Can the engineer in a garage still "make it"?

By Bill Schweber, Executive Editor -- EDN, 11/23/2000

I recently went to the IEEE's IMAPS (International Microelectronics and Packaging) conference (www.imaps.org), which focused on microelectronics components and packaging as well as pc-board-assembly topics. The conference was an eye opener. Although we at EDN concentrate primarily on design issues related to software, EDA, circuits, systems, and support components, we rarely look at the nasty, messy business of loading and soldering pc boards in a manufacturing environment. Yet, the reality is that you probably don't have a product, whether prototype or preproduction, until you build and debug at least one pc board that is reasonably representative of what you are signing over to production or showing to potential financial backers.

At the conference, I saw extraordinarily fast and precise pick-and-place systems—truly marvels of electromechanical engineering. These machines-moved effortlessly from component-source reel to component-source reel, plucking components, including fine-pitch ICs, and placing them gently but with laserlike precision. I was more amazed as the machines placed the passive components, which measured 1 mm sq (40 mils sq) or less. Some components were the size of a period on a printed page.

This precise engineering both impressed and scared me. We've all heard stories—some exaggerated, many not—of a "couple of guys in a garage" building a few rough prototypes that start their path to fame and fortune. Bill Hewlett and David Packard and their audio oscillator are perhaps the most famous, but there are many more stories.

But what happens when the pc is no longer a vacuum-tube design that you can point-to-point hand-wire, as in the design of the Hewlett-Packard unit? Or what happens when the design is no longer a hand-stuffable, manually solderable pc board with through-hole ICs and other components? Instead, it is a pc board with hair-fine tracks and spacing and microminiature active and passive components. Rather than building that demo prototype by yourself, you have to round up the small-quantity parts. (Good luck; the parts are usually available and manageable only on large-volume reels.) You probably need an outside service to build just one or two boards at a fairly large setup expense. In addition to these challenges, you will probably have to probe, debug, and change the pc board, which are not easy tasks.

I realize that not every innovative design needs the final pc board as a demonstration vehicle. For example, if you have a great idea for a talking vending-machine controller that somehow persuades users to spend more money at the machine, you can use an existing embedded controller or a PC to simulate your idea in a convincing mockup. But, for many products, the only way to convey your idea to prospective users (or venture capitalists) is with an accurate representation of the real thing. For example, a handheld PDA (personal digital assistant) is not the same in a large box; the PDA concept works only with its slim, deck-of-cards size. Today, most final packaging is more than an aesthetic function; it affects many engineering performance issues, such as EMI/RFI considerations, handling, and the hard-to-define sex appeal.

The situation for the engineer in the garage has a bright side, though. The same advances in electronics, mechanics, and software that have changed the nature of typical designs have also spawned rapid-prototyping tools that can help an engineer with a hot idea more quickly and with virtually no tooling build many parts of the prototype (Reference 1). (Then again, the original Hewlett-Packard oscillator used a low-tech sheet-metal box and standard radio knobs, which worked fine, too.) Additionally, software that simulates the user interface helps you define, refine, and demonstrate ideas (Reference 2).

So, do we have a problem today, or are new software-driven products making it easier to get design ideas up and running? Have complex-circuitry ideas become the domain of engineers who command significant design-support resources or make the right connections in the industry? I'm torn. One side of my brain says that we have a problem, and the other side says that the past few years have seen an acceleration in innovative product ideas from independent designers. I suppose I have to keep watching.

Author info