-
- News
- Books
Featured Books
- smt007 Magazine
Latest Issues
Current IssueBox Build
One trend is to add box build and final assembly to your product offering. In this issue, we explore the opportunities and risks of adding system assembly to your service portfolio.
IPC APEX EXPO 2024 Pre-show
This month’s issue devotes its pages to a comprehensive preview of the IPC APEX EXPO 2024 event. Whether your role is technical or business, if you're new-to-the-industry or seasoned veteran, you'll find value throughout this program.
Boost Your Sales
Every part of your business can be evaluated as a process, including your sales funnel. Optimizing your selling process requires a coordinated effort between marketing and sales. In this issue, industry experts in marketing and sales offer their best advice on how to boost your sales efforts.
- Articles
- Columns
Search Console
- Links
- Events
||| MENU - smt007 Magazine
Rethinking the Use of Commercial Components
December 31, 1969 |Estimated reading time: 10 minutes
By L. Britt
Components off the shelf (COTS) offer a multitude of solutions, but careful consideration of many factors is essential to their use. Each potential situation poses a unique set of issues to resolve before final selection.
By their sheer volume, the commercial industries have influenced the buying power of the military. Regardless of how critical a project, the parts needed to complete a design may not be an option. The same vendor who fulfilled critical specifications years ago is no longer interested in low-volume, poor yield, labor-intensive design-to-specification. High-volume, disposable, short life components are more profitable and impose little or no accountability. COTS are not only a consideration, but have become the mandate. The more rigid military specifications, therefore, need to be re-addressed for both actual requirement and for work-around concepts such as redesign of associated circuitry.
Figure 1. Thermal cycling performance of select COTs devices.2
For military use of commercial parts, the problems are less obvious than lack of information within specification sheets and do not guarantee that date-code lots will match other lots to include the associated costs. (Note that samples within a lot may not match others from the same lot.) While component price may be lower, this initial price pales in comparison to the costs incurred because of rework, scrap, test and maintenance. Accelerated repairs and inspections introduce safety and logistics issues. According to the Naval Air Test Center in southern Maryland, maintenance associated with changeover to COTS parts use is one of the greatest concerns.1 Maintenance becomes the cost driver because more frequent checks are necessary and components may fail more frequently. System errors result and are followed by downtime for replacement. Maintenance personnel gaining access to the platforms holding these parts call for the transport of service personnel, test/repair equipment, parts to the in-field platform for repairs and recalling the platform to port for scheduled maintenance on a more frequent basis. Further, the COTS component stands a high risk of becoming obsolete. A new commercial part must update the initial replacement. Data and calculations for the initial replacement must be retrieved and updated. It is necessary to verify that these initially were suitable based on the original design calculations for the mil part and assembly. If a redesign had been necessary to implement the first COTS replacement, is yet another redesign required? Forward thinking can relieve some of this continual redesign effort. Further, is the advantage gained by the redesign substantial, or are much time, effort and cost expended for only marginal gain? If components can be evaluated and reliability predictions made with their initial selection, and if increased effort is extended to the refinement of module-level diagnostics, then a spare kit of plug-in modules for those containing most-probable failures can be prepared and included as routinely as the maintenance tool kit. The additional storage facilities and weight would be the concern here.
Is COTS Really A Bad Word?Commercially manufactured components may actually meet the necessary requirements. In many cases, the components are simply not tested throughout the required performance range, and therefore no information as to performance or range is offered. Alternately, some testing may have taken place, but to minimal quantities and, therefore, performance is not guaranteed. Component upscreening via test may be the solution. Discrete components may be measured for nominal values followed by bag and tag of matched sets. Components may be subjected to thermal and humidity cycling, thermal shock, and shock/vibration testing. Tests normally are performed pre- and post-stress cycling; however additional operational testing while in-chamber will provide a more realistic view of failure modes such as open circuits due to hairline fractures which 're-heal' themselves when vibration ceases or temperature returns to room ambient. Collection of historical test data and vendor information submission which assists in the determination of variances (e.g., changes in composition, quarry, supplier, manufacturing process, date-code and date-code vendor definition) are all part of the solution.
Upscreening may be accomplished by a) purchasing the commercial parts, then sending them to a test facility, b) testing internally, or c) buying from a commercial vendor who will upscreen prior to shipment. The choice lies in the facility readiness and experience. If either site is prepared and lab-time/labor are available, the more cost effective and reliable data may be attained by in-house screening, because of test parameter control, test witnessing and the savings of lack of incoming rescreening/inspections. Further, houses exist whose stock is comprised of components which have been already screened in anticipation of requests. Using an 'approved vendors' list has had its advantages and drawbacks, historically. Using manufacturers who stand a reasonable chance of existing 30 years from now is a legitimate goal. By adhering to that goal, some innovative new products were dismissed for military use. COTS use in a military environment does not imply abandoning the principle behind the approved vendors list, but rather calls for a process implementation that adheres to all reliability goals, approached in a new manner.
Figure 2. Typical failure sites.2
A clarification of true system operating and storage temperature, humidity, shock and vibration may be in order. A variance in performance/environmental requirements by internal assembly or section may improve the success rate of newly found items. Are the salt/fog and vibration requirements of a ship's exterior, or the high temperature environment local to an engine, applied to a module which resides within a sheltered, shock-mounted, well-ventilated environ? The broadened range of environmental stress, beyond those that actually will occur, must be determined next. Life cycle assumptions such as simulating the passage of time for dormant storage via thermal and humidity cycling must be validated. Figure 1 illustrates a sampling of devices showing failures after repeated thermal cycling.2 Five of the 19 products exceed 90 percent acceptable military survivability (to the right of dashed vertical line). Could one interpret that one ambient cycle per day may allow the 1,000 cycle marker to represent three years of thermal and humidity cycling? What similar systems using similar components have actual time-data to aid in these assumptions? The number of thermal cycles vs. percent of devices with failures for various packaging technologies subject to -55° to 125°C thermal cycling and 130°C/85 percent relative humidity/bias HAST are shown in Figure 1.
ObsolescenceObsolescence takes on two definitions in the context of this discussion. First, the original design, containing military-qualified parts only, is in need of substitutions at the component level because of failures. The components are not all available because of manufacture cessation. Second, the original component already has been replaced with a COTS similar part and the surrounding circuitry already has been modified to compensate for any lack of component performance. Now failure of the COTS component reveals a new problem. The COTS component, too, has become obsolete. A second replacement is necessary.
For the first case, a low quantity item with low failure occurrence may allow using such services as stock houses, which offer discontinued items. End-of-life buyouts may be scattered among many stock houses, while some may hold agreements with the former manufacturer to be the only source for the product. Certain parts may no longer be available or their availability may be too low in quantity to risk future unavailability. Approved distributors may decline the commitment to carry low-volume product.
The need for a form, fit and function retrofit is in order. Several options exist. Extensive cross-referencing (where such information is provided) would be the minimal effort required. Often such lists are seen as 'help for the competition' and, therefore, are no longer provided. However, organizations are dedicated to compiling such services and listings. Houses exist that purchase commercial, plastic packaged components and repackage the die into ceramic packages. Is the new product reliable after the extensive rework effort? Specifications must be re-evaluated for necessity. Perhaps a part exists that complies with most of the specified requirements, but with all of the real requirements. Component replacement simply may not be one for one. As with COTS substitutions, the obsolete parts' replacement effort may require some redesign. An entire function may be incorporated into an application-specific integrated circuit (ASIC) which would minimize the component-quantity of future replacements, while using programmable devices would allow for device reprogramming, rather than a search and replacement effort.
For the second case, wherein a COTS replacement part has become obsolete, there are some advantages. Design re-evaluations, specification review, etc. already have been accomplished. Disadvantages include moving on to an even less worthy candidate (assuming other candidate parts had been rejected with the first replacement). However, the passage of time allowed for the COTS-conscious community to have made some strides in evaluation techniques, availability of multiple sources and test facilities, and vendor accountability. Dealing with the problem of obsolescence must be viewed as an ongoing process with each solution considerate of those that may follow. Maintenance access to the component, plug-in modularity and reprogrammability are a few characteristics that deserve as much consideration as the component's reliability.
Reverse EngineeringIn the rebuild of an old design, the starting point is an existing, functional assembly. The design is valid; the parts are unattainable. Likewise, for the COTS or obsolete parts replacement process, the new parts are not all one-for-one retrofittable with the old. Rather than expend redundant effort getting to a replacement-ready point, with respect to documentation, consider the technique of immediate computer-aided design (CAD) editing of a printed circuit board's (PCB) artwork. For older boards, soft-files may not exist. The drafting system in-house may not be compatible with that used to create the drawing(s). The artwork, in the form of film or paper, or even the depopulated board itself, may be scanned into a raster capture system to produce the artwork, edit-ready, on the engineer's or draftsman's computer monitor. Here hand-cut traces can be brought to modern standards. Fiducials and tooling can be added. Redesigned sections can be edited. Layout techniques that work with the fabricator's processes can be redefined.
The particular system in use includes as output, various file formats for artwork (board layers, silkscreens and soldermasks), drill files, and pick-and-place assembly code. Inputs may be direct from an existing Gerber file or scans of hardcopy artwork layers or the board itself. Multilayer boards can be separated and each layer scanned individually.
The benefit lies in immediate access to the artwork, vs. cumbersome translations in an attempt to force proprietory CAD systems to accept another vendor's format.
Component Selection and InterfacingMany problems associated with COTS components deal with the use of plastics and lack of hermeticity. One may question the use of the term hermetic. Is true hermeticity, even in ceramic components, ever really achieved? Or have sights been set to some vague, lower standard of acceptability? Do all plastic packaged components necessarily popcorn (because of moisture capture and expansion under temperature increase)? Have not prebaking, underfilling and other encapsulation techniques, and improvements in resin purity used in plastics advanced to a point where some plastic components survive at the same rate as their ceramic counterparts?3 Coefficient of thermal expansion (CTE) mismatch remains the most adverse parameter in the use of plastic components on a ceramic base, or ceramic dice on an organic substrate. Microcracking, which occurs at the lead to component attach points and at the lead to substrate solder joints, can be alleviated partially by using developments in the following areas:
- Refining layout techniques (such as offsetting vias, necking-down traces) alleviates solder-robbing from the joint.
- Controlling the growth of the intermetallic layers diminishes microcracking.
- Choosing advantageous array packages, and using underfills alleviate warpage, stress on joints and cleaning issues. (Column grid array (CGA) packages, rather than ball grid arrays (BGA), will absorb some stress within the column, alleviating that at the joint. Full arrays, rather than peripheral arrays, may alleviate component warpage during storage and assembly, and alleviate some cleaning problems; whereas peripheral, vs. full arrays, allow desired distance placed between the internal die (heat source) and leads (balls) residing below.)
Common microcracking sites are illustrated in Figure 2. Such tools as the scanning electron microscope (SEM) are invaluable for inspecting items when developing and refining a process. Images before and after stress screening can evaluate the acceptability of an encapsulation, layout or other process that makes COTS usage more manageable.
References
- Reverse Engineering, Emphasis, a publication of the EMPF, December 1999 issue, L. Britt.
- Naval Air Test Center in southern Maryland. Thanks to K. Carter.
- State of the Market Survey on Microelectronic Packaging, Rockwell International.
- The Appropriateness of Plastic Encapsulated Microcircuits in a Specific Wooden-round Application, IEEE Transactions on Reliability, Vol. 45, J. Gardner, March '96.
L. Britt may be contacted at ACI Technologies, One International Plaza, Suite 600, Philadelphia, PA 19113; (610) 362-1200, ext. 217; Fax: (610) 362-1290; E-mail: lbritt@aci-corp.org.