In the early 1980s, a small company that developed weather data
collection and display systems was interested in a potential
contract to supply equipment to monitor the water level in rivers.
They had already studied the feasibility of such a system and were
interested in a fully functional prototype. I met with the owner and
his engineers, and soon thereafter, I had my first project as an
independent consultant.
I began with my client's mechanical concept of an ultra-sonic
device that included a calibration target and a method of mounting
an ultra-sonic transducer. After studying the project and producing
a plan, I hired another electrical engineer to help me with the
development. We designed and prototyped the analog driver and
receiver electronics and the digital logic design for the
microprocessor, memory, and user interface. The design was driven by
the need for low power consumption because the user intended to
place the system in a remote location for extended periods of time
and it had to operate on battery power. Of course, ultra-low power
components are readily available now, but in those days, high speed
CMOS devices were just becoming commercially available.
We wrote the software in Pascal and assembly language and a few
of you might remember my development machine: the Osborne Executive
personal computer that ran the CP/M operating system. We wrote all
of the real-time software to drive the ultrasonic transducer,
synchronize the reception of the echo signals, store the echoes, and
analyze them using statistical correlation techniques. We included
LED displays, a keypad, and a RS232 port for a remote terminal user
interface connection. Later, we added a modem so the users could
download the river height data over a phone line.
After my client evaluated the prototype, I developed the printed
circuit board layout, designed a rugged packaging scheme, and
produced a field-ready unit. Then, I completed my part of the
program by delivering all of my technical documentation to the
client.
In retrospect, this product seems primitive if we think of how we
could design it today with wireless technology, the internet, and
ultra-low power components, but even with these technological
improvements, we would still need basically the same user interface,
data presentation, the analog connections to the ultra-sonic device,
and most importantly, the physics underlying the measurement.
