Page 60 - Analog Circuit Design Art, Science, and Personalities

Samuel Wilensky

Table 6-1
Construction. Technology Bi na ry..W eig hted Current Diviio~-
..
.
.
Printed circuit I2 resistors 36 resistors
Chip and wire . . 20 megohms . . 0. IS . megohms
.
.
.
_- . . .
If size were the only consideration, the binary weighting would be selected for the
printed circuit assembly and the current division would be the selected for the hybrid.
Since current division is faster than the binary weighted design, one might think
that the hybrid possesses the best of both worlds--small size and good performance.
Rut alas. the First LAW of Engineering' has not been repealed. The hybrid is smaller
and has better performance than the discrete component model, but to obtain these
improvements the designer must compromise on cost. A precision chip and wire
hybrid is always more expensive than an equivalent printed circuit dcsign. If size is
important. the user must be willing to pay for the decrease in size with an increase
in price.
A designer will usually have several circuit configurations from which to choose
to perform a desired function. The designer should evaluate all circuit possibilities
ana select the configuration best suited for the job. To make the proper selection, ii
designer must evaluate every component of the circuit and be able to integrate these
components into an optimum system.
The paper dcsign oE the circuit is only one aspect of product development.
Packaging: assembly? documentation, repair, trimming, testing, and last but not
least, helping the end user with application problems are all important parts of pro-
ducing ii usable product. A good designer becomes involved in every aspect of
product development. The designers name is on the product, and a good designer
should do everything possible to iissure its success. The designer should feel per-
sonally responsible when the product develops a problem.
At some point in the product development process. hardware, in the form of a
breadboard, will appear. This is a decisive moment. One now has a circuit to which
pow-ei can be applied. Before the breadboard was available, the design only existed
on paper. You now find out if your theoretical design performs as you predicted. A
word of advice: if the breadboard is completed on Friday afternoon. don't power it
up until Monday morning. Enjoy the weekend.
The breadboxd evaluation is a time to compare the actual perforrnance to the pre-
dicted performance. If the predicted and actual results do not agree, hewure. Don't
casually dismiss the difference. lnvestigate thoroughly until you find the discrepancy
between the paper design and the breadboard. I cannot emphasize enough the impor-
tance of attaining agreement between the paper design and actual circuit operation.
Occasionally during a breadboard evaluationl even though everything seems to
be operating properly, I will get a second sense that something is riot right.. It's hard
to describe, but the feeling is there. It might be a wave form that has an insignificant
wiggle or a voltage that is close to the proper value but not exact. When I get this
feeling, I investigate the circuit more thoroughly than I normally would. More times
than Rot I find a hidden problem. II'the problem is not solved then, it will appear at a
later time and ieally bite me in the rear end. If you sense a circuit is not operating
properly. take hecd: it probably isn't. Place your trust in the "Force" and investigate.
Working wirh customers on application problems is challenging and can be

3. Tlic Fi:-st Law ofEnginccrin~,"You don't get somctlling for nolhing." is a result of the First Law of
'i'hermodynarnics. The First Law of Engineering also has applications in economics. business. and
politic!,.

41