In The Trenches
times — pretty much outright lies.
Way back when the number of
transistors in a radio was significant,
manufacturers would simply solder in
non-functioning transistors to pad
their count. Early stereo amplifiers
used to be rated in "music power"
watts. (This was, apparently, the
maximum peak power reached just
before the amplifier exploded.)
Currently, there is the ridiculous
wattage rating of some computer
multimedia speaker/amplifiers. I'm
looking at one in the May 2004
catalog of a very well-known hobbyist
electronics/computer supplier. These
speaker/amplifiers are rated at 480
watts and cost $39.95. It's interesting
to note that they look identical to
mine — which are rated at 80 watts
(also a joke). I took my speakers
apart and found that the amplifier is a
BA5406, which is rated at 3 watts per
channel.
Note that I use one specification
(and common sense) to contradict a
different specification. This is not
uncommon in engineering.
After all, a specification is just a
measurement and measurements of
the "same thing" can differ by
method.
Let's look at a fictional — but
realistic — example. You've designed
a logic chip with a propagation delay
of 5 nS. Suppose this is typical for
this type of device. Your company
wants a chip that's faster than the
competition’s. Can you accommodate your company without
redesigning the chip?
How did you measure the
propagation delay? You reply, "With
the standard 20 pF load." Suppose
you measured it with only a 10 pF
load? That would reduce the
apparent propagation delay.
Now, you have a chip with only a
3 nS delay. You've "improved" your
chip considerably — with no
increased expense. The company is
happy, marketing is happy, and you
get a raise.
People continuously find loopholes in standard measurement techniques in order to give their product a
perceived advantage in the marketplace. This is called "specsmanship."
While it's not really good engineering
practice, it's fairly common in
marketing (again, engineering vs.
marketing).
You clearly show on the spec
sheet that all measurements were
with a 10 pF load, so you are being
completely honest. Nowhere is it
said that your chips are faster than
the competition’s. If those
companies choose to use 20 pF to
measure their chips, that's their
decision. You have presented your
measurements and how you
obtained them. It is up to the
consumer to determine if your
product will suit their needs.
This illustrates the constant tug-of-war between engineering and
marketing. Engineering likes standard methods so that measurements are easy and consistent.
Marketing wants a product that
stands out so it will sell. It's
important to see that both points of
view are valid. There is nothing
unethical about making your product look good. There is nothing
wrong about emphasizing the strong
points and down-playing the weak
points, but it is also important to
realize that, in this case, your chip is
really no better than the competition’s. If you compare your chip in
the same circuit as your competi-tor's do, they perform the same.
Two Types of
Specifications
There are two general types of
specifications: performance and
target.
Performance specifications are
those that the product is guaranteed
to meet (hopefully).
Target specifications are those
that the product is designed to meet.
F
o
r
E
l
e
c
t
r
o
n
i
c
s
NUTS & VOLTS
E
v
e
r
y
t
h
i
n
g
84
Engineers usually define the performance specifications. Marketing
usually defines the target specifications. Generally, these target
specifications are based upon feedback from customers who say that
they need a product that does "___."
It's important to keep these types of
specifications in their proper place.
Unfortunately, too few engineers
understand target specifications. All
they see is that marketing has committed them to create a product with
SEPTEMBER 2004
