-JULY 2008-
-Circuit Cellar readers change the way Renesas builds
development kits
-Microchip Technology Gears Up for Latest Circuit
Cellar Sample Distribution
-Catching Up with Circuit Cellar Author, Design
Contest Winner Jens Altenburg
Circuit Cellar readers change the way Renesas builds
development kits
52% of Circuit Cellar survey responders said that software development ate up more than half of their allotted project time. Independent studies go further, placing many developers in a category where up to 75% of their time is spent on software-related development issues.
Renesas Technology responded recently by taking a serious look at how its development kits can help resolve software issues right out of the box. In this case, it turns out Circuit Cellar readers’ distaste for user manuals led the way for sweeping change.
As the electronics market continues toward a commodity market, the cost
associated with a design has shifted from hardware to software. With the
reduction in cost of the hardware, more concern is being placed on the design
and sustainability of the application. In many cases, the development time for
the software component has exceeded 75 % of the design cycle and, as such, the
cost associated with the development of an application has rooted itself in
software.
Knowing this, it’s no wonder that most engineers consider the selection
of development tools and the access to predeveloped code, such as drivers,
middleware, and sample projects, as a primary decision factor when selecting a
processor vendor. Therefore, after suitable microcontrollers (MCUs) or microprocessors
(MPUs) have been identified, the focus of attention shifts to the software
tools, drivers, sample code, etc., all of which must be evaluated before the
best device for the system can be selected. A project’s success truly depends
not only on the quality of the code produced for the embedded design, but also
on the quality and usability of the software tools used by the engineers who
develop that code.
To engineers, having access to information about the latest tools and
software is obviously critical in making their embedded project successful.
Historically, to do this, hardware companies sent an engineer to the customer’s
site to discuss and demonstrate the capabilities of devices and support
products—a manpower-intensive solution that has to be scheduled in advance, but
works well. Now, there is an alternative method to enable engineers to get a
great, self-guided interactive demo/tutorial, conveniently, whenever it is best
for them to do so.
-SPB White Paper, Renesas Technology America
Analysis of the Innovative Subatomic Particle Demonstration Platform. Click here to download this white paper from Renesas Technology America, Inc.
Microchip Technology Gears Up for Latest Circuit
Cellar Sample Distribution
Is USB PIC in your
future?
It’s no secret that Circuit Cellar runs a number of design contests each year. Historically, these contests have included significant quantities of free hardware samples made available to qualifying Circuit Cellar readers. Fortunately, Circuit Cellar readers are really good at utilizing these samples in unique projects and passing along many new applications. The result: Microchip Technology is gearing up to distribute a wide variety of its USB-ready PIC MCUs to qualifying Circuit Cellar readers.
How to make sure you’re notified about availability:
If you’ve recently submitted a
project to a Circuit Cellar design contest, you’ll be first on the list to
receive notification of premium sample distributions. You’ve proven that you
have what it takes to convert a sample into project worthy of attention.
If you’re a print magazine
subscriber and new to Circuit Cellar, or if you haven’t yet had a chance to
participate in past design contests, then you may want to make sure you’re a
part of Circuit Cellar’s Design Notification Network (DNN). The DNN system will
keep your e-mail on file for early notification about important sample
distribution efforts. To sign up for the DNN, complete and submit the application
form found at http://www.circuitcellar.com/network/. Please note that Electronic Edition subscribers who
haven’t opted out of Circuit Cellar’s e-mail notifications will automatically
receive e-mail notification of these sample programs.
To learn more about some of the
devices associated with this upcoming sample distribution, please visit Microchip USB Documentation.
Catching Up with Circuit Cellar Author, Design Contest
Winner Jens Altenburg
From Sophocles to
UAVs
Jens’s Introduction
to Circuit Cellar
I got in touch with Circuit
Cellar in 2001. A colleague of mine drew my attention to the 2001 Texas
Instruments MSP430 Ultra-Low Power Flash MCU Design Contest. It was the first
time I tried to design an application for an international competition. The
result overwhelmed me completely. I won Third Place in the contest. It could be
possible that the "contest virus" infected me totally at this
time.
Embedded
Perspective
I've worked with microcomputers for more
than 25 years. At the beginning, I had to use separate chips for CPU, RAM, and
ROM. Now, I work nearly exclusively with MCUs in the range from 8 to 32 bits.
I don't know exactly when I built my first
computer. I think it was during my last year (or the year before) at school. A
teacher presented a Z80 microprocessor to me. It was horribly expensive and I
had a lot of respect for the "intelligent" silicon. It had no
external RAM. Only a small EPROM with a capacity of 256-byte ROM was available.
The design was built on a breadboard, and I had to connect each pin of the CPU,
ROM, and external components with single wires. Because I didn't have a chance
to use a development system, I programmed the code manually into the EPROM. The
system had nearly no functionality, but it ran.
My first fully operating design was based
on the legendary Z8. The board, developed in 1991, was equipped with external
RAM, an RS-232 interface, and eight pulse-width-modulated outputs for servos.
The design could be programmed using BASIC (located in a special ROM version of
the Z8) as well as directly in assembly language. Clocked with an 8-MHz quartz, the board had a power consumption of about
1.5 W. The original prototype is still in service.
On Robotics
In 1978 a competition took place in Sömmerda,
Germany. All the local schools sent their representatives showing their ideas.
I took part with a hard-wired, light-beam-following robot. It was a very simple
system. The robot turned at its starting point as long as it detected the light
beam. With the help of two phototransistors, the robot followed the beam.
Unfortunately, I don’t have any pictures of the robot.
My first MCU-controlled robot was the
ROBOTECH 1. The robot was also programmable in BASIC. Because I used an STMicroelectronics ST9 microcontroller, I had to redesign
the original Z8 BASIC source code. The robot was fitted with two stepper motors,
some optical sensors, and a powerful ultrasonic range measurement system. It
was shown at the Embedded World tradeshow in Nuremberg in 1996.
One year later, another robot called Willi
was designed. The small vehicle (100 mm × 100 mm × 50 mm) was driven by two DC
micromotors and fitted with two phototransistors, a microphone, and a bumper.
The robot’s “operating system” supported a few specialized commands like STEP
and TURN for control purposes. The user didn't have to know complex information
about a programming language to experiment with this robot.
Robotics projects fascinate me because so
many things have to come together to make a design work. You need hardware,
software, and a little bit of imagination. But let me note that robotics is not
my only interest. I have a wonderful family. My wife is an English teacher and
my son drives my car as often as he can. I read lots of books and I like it to
discuss problems—robotics-related and more—with friends and partners at
different places all over the world.
Planning for an
Autonomous Flying Robot
Over the years, I've come to understand
that there is a huge difference between laboratory conditions and the real
world. Experiments working well under predefined conditions often fail in the
field. Practical knowledge has also shown that a lot of the necessary equipment
for small-sized robot systems is not available as off-the-shelf components.
Because I don't have support
from a huge company or enough money from lottery prizes to realize my ideas, it
is absolutely necessary to work in a step-by-step fashion to avoid failures.
When designing flying systems, this is a very important thing. You can imagine
that a lot of tests had to be made, and some of these test results will involve
crashes.
For minimizing the development risks and
costs, I decided to divide the project into different parts. In the first step
I developed the base station. With that piece of hardware it is possible to
program the skeleton of the software and check some basic stabilization
algorithms. The use of a highly sophisticated model aircraft simulator has
reduced field tests as well as the costs for expensive model aircrafts. In the
next step, I'll check some mathematical algorithms for controlling the
simulated aircraft.
Groups of students and engineers that
develop micro UAVs are located on the Internet. Most of them use only one type
of model aircraft. In that case, I think it will be difficult to use the
results in a general way. Even small changes need a lot of extra developing
time. Testing other models or modifications with the simulator first should be
the better way.
At the moment, I’m working with the
prototype of the base station. A powerful ARM9 core is the heart of the
hardware. There are lots of peripheral interfaces on board, such as UART, PWM,
I2C, SPI, and timers. The flight simulator communicates via a serial interface
with the base station. The simulator accepts pulse width-modulated signals as
an input. So, it's easy to create a complete control loop. Simulation data (e.g.,
speed, altitude, and attitude) are sent to the base station and calculated
control signals come back to the simulator.
In the background of the base station,
software runs a BASIC interpreter. “Background” means that the real-time
behavior of the autopilot isn't affected by BASIC programs. There are two
reasons for implementing this extra software.
With the help of user calls, the BASIC
interpreter enables access to internal data or parameters of the autopilot
software. That makes debugging easier.
For adapting the autopilot software to
other models, a number of parameters of the control software have to be
changed. I hope it will be possible to make these modifications with the help
of BASIC programs. In this case, the autopilot doesn't need any changes to
adapt to different models. This feature makes the project very interesting.
The flight simulator (www.reflex-sim.de) offers the
possibility to create your own models. So, together with the base station, a
semiprofessional simulation system is a reality.
My BASIC
interpreter
The BASIC interpreter was written in ANSI-C
by my brother Uwe Altenburg. Originally, the interpreter was designed for
beginners. My first MCU project that used Z8 had exactly 19(!) registers for
controlling timer, interfaces and core behavior. Now, a standard microcontroller
can have hundreds of internal registers. What newcomer or beginner does have a
chance to overlook that?
The BASIC interpreter consists of two
parts, the PC-based IDE and the run-time kernel. The IDE manages source code
editing, syntax check, and software download. Smaller micros are connected via
serial interface to the IDE. Microcontrollers with SD card support can handle
BASIC programs that are precompiled and stored at the card.
Another essential point is the use of
different microcontrollers without changing existing BASIC programs. Because
the run-time kernel is written in ANSI-C, it can be compiled for a lot of
different cores. The ARM9 compilations need about 32 KB of ROM and less than 2 KB
of RAM.
A detailed description and the PC-based IDE
are available for free at www.tinybasic.de.
Looking Forward
Aside from my usual business, my main focus
is set on finishing my flight simulation system. Afterwards real flight tests
are the next steps.
It might sound a little bit freaky, but I
could imagine that it would be possible to develop a small civilian unmanned
aircraft that could be found flying around the world. In that case, one of my
most ardent wishes is to be a part of the design team of such a UAV.
About
Jens Altenburg
Jens Altenburg grew up in East Germany. He
explains that most of his lifetime was spent in his hometown of Sömmerda, where the closest big city is Erfurt, the
capital of Thuringia. The region has a remarkable history. It was the birth and
work place for Martin Luther, Johann A. Roebling, and Carl Zeiss.
In 1985, Jens went to study at the
Technical University of Ilmenau, finishing his degree in 1989, the year of the
German revolution.
Today, Jens works for a mid-sized company
in the research and development department, helping to design special equipment
for industrial customers. He has won numerous Circuit Cellar design contest awards
and is a recurring Circuit Cellar contributing author. Click the following link
for a sample of his work: Image Processing
for Robots
Target: Sensors
Expo
Circuit Cellar representatives attend industry events that are most aligned with hands-on embedded developers. The Sensors Expo, held in Rosemont, Ill, has started to draw in exhibitors that many Circuit Cellar readers might find noteworthy. Here is a brief selection of companies that captured our attention.
www.thisisant.com – Ultra low-power wireless
www.cymbet.com – Energy harvesting & battery systems to make the most of wireless sensor systems
www.greenpeak.com – Ultra low-power wireless
www.jennic.com – Low-cost wireless sensor networks
www.rlpenergy.com – Energy harvesting, sensors
www.advancedcerametrics.com – Ceramic fibers for energy harvesting and sensor systems
What’s Your
Favorite Circuit Cellar Article?
(Can it beat an article about barking like a dog to outsmart an HCS alarm system?)
After more than 20 years and thousands of design projects, there have been plenty of Circuit Cellar articles that have topped our buzz list. Sometimes it’s the style of the writer or the unique application that garners some attention; other times, it’s an invaluable insight from an expert in the field. Whatever your reasoning, Circuit Cellar wants to know your all-time favorite Circuit Cellar magazine article. The winning article will be posted on our home page as your gift to the embedded community.
Make your comments here: http://www.circuitcellar.com/comments/
And while you’re at it, enjoy a favorite from Steve Ciarcia’s earlier HCS projects: The Circuit Cellar Home Control System II
Entrants to the Microchip Technology 16-Bit Embedded Control Design Contest were asked to select the Distinctive Excellence project that impressed them the most. Congratulations go to George Adamidis of Greece who won the Contestants’ Choice Award for his project, Audio Spectrum Analyzer. You can embed this digital audio frequency spectrum analyzer in an audio device or use it as a stand-alone unit. The real-time system computes the distribution of the audio signal energy to 20 specific frequency bands and displays it on a 20 × 20 LED display.
