THE MAGAZINE FOR COMPUTER APPLICATIONS
Circuit Cellar Online offers articles illustrating creative solutions
and unique applications through complete projects, practical
tutorials, and useful design techniques.

DECISIONS, DECISIONSý

Choosing the Right Technology
by George Noveck

Start ý The Cost ý The Traps ý Take the Pick ý Logic Processing ý Software ý Whatýs Next? ý Sources and PDF

THE COST

Time is money, so they say, especially with the cost of engineering time these days. And because the project cost and schedule are closely related, the fundamental question boils down to deciding what to use. Is it better to use complex, sophisticated, more expensive components, bringing the design to market quickly, or spend more time designing in order to use simpler, less expensive, common components? In your business model, you also have to include the cost of manufacturingýweighing a single FPGA, for example, against a PCB populated with several hundred discrete components, the cost of troubleshooting, testing, and product support several years down the road. One way or another, the development cost will be reflected in the selling price of the product. Completely different scenarios will be present in a consumer-oriented product, like a cellular phone, which is built by the thousands and has a lifecycle of a few years. Place this in opposition to an avionic controller with perhaps 300 units manufactured over a period of 20 years (with no re-design), or a customized one-shot deal design.

In a practical sense, life is too short to be wasted on endless studies. As a manager, you are expected to fall back on your general knowledge and experience, cut through the maze of useless data, and make your decision quickly without agonizing over it as the days turn into weeks. Anything that shortens the time to market, reduces the risk of not meeting specification requirements, and simplifies the manufacturing process is a winner in my book, even though component cost may be greater. I will consider this approach first and, unless there is a bona fide showstopper, I wonýt waste time reviewing other traditional methods.

There are three reasons for this sentiment. You never fully appreciate the gain from being quick, but you will definitely feel the pain from being late. Also, microelectronics prices are always on a sliding scale. A complex, expensive part today may be inexpensive by the time your design is finished, but be careful not to fall into the trap I touch on next. Finally, engineering and troubleshooting time can be expensive, difficult to predict, and always expanding beyond original expectations. With lower proportion engineering represented in the overall cost of your project, the completion estimates will be better.

PREVIOUS • NEXT