cases, it is possible to source a broad range of exact replacement equipment from the secondary market supply chain that can directly replace a piece of unserviceable equipment in a test system. In a perfect world, this approach is clearly the lowest-cost alternative because no changes in procedure, practice, or documentation are required. In addition, many suppliers in the secondary market supply chain operate refurbishing programs that can bring an existing malfunctioning unit up to the condition and specifications of a new unit.
Key elements in the secondary marketplace for a support chain are the understanding of a prospective item's configuration and condition. Understanding these elements begins at vendor selection. Prospective vendors must have documented systems for configuration control. Capability must exist to reconfigure an existing piece of equipment to make it an exact form, fit, and function replacement for an existing item. Further, the vendor must have a systems process for refurbishment to Federal Condition Code A. This process must include critical evaluation of all power-on life-critical items such as mechanical switches and rotating components.
A vendor should be able to supply a before-and-after report of condition and calibration that documents all key elements of a product's configuration and condition. Finally, the vendor should offer a warranty of at least six months and make a support commitment of an additional three years against that particular model and configuration.
Peter Ostrow is president and chief executive officer of TestMart in San Bruno, Calif., where he has been since early 1999. Prior to joining TestMart, he worked for Narrowline, a San Franciscobased Internet transaction and research company. From 1988 to 1997, Mr. Ostrow held various positions at The New York Times, including managing director for The New York Times Magazine.