noise amplifier. There’s a significant price penalty, however.
This price-versus-performance issue applies to
everything from connectors — gold plated versus tin-plated
— to the volume control. You can go with an inexpensive
carbon log potentiometer ($1.89), an inexpensive ALPS
step attenuator ($45), or a DACT audiophile-quality step
attenuator ($189). Step attenuators — which are made with
switched fixed resistors — have a nice feel and largely avoid
the scratchy noise associated with a typical potentiometer.
But there is the issue of diminishing returns.
Component selection is usually a compromise, even
when you know what you want. You might work with an
online database and catalog for hours, only to reach the
end of your component shopping list to discover that a key
component either isn’t available or must be purchased in
lots of 10, 100, or 1000. Then you search another site and
find the component is available in the quantity you need.
Do you start over with or incur an extra shipping charge
and order from both suppliers? Or do you reconsider
your design and use a component that is available from the
Furthermore, how many ‘spare’ components should
you order? If you’ve worked with surface-mount
components, you know what I mean. An inappropriately
timed blast from a hot air pencil can send a component
flying across the room. Do you really want to spend an hour
searching for an eight cent capacitor?
The complexity of your project may dictate a printed
circuit board over a breadboard. If so, you’ll have to select
from among the half dozen or so board development
systems and design your boards. (I like ExpressPCB for
simple projects and Eagle for complex layouts.)
Back to the headphone amp example, given the design
constraints, you’ll have to decide whether to use a single
board or the more expensive option of using a separate
board for the power supply. In either case, you’ll also have
to determine whether to mount the power transformer on
the board or directly on the chassis.
Circuit development is about making a series of
prototypes. The endpoint is never perfection, but
something close enough that you can live with it — and feel
good about sharing it with others. I’ve been developing
circuits for decades, and have learned that, despite my
best efforts, a circuit is a prototype until at least version
two. Sometimes it’s as simple as forgetting to specify a
trace-free zone around the bolt holes of a circuit board.
Sometimes it’s a complex ground loop problem that
requires a new layout.
As you read the projects in this issue, consider the time,
resources, and effort that each author has expended in sharing their designs. If you’re new to the development process,
start with one of these projects, and gradually work your way
up to creating your own designs. Enjoy the journey. NV
March 2008 7