Page 109 - Analog Circuit Design Art, Science, and Personalities

The Process of Analog Design

ment the resultant plan). You’ll probably also have some preconcieved notions about
what you want to do and how to do it. I won‘t fault you for having these (although
there are those who say that even having any preconceptions is harmful), but keep
them well into the background in the beginning. You can be sure that you will be
dragging them out often enough later to compare to your predictions or experiences
with your “creation.” Just dive in first. You have to get some movement and fermen-
tation going.

Getting Rid of Excess Baggage

A good way to start is to look at the goal-but not necessarily the one you think
you have. Your original goal is often cluttered with all thc assumptions about “what
can’t be done” that you have chained to it or have allowed others to do so. Try start-
ing with a clean sheet of paper. If there were no knowns, if there were no limits,
what would be your goal? “But!” you say‘? No “buts” for now! Put them all aside-
you can retrieve them later if necessary. Take some time with this step. After all,
why not end up with the thing you would really like to have as opposed to some-
thing that is merely in the ballpark‘? The effort you will have to expend is probably
not much greater and, in fact, can be considerably less as a result of having a clear
goal in mind. Tell yourself “this is what I want the result to be” and be specific!
Start with the big picture and qualify. Then qualify and qualify some more.

Analyze Later
Analysis-to use a tool or set of tools to predict or qualify systems-is important, as
they taught us in school. There will be considerable opportunity during the design
process to demonstrate that we can analyze. First, however, we need to generate
something to analyze. My experience is that it is exceedingly difficult (read impos-
sible) to be in generation and analysis mode ut the sume time. So here in the “gener-
ation of ideas‘’ phase, let the ideas have a chance to form (if not blossom) before
subjecting them to rigorous scrutiny-yours or someone else‘s

Don’t Forget the Fixed Overhead
Fixed overhead effort is something we usually fail to take into account. If the dif-
ference in design effort is, say, 2: 1 to design a really good part (we won’t say great
because it is considered impolite) we tend to think that is too high a price to pay.
But if the design effort is only 20% of the total effort required to get the job done,
then the “really good part” costs only an extra 10%. Is that worth it:’ We all get to
answer that question for ourselves. I say yes.

Forget the Window Dressing-For Now

When you are trying to get a new idea down, don’t worry about the window
dressing, c.g. the grammar, the neatness, drawing within the lines. There is plenty
of time for that later. It gcts in the way and can impede or stop the whole process.
It may even be that someone else winds up doing the “window dressing.”

Synergy!

Synergy works. Get a group of people together and brdillStorm. or generate ideas.
The people you choose should bc bright, eager, intcrested. and open people, but if

90