There was a wonderfully cute product designed a few years ago by an analog semiconductor manufacturer that nearly buried the company. Without getting into too much detail, and thereby identifying the company, let's just say that it was a composite product containing a great selection of analog functionality that was totally programmable. Super design, extremely flexible, useful associated software to drive it, not expensive. What went wrong?

At lunch with another company we passed by the CEO on the way out, and the VP of Engineering at this second company summed it up in a nutshell, "The digital people don't know they need it, and the analog people want to design it themselves." And that is a problem analog engineers need to face up to sometimes; we simply do not want to get a working product off the shelf when we can do it ourselves for less. That doesn't make us directionless, but it does make us perverse.

When the semiconductor manufacturers make product these days they are generally making the parts as easy as possible to use; with little difficulty a designer can use some of the input range, or output range, to get a most reasonable result without scaling problems, for example. But most designers will still actually do the design work rather than get a "complete" solution in a manufactured part. This is especially true of the data acquisition market where complete data acquisition systems (DAS) are available covering many of today's needs.

So, why?

I think the analog designer is rather akin to the person who hates to ask directions (me) either while driving or in the supermarket. Analog engineers are loners, I have found, and it is one of the things that adds spice or magic to the art; we do it behind closed doors, nothing visible until the project is completed. We hate design meetings forcing us to guess progress dates and costs because there simply aren't enough people in the room who would understand what we are going through to get there.

If that is all true then we have egos that are unstoppable. And in the work environment that may not be bad: We know we can do it, we just don't know how yet!

But if that is how we think we need to get a grip. If there is a way of doing something, to the specifications we want, at the right price, we should do that. The time we save can be more usefully employed in solving those analog problems that just have to be solved. And peer advice, rather than pressure, can be a great solution-helper; but too often design reviews are not used to help people avoid blind allies where others have been before -- they too often turn instead into finger pointing and incredibly childish behavior.

In the same vein, if there is a solution that can reach your objective, within your budget and other parameters, that uses a digital method, use it.

By: Paul McGoldrick
Sr. Technology Editor, EDTN

Analog Main | Product of the Week | Columns | Editorial | Tech Notes

		Click here to get your listing up.