SIGNALsolutions - simpson.net.au

Transcription

CIRCUIT CELLAR | ISSUE 291 | OCTOBER 2014
SIGNAL PROCESSING
OCTOBER 2014
ISSUE 291
circuitcellar.com
SIGNAL solutions
.....
..... plan & construct a real-time
wireless laser tag system .....
long-distance wireless
radio control .....
tips for selecting &
caring for RF connectors ..... embedded audio &
signal recognition .....
Solar Power Revisited | Q&A: Design & EE Outreach
BlasterWorks Project
circuitcellar.com
| LMR Wireless Control | AC Monitor | ComputerScope Test Instrument
Capacitors 101 | Build a Linux System | Tips for Working with
RF Connectors | Embedded Voice Recognition
The Future for TXCOs
MIXED SIGNAL
OSCILLOSCOPES
4 ANALOG + 16 DIGITAL CHANNELS
RAPIDLY DEBUG COMPLEX MIXED SIGNAL DESIGNS
• USB 3.0
• ULTRA DEEP MEMORY
• SEGMENTED MEMORY
RAPID
• RA
R
PID TRIGGERS
INCLUDES AUTOMATIC MEASUREMENTS, SPECTRUM ANALYZER, SDK,
ADVANCED TRIGGERS, COLOR PERSISTENCE, SERIAL DECODING
(CAN, LIN, RS232, I²C, I²S, FLEXRAY, SPI), MASKS, MATH CHANNELS,
ALL AS STANDARD, WITH FREE UPDATES . 5 YEAR WARRANTY
3204D MSO
Channels
3205D MSO
3206D MSO
3404D MSO
2 analog, 16 digital
3405D MSO
3406D MSO
4 analog, 16 digital
Bandwidth
60 MHz
100 MHz
200 MHz
60 MHz
100 MHz
200 MHz
Buffer memory
128 MS
256 MS
512 MS
128 MS
256 MS
512 MS
Max. sampling rate
Signal generator
Digital inputs
1 GS/s
Function generator + Arbitrary waveform generator
100 MHz max. frequency, 500 MS/s max. sampling rate
www.picotech.com/pco527
2
CIRCUIT CELLAR • OCTOBER 2014 #291
EDITOR’S LETTER
Issue 291 October 2014 | ISSN 1528-0608
CIRCUIT CELLAR® (ISSN 1528-0608) is published monthly by:
Circuit Cellar, Inc.
111 Founders Plaza, Suite 300
East Hartford, CT 06108
Periodical rates paid at East Hartford, CT, and additional offices.
One-year (12 issues) subscription rate US and possessions
$50, Canada $65, Foreign/ ROW $75. All subscription orders
payable in US funds only via Visa, MasterCard, international
postal money order, or check drawn on US bank.
SUBSCRIPTIONS
Circuit Cellar, P.O. Box 462256, Escondido, CA 92046
E-mail: circuitcellar@pcspublink.com
Phone: 800.269.6301
Internet: circuitcellar.com
Address Changes/Problems: circuitcellar@pcspublink.com
Postmaster: Send address changes to
Circuit Cellar, P.O. Box 462256, Escondido, CA 92046
ADVERTISING
Strategic Media Marketing, Inc.
2 Main Street, Gloucester, MA 01930 USA
Phone: 978.281.7708
Fax: 978.281.7706
E-mail: circuitcellar@smmarketing.us
Advertising rates and terms available on request.
New Products:
New Products, Circuit Cellar, 111 Founders Plaza, Suite 300
East Hartford, CT 06108, E-mail: newproducts@circuitcellar.com
HEAD OFFICE
Circuit Cellar, Inc. 111 Founders Plaza, Suite 300
Phone: 860.289.0800
COVER PHOTOGRAPHY
Chris Rakoczy, www.rakoczyphoto.com
COPYRIGHT NOTICE
Entire contents copyright © 2014 by Circuit Cellar, Inc. All
rights reserved. Circuit Cellar is a registered trademark of
Circuit Cellar, Inc. Reproduction of this publication in whole
or in part without written consent from Circuit Cellar, Inc. is
prohibited.
ELECTRICAL ENGINEERING MASH-UP
This issue is an electrical engineering content mash-up comprising new technologies,
revisited technologies, and even recent spins on old engineering concepts and projects.
We revisit a classic solar power installation. We present a modern take on laser gun
gaming, which first became popular in the 1980s. We highlight new uses for wireless
mobile radio, which has been used for years by public service providers such as police
departments and emergency medical teams. And much more. Let’s go over the specifics.
I’m excited to announce that we’re re-releasing Circuit Cellar founder Steve Ciaria’s
three-part series, “Solar-Powering the Circuit Cellar.” Turn to page 6 for an excerpt from
the first article in the series. You can download the entire series for free for a limited time
at http://bit.ly/1tHMOQW.
Starting on page 8 is an interview with Bill Porter, a Navy engineer who runs a
technical outreach program in his spare time. We’re sure you’ll be amazed by the ways in
which he incorporated engineering in his wedding ceremony.
If you’re interested in laser technology, check out Max Rothstein and Richard
Speranza’s article about their BlasterWorks project (p. 18). They designed a laser tag
system that uses real lasers.
Working on a wireless control application? If so, check out long-time Circuit Cellar
author Ken Merk’s article on Land Mobile Radio (LMR, p. 30). He explains how to build a
wireless interface with eight output arrays and eight digital inputs.
If you need a simple way to determine voltage values, check out the circuit Larry
Cicchinelli presents on page 40. With the circuit and a 10-segment LED display, he can
quickly tell if voltage is high, nominal, or low.
On page 44 George Novacek presents another handy review of an essential electrical
engineering topic: capacitance. He presents common capacitor types and explains basic
capacitance-related equations.
Like Novacek, Bob Japenga presents a short review of an essential topic: embedded
Linux (p. 48). This month he offers tips for building your Linux system.
Turn to page 52 for a special feature from our friend and colleague, Jan Buiting. In this
“retronics” article, he reviews how a young engineer built an oscilloscope add-on for a PC.
How do you select the right RF connectors for a project? How do you keep your delicate
connectors in proper working order? On page 58 Robert Lacoste explains how to properly
select and care for your RF connectors.
On page 64 Jeff Bachiochi begins a series on embedded voice recognition. This month
he introduces the topic and covers the basics.
Engineer John Esterline wraps up the issue with his thoughts on the future of
temperature-compensated crystal oscillators. Visit our social pages to share your
thoughts on TCXOs.
C. J. Abate
cabate@circuitcellar.com
DISCLAIMER
Circuit Cellar® makes no warranties and assumes no
responsibility or liability of any kind for errors in these
programs or schematics or for the consequences of any
such errors. Furthermore, because of possible variation in
the quality and condition of materials and workmanship of
reader-assembled projects, Circuit Cellar® disclaims any
responsibility for the safe and proper function of readerassembled projects based upon or from plans, descriptions, or
information published by Circuit Cellar®.
The information provided by Circuit Cellar® is for educational
purposes. Circuit Cellar® makes no claims or warrants that
readers have a right to build things based upon these ideas
under patent or other relevant intellectual property law in
their jurisdiction, or that readers have a right to construct or
operate any of the devices described herein under the relevant
patent or other intellectual property law of the reader’s
jurisdiction. The reader assumes any risk of infringement
liability for constructing or operating such devices.
© Circuit Cellar 2014 Printed in the United States
THE TEAM
EDITOR-IN-CHIEF
C. J. Abate
ART DIRECTOR
KC Prescott
ADVERTISING COORDINATOR
Kim Hopkins
PRESIDENT
Hugo Van haecke
COLUMNISTS
Jeff Bachiochi (From the
Bench), Ayse K. Coskun
(Green Computing), Bob
FOUNDER
Japenga (Embedded
Steve Ciarcia
in Thin Slices), Robert
PROJECT EDITORS
Lacoste (The Darker
Chris Coulston, Ken
Davidson, and David
Tweed
Side), Ed Nisley (Above
the Ground Plance),
George Novacek (The
Consummate Engineer),
and Colin O’Flynn
(Programmable Logic in
Practice)
CUSTOMER SERVICE
Debbie Lavoie
circuitcellar.com
US/UK
Don Akkermans
+31 46 4389444
d.akkermans@elektor.com
OUR NETWORK
ELEKTOR LABS
Wisse Hettinga
+31 46 4389428
w.hettinga@elektor.com
GERMANY
Ferdinand te Walvaart
+49 241 88 909-17
f.tewalvaart@elektor.de
FRANCE
Denis Meyer
+31 46 4389435
d.meyer@elektor.fr
NETHERLANDS
SUPPORTING COMPANIES
2015 International CES
33
Ironwood Electronics
78
Harry Baggen
+31 46 4389429
h.baggen@elektor.nl
Accutrace7
Lauterbach55
SPAIN
All Electronics Corp.
78
Lemos International Co., Inc.
79
ARM TechCon 2014
21
MaxBotix, Inc.
79
CadSoft Computer GmbH
25
microEngineering Labs
79
Custom Computer Services
78
Micro Digital, Inc.
35
Eduardo Corral
+34 91 101 93 95
e.corral@elektor.es
ITALY
Maurizio del Corso
+39 2.66504755
m.delcorso@inware.it
SWEDEN
Wisse Hettinga
+31 46 4389428
w.hettinga@elektor.com
Elektor
28, 29
MyRO Electronic Control Devices, Inc.
78
Elektor
56, 57
NetBurner, Inc.
C2
Elektor71
Pico Technology
1
João Martins
+31 46 4389444
j.martins@elektor.com
Elektor
R.E. Smith, Inc.
43
INDIA
Elprotronic, Inc.
39
Reach Technology, Inc.
78
EMAC, Inc.
39
RoboBusiness Conference
47
ExpressPCB23
Saelig Co., Inc.
43
Humandata, Ltd.
78
Technologic Systems
Zeynep Köksal
+90 532 277 48 26
zkoksal@beti.com.tr
IAR Systems
11
Triangle Research International, Inc. SOUTH AFRICA
Imagineering, Inc.
C4
BRAZIL
João Martins
+31 46 4389444
j.martins@elektor.com
PORTUGAL
Sunil D. Malekar
+91 9833168815
ts@elektor.in
RUSSIA
Nataliya Melnikova
+7 965 395 33 36
elektor.russia@gmail.com
TURKEY
Johan Dijk
+31 6 1589 4245
j.dijk@elektor.com
CHINA
Cees Baay
+86 21 6445 2811
ceesbaay@gmail.com
76, 77
15, 37
79
NOT A SUPPORTING COMPANY YET?
Contact Peter Wostrel (circuitcellar@smmarketing.us, Phone 978.281.7708, Fax 978.281.7706)
to reserve your own space for the next edition of our members’ magazine.
3
4
CONTENTS
OCTOBER 2014 • ISSUE 291
SIGNAL PROCESSING
THE BLASTERWORKS REAL-TIME LASER TAG SYSTEM
CC COMMUNITY
06 : CC WORLD
08 : QUESTIONS & ANSWERS
Electronics Innovation & Outreach
An Interview with Bill Porter
A Florida-based Navy engineer on unmanned system
design and engineering-focused educational outreach
INDUSTRY & ENTERPRISE
12 : PRODUCT NEWS
17 : CLIENT PROFILE
Newhaven Display International (Elgin, IL)
FEATURES
18 : BlasterWorks
A Real-Time Wireless Laser Tag System
By Max Rothstein and Richard Speranza
A DIY real-time laser tag system complete with an LCD
screen and RF transceiver for scoring and configuration
30 : Land Mobile Radio Wireless Control
By Ken Merk
Build a Land Mobile Radio (LMR) wireless system for
long-distance monitoring and control
40 : AC Monitor
By Larry Cicchinelli
A handy circuit if you need an easy-to-use,
straightforward way to determine voltage values
52 : Elektor ComputerScope (1986)
Goodbye Green-Only 'Scope CRT
By Jan Buiting
A look back at a young engineer’s ambitious test
instrument, which was like a ‘scope with PC support
COLUMNS
44 : THE CONSUMMATE ENGINEER
The Not So Humble Capacitor
BILL PORTER DESIGNING AT HIS WORKBENCH
By George Novacek
A look at common capacitors and a variety of basic
capacitance-related equations
circuitcellar.com
CONTENTS
LAND MOBILE RADIO FOR LONG-DISTANCE MONITORING & CONTROL
48 : EMBEDDED IN THIN SLICES
Linux System Configuration (Part 3)
Build Your Linux System
By Bob Japenga
Various ways to build your embedded Linux system
58 : THE DARKER SIDE
RF Connectors
By Robert Lacoste
Tips for selecting and caring for RF connectors
64 : FROM THE BENCH
Embedded Voice Recognition (Part 1)
Recognition as a Peripheral
CONNECTOR CARE
TESTS & CHALLENGES
74 : CROSSWORD
75 : TEST YOUR EQ
TECH THE FUTURE
80 : The Future of Temperature-Compensated
Crystal Oscillators
By John Esterline
A look at TCXO technology and the artificial neural
network’s potential
By Jeff Bachiochi
Add effective speech and voice recognition capabilities
to an electronics system
EMBEDDED VOICE RECOGNITION
AIR TUNING CAPACITOR
@editor_cc
@circuitcellar
circuitcellar
5
6
CC WORLD
FREE SOLAR SERIES
By CC Staff (US)
COMMUNITY
Various Connecticut, USA-based CC staffers have noticed
new local television ads promoting home solar power
systems. Many towns in our area are even promoting solar
programs with incentives for homeowners to switch to
solar. We aren’t quite sure why there has been a push for
solar during the past several weeks (in comparison to, say,
three months ago), but it’s been quite obvious.
In the spirit of DIY engineering and solar power
innovation, we’re re-releasing Circuit Cellar founder Steve
Ciaria’s three-part series, “Solar-Powering the Circuit
Cellar.” An excerpt from the first article in the series
appears below. You can download the entire series for free
for a limited time (http://bit.ly/1tHMOQW). Enjoy!
Photo 1 shows the roof-mounted solar panels
that produce approximately 40% of the total PV
power. I know it sounds like a joke that the first PV
system consideration is walking around the house
and looking for the sun, but you can’t generate much
energy if your panels are always shaded. When you
live in the middle of the woods, finding the sun is
often easier said than done.
Array orientation determines how much energy
you can produce. Solar panels are typically aimed
due south at a specific tilt angle that optimizes the
incidence angle of the sunlight striking the panel.
Maximum energy is produced when this tilt angle
is equal to the latitude of the location (reduced by a
location correction factor). Typically, the optimal tilt
angle during the summer is the latitude minus 15°,
and the optimal angle for the winter is the latitude
plus 15°. Hartford, CT, is located at 42° latitude
and the optimum tilt angle (minus an 8° correction
factor) ranges from 19° in the summer (34° – 15°)
to 49° in the winter (34° + 15°). The Connecticut
rebate program suggests that if a fixed tilt is used,
it be set at 35°. Of course, these are computergenerated optimizations that don’t necessarily
accommodate real-world conditions. While it
Photo 2
Attaching the first SunPower SPR-210 panel to the UniRac mounting rails on the roof
Photo 1
Approximately 4,200 W of PV power is generated from 20 roof-mounted SunPower SPR-210
solar panels. The other 6,560 W comes from pole-mounted arrays behind this area.
requires some nontrivial computer calculations to
show authenticity, it is my understanding that as
long as the non-optimal differences in azimuth and
tilt are less than 20°, the loss in maximum power
production is typically only about 5%. It is exactly
for that reason that the most cost-effective PV
installation is typically a fixed-pitch roof-mounted
array.
My system includes both variable and fixedpitch arrays. The roof-mounted panels are located
on the solarium roof and oriented at a fixed pitch of
17.5° facing SSW (see Photo 1). According to Sunlight
Solar Energy’s calculations, efficiency is still about
92% of the desired maximum because the 17.5° roof
angle actually allows higher efficiency during longer
summer hours even though it isn’t the optimum tilt
for winter.
Pole-mounted arrays are more efficient than a
fixed-pitch roof array by design. My configuration is
single-axis adjustable. The pole-mounted arrays are
oriented due south and enable seasonal adjustment
in the tilt angle to optimize the incidence angle of
the sun. For everyone ready to e-mail me asking
why I didn’t put in a tracking solar array since this
is a pole mount, let me just say that you can also
send me financial contributions for doing it via the
magazine.—Steve Ciarcia, “Solar-Powering the Circuit
Cellar (Part 1: Preparing the Site),” Circuit Cellar 209,
2007.
You can download Steve
Ciarcia’s three-part article series, “Solar-Powering
the Circuit Cellar,” for free
(for a limited time only):
http://bit.ly/1tHMOQW.
8
QUESTIONS & ANSWERS
Electronics Innovation
& Outreach
An Interview with Bill Porter
COMMUNITY
Bill Porter is a Panama City Beach, FL-based electronics engineer working for the US Navy. When he isn’t working on unmanned systems for
the Navy, he spends his time running an engineering-focused educational outreach program and working on his own projects. In this interview,
Bill talks about his first designs, technical interests, and current projects.
CIRCUIT CELLAR: You’re an electronics
engineer for the United States Navy. Can
you describe any of the projects you’re
involved with?
Bill Porter at his home workspace
BILL: I work with unmanned systems, or robots
that are teleoperated and/or autonomous.
This includes systems that swim under water,
on the surface, or across the land. The Navy is
working hard to develop robots to do the jobs
that are dirty, dangerous, or dull and help
keep the sailor out of harm’s way. One such
system is called MUSCL, or Modular Systems
Craft Littoral. MUSCL is a small, man-portable
surface vehicle that is used by Riverine Patrol
for remote surveillance and reconnaissance.
I was the lead electrical engineer for the
project.
Besides robots, I am also working on a
few education outreach programs that work
towards getting more students interested in
STEM careers.
CIRCUIT CELLAR: Tell us about The Science
Brothers nonprofit outreach program. How
did the program start?
BILL: The Science Brothers is my main
educational outreach program run out of my
Navy base. For two Fridays every month, a
few of my coworkers and I will visit a local
elementary school to put on a show. The
script of the show centers on the dynamics of
two brothers, who specialize in different fields
and argue over whose science is “cooler.”
The result is a fun and wacky trip exploring
different premises in science, such as light,
sound, and energy, with examples and
demonstrations from the realms of chemistry,
physics, and electricity.
The program restarted when a few
coworkers and I sat down and decided to bring
back an old program that had existed on the
base in the ‘90s called “Dr. Science.” The goal
of the program was to bring science-based
experiments to the schools using equipment
they otherwise were not able to afford. By
wowing the students with the spectacularlooking demos, we get them excited about
science and yearning to learn more.
circuitcellar.com
9
QUESTIONS & ANSWERS
CIRCUIT CELLAR: Your website (BillPorter.
info) includes projects involving 3-D
printing, motor controllers, and LEDs. What
types of projects do you prefer working on
and why?
CIRCUIT CELLAR: Do you have a favorite
project?
BILL: Yes! My wedding of course! I married
the girl of my dreams who is just as much as
a geek as I am, and as a result, we had an
extremely geeky wedding over a year in the
making involving many projects throughout.
So much so that our theme was “Circuit and
Swirls” and we carried the motif throughout.
We designed and made our own wedding
invitations involving LEDs, a microprocessor,
and a clever Easter egg. Furthermore, we
3-D-printed our centerpieces, built up our own
“e-textile” wedding attire with LEDs and EL
wire, and we even had a “soldering ceremony”
during the event. It made our parents nervous,
but in the end, everyone had a good time. Did
I mention I asked her to marry me on a PCB
she designed for a project?
CIRCUIT CELLAR: Do you have a favorite
Science Brothers project?
BILL: My favorite Science Brother project has
to be Tesla Hero. I took apart a Playstation2
Guitar Hero controller and modified it to send
signals over a fiber optic cable to my Tesla
Coil. Instead of hitting digital rifts, now I
can play a few notes through my Tesla Coil.
Nothing gets the kids more excited than to see
the “Million Volt Electric Guitar.”
CIRCUIT CELLAR: Give us a little background
information. How did you become
interested in technology? Where did you
do to school? What did you study?
BILL: I studied electrical engineering in college
and build robots for fun. I spent way too much
COMMUNITY
BILL: I am a hardware guy. I love to fire up my
favorite PCB CAD software just to get an idea
out of my head and on the screen. I do not
breadboard very often, as I would rather take
my chances trying some new idea on a board
first. Either it works, or I have an excuse to
design another PCB. Thankfully, group-order
PCB services have enabled my addiction
tinkering at a very low cost. I wish I was
stronger at mechanical CAD design to really
get the full potential out of my 3-D printer, but
I have done well enough without it. It really
does come in handy at times, whether it is a
quick project enclosure, a mount, or a part for
our garden.
Images from Bill Porter's EEthemed wedding
10
QUESTIONS & ANSWERS
of my allowance on the first Lego Mindstorms kits in
high school (which got me hooked on robotics) and
liked to mod Xboxs for fun. If there was a way to
attach an LED to something, I would find it.
CIRCUIT CELLAR: Tell us about the first embedded
system you designed. What did you learn from
the experience?
COMMUNITY
BILL: I guess my first real embedded project was
the robot I helped design from the ground up during
the capstone project of my last year in college. In a
team of four, we had to design a robot that would
autonomously locate, sort, and store recyclable
“trash” consisting of Coke can, plastic, and glass
bottles. The project was a smashing success and
lead to my career with the US Navy. The experience
taught me about how much involvement is required
to take a project from an idea to the final product.
There are a lot of little things for which one has to
contemplate, plan and design solutions.
CIRCUIT CELLAR: Tell us about last electronicsdesign related product (e.g., a microcontroller,
sensor, etc.) you purchased and the project you
worked on with it.
BILL: My wife and I just decided to start a small herb
garden on our apartment balcony. Never missing an
excuse to tinker, I started playing with designs for an
automatic gravity-fed drip irrigation system. I just
purchased a few soil conductivity sensors to sense
when the plants will need to be watered. I’m hoping
to automate and data log things like soil moisture vs.
relative humidity, temperature, sunlight level, etc.
CIRCUIT CELLAR: Are you currently working on
or planning any projects? Can you tell us about
them?
BILL: I have one main project that is taking up all my
time at work and at home. A coworker and I are the
technical directors for the first-ever Maritime RobotX
Challenge. The challenge, sponsored by Association
for Unmanned Vehicle Systems International (AUVSI)
Foundation and by the Office for Naval Research
(ONR), will take place this October in Singapore. It
will include 15 teams of college students from five
participating nations and put them to the test by
challenging them to design a robot that will complete
five tasks autonomously. As one of the technical
directors, I have been helping design and build the
interactive course elements that the teams’ robots
will be facing. Find out more at Robotx.org.
CIRCUIT CELLAR: What new technologies excite
you and why?
Science Brothers shows
BILL: I have always been infatuated by LEDs and
ways to conserve energy, so I am most excited to see
how efficient LEDs are starting to take over as the
new source of light in the household.
C-RUN is a high-performance runtime analysis add-on
product, fully integrated with world-leading
C/C++ compiler and debugger tool suite
IAR Embedded Workbench.
C-RUN performs runtime analysis by monitoring applica-
tion execution directly within the development environment.
The tight integration with IAR Embedded Workbench improves
development workflow and provides each developer with access to
runtime analysis that is easy-to-use.
Join us at ARM TechCon!
Booth 512
12
PRODUCT NEWS
UPGRADES TO HIGH-ISOLATION, MULTI-THROW SWITCHES
Peregrine Semiconductor Corp. recently
introduced the UltraCMOS PE42442 and
PE42452, which are high-isolation, multithrow switches. In comparison to its
PE42451, the SP4T and SP5T switches include
several additional features to meet wireless
infrastructure demands:
• extended temperature range to 105°C
• control logic support of 1.8 and 3 V
• a 1.8-V logic control was included to
enable lower power consumption
• a wider power supply range of 2.3V
to 5.5V
• loss, isolation, and ESD protection
A 24-lead, 4 × 4-mm QFN package is
available. For orders of 10,000 or more, the
PE42442 costs $1.60 each and the PE42452
costs $1.92 each. Evaluation kits are also
available.
Peregrine Semiconductor
www.psemi.com
NEW PIC32 BLUETOOTH STARTER KIT
Microchip Technology recently announced the new PIC32 Bluetooth Starter Kit, which is
intended for low-cost applications such as a Bluetooth thermostat, wireless diagnostic tools, and
Bluetooth GPS receivers. According to Microchip, the kit includes “a PIC32 microcontroller, HCIbased Bluetooth radio, Cree high-output multi-color LED, three standard single-color LEDs, an
analog three-axis accelerometer, analog temperature sensor, and five push buttons for user-defined
inputs.”
PICkit On Board (PKOB) eliminates the need for an external debugger/
programmer, USB connectivity, and GPIOs for rapid
development of Bluetooth Serial Port Profile (SPP), USB and
general-purpose applications. The starter kit also features a
plug-in interface for an audio CODEC daughter card. The kit’s
PIC32MX270F256D microcontroller operates at 83 DMIPS with
256-KB flash memory and 64-KB RAM.
The PIC32 Bluetooth Starter Kit is supported by Microchip’s
free MPLAB X IDE and MPLAB Harmony Integrated Software
Framework. Additionally, the free Quick Start Package is available
with an Android application development environment. It also
includes a free SDK with the application source code and binary
for Microchip’s Bluetooth SPP library. Both are optimized for the
on-board PIC32 MCU and are available for free at www.microchip.
com/get/1AVL.
The PIC32 Bluetooth Starter Kit costs $79.99.
Microchip Technology
www.microchip.com
circuitcellar.com
13
PRODUCT NEWS
DUAL STEP-DOWN REGULATOR
•
•
•
•
Dual step-down outputs: 1 A per channel
VIN range: 2.7 to 17 V
VOUT range: 0.6 V to VIN
Up to 95% efficiency
•
•
•
•
•
•
•
•
•
No-load IQ = 5 μA (both channels enabled)
< 4 μA (one channel enabled)
High efficiency, low dropout operation
(100% duty cycle)
Constant frequency (1 MHz/2.25 MHz)
with external synchronization
±1% output voltage accuracy
Current mode operation improves line
and load transient response
Phase shift programmable with external
clock
Selectable current limit
Internal compensation and soft-start
Compact 14-pin DFN (3 mm × 4 mm)
package
The regulator is available now with a per-unit
cost starting at $3.75 for orders of 1,000 units.
Linear Technology
www.linear.com
HIGH-SENSITIVITY
ACCELEROMETER FAMILY
Freescale recently introduced a new range of
three-axis accelerometers offering high sensitivity
at low power consumption. According to Freescale,
the FXLN83xxQ family is capable of detecting
acceleration information often missed by less accurate
sensors commonly used in consumer products such
as smartphones and exercise activity monitors. In
conjunction with appropriate software algorithms, its
improved sensitivity allows the new sensor to be used
for equipment fault prognostication (for predictive
maintenance), condition monitoring, and medical
tamper detection applications.
The 3 mm × 3 mm chip has a bandwidth of 2.7 kHz
and uses analog output signals for direct connection to a
microcontroller’s ADC input. Each chip has two levels of
sensitivity that can be changed on the fly. The complete
family covers acceleration ranges of ±2, ±4, ±8, and ±16
g, with gains of, 229.0, 114.5, 57.25, and 28.62 mV/g,
respectively. Zero g is indicated by an output level of
0.75 V.
The FXLN83xxQ family features:
•
•
•
•
FXLN83x1Q ±2 or ±8 g range
FXLN83x2Q ±4 or ±16 g
FXLN836xQ 1.1 kHz x- and y-axis bandwidth (Z = 600 Hz)
FXLN837xQ 2.7 kHz x- and y-axis bandwidth (Z = 600 Hz)
The sensors operate from 1.71 to 3.6 V (at 180 µA
typically, 30 nA shutdown). The company has also made
available the DEMOFXLN83xxQ evaluation break-out board
with a ready-mounted sensor to simplify device integration
into a test and development environment.
Freescale Semiconductor
www.freescale.com
Linear Technology has announced an addition
to its family of power regulator solutions. The
LTC3622 is a dual step-down regulator in a small 3
× 4 mm package that provides two independently
configurable 1-A outputs operating from a 2.7- to
17- V input. External voltage divider networks
define the two output voltages or alternatively a
range of fixed output voltage versions result in a
lower component count. The input voltage range
makes it suitable for operation from single or
multiple lithium cells or from a vehicular supply.
The regulator can operate in Burst mode to
give highest efficiency at light loads or PulseSkipping mode to give lower ripple noise. The
system clock can be synchronized to an external
source to help to reduce system noise bandwidth.
Some of the many features include:
14
PRODUCT NEWS
DE0-NANO CYCLONE FPGA DEV BOARD
With a DE0-Nano Cyclone FPGA Development Board,
you can create your own sophisticated hardware using
programmable logic. The development board includes an
Altera Cyclone IV and additional components to connect and
test hardware designs. It comes a pre-wired Cyclone IV FPGA
for programming and connection to internal or external
devices and circuit.
With the board, you can create sophisticated logic
hardware fairly quickly using a hardware description
language. Possible applications include dedicated digital logic
processors, robotics, and DIY autonomous systems.
Parallax
www.parallax.com
NEW DUAL-CHANNEL FUNCTION/ARBITRARY
WAVEFORM GENERATORS
Analog Devices has announced a software-defined radio
(SDR) rapid prototyping kit with dual 2 x 2 AD9361 RF
transceivers to simplify and rapidly prototype 4 × 4 MIMO
wireless transceiver applications on the Xilinx Zynq-7000 allprogrammable SoC development platforms.
The AD-FMCOMMS5-EBZ rapid prototyping kit provides
a hardware/software ecosystem solution addressing the
challenges of SDR transceiver synchronization experienced
by RF and analog designers when implementing systems
using MIMO architectures. A webinar is available on how to
synchronize multiple RF transceivers in high-channel density
applications.
The AD-FMCOMMS5-EBZ rapid prototyping
kit includes the following:
•
•
•
•
•
•
An FPGA mezzanine card (FMC)
featuring two of Analog AD9361 2 x 2
RF transceivers and support circuitry
Reference designs
Design and simulation tools for
MathWorks
HDL code
Device drivers for Zynq-7000 All
Programmable SoCs
Online support at ADI’s EngineerZone for
rapid prototyping to reduce development
time and risk.
The AD-FMCOMMS5-EBZ rapid prototyping
kit is the fifth SDR rapid prototyping kit ADI has
introduced in the last year to help customers
address the global SDR market. SDR MIMO
applications range from defense electronics and RF
instrumentation to communications infrastructure
and include active antennas, transmit beamforming,
receive angle of arrival systems, and open-source
SDR development projects.
The AD9361 operates over a frequency range of 70 MHz to
6 GHz. It is a complete radio design that combines multiple
functions, including an RF front end, mixed-signal baseband
section, frequency synthesizers, two analog-to-digital
converters and two direct conversion receivers in a single
chip. The AD9361 supports channel bandwidth from less than
200 kHz to 56 MHz and is highly programmable.
Analog Devices
www.analog.com
DESIGN YOUR SOLUTION TODAY
CALL 480-837-5200
TS-8820-BOX Industrial Controller
TS-7680 Single Board Computer
Opto-Isolated Analog IO
AC Power, Relays, WiFi, & Bluetooth
Powered by:
TS-SOCKET
Modules
TS-8820-4700
Pricing starts at
$608 $659
Qty 100
Qty 1
TS-8820-4800
Pricing starts at
$648 $699
Qty 100
Features:
Qty 1
Pricing starts at
$159 $203
Qty 100
(Shown with all options)
Qty 1
Features:
800 MHz ARM CPU
2x USB Host
454 MHz ARM CPU
Wireless 802.11b/g/n
256 MB RAM
8x Opto-Isolated Inputs
128 or 256 MB RAM
Bluetooth 4.0+EDR
256 MB Flash Storage
6x Digital Inputs, 6x Out
2 GB Flash Storage
Wireless 802.11b/g/n
1x microSD Socket
16x ADC, 4x DAC Ports
1x microSD Socket
24 VAC or 8-24 VDC Power
1x 10/100 Ethernet
PWM, Counter, & H-Bridge
2x 10/100 Ethernet
DIO, CAN, Modbus, Relays
Benefits:
Benefits:
Provides a variety of electrically isolated IO
Low power and low cost industrial grade SBC
Metal enclosure with rugged screw connectors
Rugged 24-Pin screw terminal connector
Convenient Power over Ethernet (PoE) powers SBC
Flexible power inputs including AC and DC
Customizable with programmable 5 KLut FPGA
Wireless data aquisition via WiFi and Bluetooth
-40 to 85 ºC industrial temperature range
L
NA
GI
Support every
step of the way
with open
source vision
O
RI
Embedded
systems that
are built to
endure
N
TS-7670 Low Power SBC with GPS Radio & Cell Modem
O
PE
D
GE
RU
G
LO
NG
LI
FE
Boots Linux 2.6 in less than 3 seconds
We’ve never
discontinued a
product in 30
years
30 V tolerant DIO, analog IO, 3 A relays, and more
Also Available:
Unique embedded
solutions add
value for our
customers
16
PRODUCT NEWS
USER-EXTENSIBLE FDA FOR REAL-TIME OSCILLOSCOPES
Keysight Technologies recently announced the availability of
a frequency domain analysis (FDA) option, a user-extensible
spectrum frequency domain analysis application solution for
real-time oscilloscopes.
The FDA option extends the capabilities of Keysight Infiniium
and InfiniiVision Series oscilloscopes by enabling you to acquire
live signals from the oscilloscope and visualize them in the
frequency domain, as well as make key frequency domain
measurements. Option N8832A-001 includes the application,
the application source code for user extensibility, and MATLAB
software. These tools enable you to extend an application’s
capabilities to meet their current and future testing needs.
With the FDA application, you can address a variety of FDA
challenges such as:
•
•
•
•
•
Power spectral density and spectrogram visualization
Frequency domain measurements in an application
including relevant peaks in the PSD and measurements
such as occupied bandwidth, SNR, THD, SFDR, and
frequency error
Oscilloscope configuration through the application
to allow for repeatable instrument configuration and
measurements; optionally includes additional SCPI
commands for more advanced instrument setup
Insertion of additional custom signal processing
commands prior to frequency domain visualization,
as needed, for more advanced analysis insight
Live or post-acquisition analysis of time-domain data
in MATLAB software
Keysight Technologies
www.keysight.com
QUARTUS II SOFTWARE ARRIA 10 EDITION V14.0
Micro Altera Corp. has released Quartus II software Arria 10
edition v14.0, which is an advanced 20-nm FPGA and SoC design
environment. Quartus II software delivers fast compile times and
enables high performance for 20-nm FPGA and SoC designs. You
can further accelerate Arria 10 FPGA and SoC design cycles by
using the range of 20-nm-optimized IP cores included in the latest
software release.
Altera’s 20-nm design tools feature advanced algorithms. The
Quartus II software Arria 10 edition v14.0 provides on average
notably fast compile times. This productivity advantage enables
you to shorten design iterations and rapidly close timing on 20nm design.
Included in the latest software release is a full complement
of 20-nm-optimized IP cores to enable faster design cycles. The
IP portfolio includes standard protocol and memory interfaces,
DSP and SoC IP cores. Altera also optimized its popular IP cores
for Arria 10 FPGAs and SoCs, which include 100G Ethernet, 300G
Interlaken, Interlaken Look-Aside, and PCI Express Gen3 IP. When
implemented in Altera’s Arria 10 FPGAs and SoCs, these IP cores
deliver the high performance.
The Quartus II software Arria 10 edition v14.0 is available now
for download. The software is available as a subscription edition
and includes a free 30-day trial. The annual software subscription
is $2,995 for a node-locked PC license. Engineering samples of
Arria 10 FPGAs are shipping today.
Altera
www.altera.com
NEW 40-NM MICROCONTROLLERS FOR MOTOR CONTROL
Renesas Electronics Corp. recently announced the RH850/
C1x series of 32-bit microcontrollers (MCUs), which it said are
designed for motor control in hybrid electric vehicles (HEVs) and
electric vehicles (EVs). Based on Renesas’s 40-nm process, the
RH850/C1x series features the RH850/C1H and RH850/C1M MCUs.
The new RH850/C1x devices can be used with the
RAA270000KFT RH850 family power supply management IC
(PMIC). The power management IC integrates into one device
all the power supply systems required for MCU operation, two
external sensor power supply tracks, and a full complement of
monitoring and diagnostic functions.
The RH850/C1H and RH850/C1M MCUs incorporate large
memory capacities achieved through 40 nm MONOS process
technology. The RH850/C1x series is based on Renesas’s metal
oxide nitride oxide silicon embedded flash. The RH850/C1M and
RH850/C1H devices offer memory capacities of 2 MB and 4 MB,
respectively. In addition, 32-KB data flash memory, with similar
functionality as EEPROM, is included for data storage.
The microcontrollers also feature an extensive set of peripheral
functions for HEV/EV motor control. The RH850/C1x MCUs can
implement three types of motor control in hardware: sine wave
PWM, over modulation, and square wave.
Samples of the RH850/C1H and RH850/C1M MCUs are
scheduled to be available from the beginning of 2015 and will cost
$45 and $50 per unit, respectively.
Renesas Electronics Corp.
www.renesas.com
circuitcellar.com
17
CLIENT PROFILE
Newhaven Display
International, Inc.
www.newhavendisplay.com
CONTACT: Melissa Campbell (mcampbell@newhavendisplay.com)
WHY SHOULD CC READERS BE INTERESTED? Intn The
Full Color OLED displays feature 262k colors with a 2001:1
contrast ratio, >160° view, and a 10-μs response time. Each
of these displays is 5-mm thick. They are built as complete
modules that include all necessary logic. They can withstand
temperature ranges from –30°C to +70°C without affecting the
response time of the display. The Full Color OLEDs also support
a 256-step brightness and have a built-in sleep mode as well
as additional features that you won’t find in your typical LCD.
SPECIAL OFFER: For a limited time, take 10% off your first
COLOR OLED online purchase at www.newhavendisplay.com.
Use coupon code CCFullColor during check out. Limit one per
customer; cannot be combined with any other offers.
Circuit Cellar prides itself on presenting readers with information
about innovative companies, organizations, products, and services
relating to embedded technologies. This space is where Circuit
Cellar enables clients to present readers useful information, special
deals, and more.
Get published. Get noticed. Get paid.
Circuit Cellar feature articles are contributed by professional
engineers, academics, and students from around the globe.
Each month, the editorial staff reviews dozens of article
proposals and submissions. Only the best make it into the
pages of this internationally respected magazine.
Do you have what it takes?
Contact C. J. Abate, Editor-in-Chief, today to discuss the
embedded design projects and programming
applications you’ve been working
on and your article could be
featured in an upcoming issue
or online at circuitcellar.com.
Email: editor@circuitcellar.com
FEATURED PRODUCTS: The Full Color OLEDs of fer any
monochromatic applic ation a boos t of vibranc y and
enhanced performance. The NHD-1.27-12896UGC3, NHD1.5-128128UGC3, and the NHD-1.69-160128UGC3 are perfect
for any application, but are ideal for handheld or low-power
applications looking to add a sharp-looking display with the
ability to produce smooth motion graphics and support a wide
viewing angle. To learn more about these Full Color OLEDs,
visit http://bit.ly/ColorOLEDs.
18
FEATURES
BlasterWorks
A Real-Time Wireless
Laser Tag System
Unlike most commercial “laser tag” combat
games that use infrared light-based systems, the
innovative system presented in this article uses
real lasers. Wireless updates work seamlessly
and the sensor system can accurately detect
hits from reasonable ranges.
By Max Rothstein and Richard Speranza
A
s our final project for the course “ECE4760:
Designing with Microcontrollers” at
Cornell University, we built a wireless, realtime laser tag system—or, as we call it,
BlasterWorks (see Photo 1). Having played
standard laser tag games, we were interested
in expanding the game beyond its traditional
boundaries. Our ultimate goals were to create
a system that mirrored combat as closely as
possible and to add new features that are
normally only found in video games. We also
PHOTO 1
The player operates four buttons in the center. From the top left to the right, they correspond to Reload,
Change Weapon, Fire Blaster, and Change Screen.
felt laser tag would be cooler with real lasers,
as opposed to using infrared light, which all
commercially available systems implement.
With this in mind, we started with the
essential components—namely, the laser and
detector system—and expanded the game by
adding an LCD screen and RF transceiver to
provide score updates and configurability.
GAME OVERVIEW
At the highest level, our system has three
main components: a laser blaster, a base
station, and a shot detection vest. Figure 1
depicts their high-level connections.
Each player’s vest is instrumented with
sensors to detect shots coming from other
players. Each blaster has an RF transceiver,
an LCD, a sensor, and a laser. The laser is
used to fire at other players and only turns
on when the player presses the fire button.
The RF transceiver allows each player to send
and receive real-time information about the
game. This information, such as player health,
is displayed on the blaster’s LCD. Finally, each
blaster is equipped with a sensor to determine
if the blaster itself has been hit by another
player. The final component, the base station,
serves as a relay point for the real-time
updates and logs the game’s activity. It also
sets the game parameters, such as starting
health and ammo, and then begins the game.
Each blaster has four buttons that control
its actions. The first button fires the laser on
the blaster. The second button reloads the
blaster, replenishing the ammo count. The
third button cycles through LCD views, and
the fourth button switches weapon types.
circuitcellar.com
Base station
PC
PuTTY
19
DC Supply
SPI
RF Transceiver
COM Port
5V
DC/DC
MAX232
FEATURES
ATmega644
MCU
USART
Serial/RS-232
9-V
Battery
Player 1 Blaster
SPI
SPI
RF Transceiver
DC/DC
9-V
Battery
Player 2 Blaster
RF Transceiver
5V
5V
Wireless comm
Laser
ATmega644
MCU
Lens
Laser
Blaster sensor
Blaster sensor
LCD
LCD
ATmega644
MCU
Player 1 Vest
PWR
Outer sensor
Outer sensor
Inner sensor
DC/DC
Lens
Player 2 Vest
Outer sensor
Outer sensor
PWR
Inner sensor
Data
Data
Outer sensor
Outer sensor
Photo 1 the shows user interface.
The LCD has three different views: selfinfo, scoreboard, and weapon type. The user
has the ability to scroll between these screens
by pressing the change screen button. On the
self-info screen, the player can view his or her
own health and ammo count. By looking at
the scoreboard, the player can see the health
of all of the players in the game. The weapon
screen tells the player if they are in single
shot or automatic mode.
The objective of the game is to be the last
player standing. Players can create their own
custom games with their own parameters, or
choose default values. Default values assign
players with 100 health points, 100 rounds
of ammunition for the automatic blaster, and
unlimited ammunition for the pistol blaster.
However, each pistol clip has only 10 shots, so
a player needs to reload. When a player is hit
on the vest, points are deducted from his or
her total health. If hit in the inner ring of the
vest, 10 points are deducted. If hit in the outer
ring instead, only five points are deducted.
Hitting a player on his or her blaster sensor
prevents them from shooting for the next 5 s
Outer sensor
Outer sensor
and will also deduct five health points.
HARDWARE DESIGN
The hardware can be split into two main
categories: shot detection and microcontroller
processing and peripherals. Both the blasters
and base station have processing/peripheral
blocks, but the base station does not have
a shot detection block. The circuitry for the
blaster and base station are depicted in
Figure 2 and Figure 3, respectively.
The shot detection block’s mission is to
catch any lasers that are shot at the user and
to give information to the microcontroller so
it can determine where the user was hit. As
you can see in Figure 4, there are a total of
six LTR-301 phototransistors. Q1 through Q4
on the schematic correspond to the four outer
detectors on the user’s chest plate. Q5 is the
sensor for the inner detector on the chest
plate. Finally, Q6 is the last sensor, which is
placed on the user’s blaster. These sensors
can be thought of as a normal BJT; however,
instead of biasing the base manually, the
laser does the work for us. Without any light
shining on the base of the phototransistor, the
FIGURE 1
The full system
20
FEATURES
FIGURE 2
The Blaster’s schematic
FIGURE 3
The base station’s schematic
device is turned off and there is an effective
open between the collector and emitter. As
light hits the component, the BJT begins to
turn on, allowing the 5 V from the collector to
flow to the emitter. Something worth noting
is that this device is not an “all-or-nothing”
device. The voltage at the emitter is not
always either 0 or 5 V. Depending on how
hard the component is hit by light, the device
may only be partially on. This behavior is the
reason why comparators are needed.
The purpose of the fast-switching diodes
(D1–D9) in the circuit is to create a logic OR
function. Regardless of which sensor is hit, we
want a comparator to flip to throw an external
interrupt on the microcontroller. By using an
OR circuit, we can obtain this functionality. A
diode OR circuit allows the highest voltage in
the OR to pass through and then back-biases
the remaining diodes.
The LM358 operational amplifiers (opamps) in our circuits are used as comparators
and Schmitt triggers to send effective “bits”
to our microcontroller. The bottom op-amp
is configured as a Schmitt trigger. This is
essentially a comparator with hysteresis. A
Oct 1-3, 2014
Santa Clara Convention Center
Santa Clara, CA
TECHNICAL SESSIONS | HUNDREDS OF EXHIBITORS | KEYNOTES
PRIZES & GIVEAWAYS | NETWORKING EVENTS
Please visit www.armtechcon.com
trigger, similar to the shot comparator. This
Schmitt trigger is attached to the sensor that
is placed on the blaster. A reference voltage
is set with a 10-kΩ potentiometer on the
inverting input. When the user is shot on the
blaster, the noninverting input goes above
this threshold and causes the output of the
op-amp to flip. Whenever the bottom op-amp
flips and generates an external interrupt, the
two bits from the top LM358 are read.
FEATURES
22
PROCESSING & PERIPHERALS
FIGURE 4
The shot detection schematic
FIGURE 5
The laser driver schematic
10-kΩ potentiometer sets a reference level at
which the Schmitt trigger flips. If a shot is
detected from any of the six sensors, then the
voltage at the noninverting input raises above
the set threshold voltage. Due to positive
feedback, the output of the Schmitt trigger
is driven high, which causes an external
interrupt on INT0.
The LM358 package conveniently has
two op-amps inside of it. On the top LM358,
the first op-amp is configured as a simple
comparator. It makes a comparison between
the inner sensor and four outer sensors. By
making this comparison, we can effectively
tell if the user was hit closer to the inner
sensor or closer to the outer sensors. The
1-MΩ resistors on each network pulls voltage
through the first layer of diodes even if the
second layer is biased off. The second op-amp
in the package is configured as a Schmitt
This is the final hardware stage in
our design. This stage includes all of the
remaining circuitry that interface with
the microcontroller to give us our desired
functionality.
Most of the following peripherals are
controlled by the user’s interaction with a set
of four buttons. As we described previously,
the functions are fire, reload, change
screen, and change weapon. The buttons
are connected to ground though a currentlimiting resistor. Since the microcontroller I/O
pins are set to Pull-Up mode, when a button
is pushed, the signal will switch to low. By
reading these pin values through software,
we can determine what the user is trying to
do and respond accordingly.
A special circuit is needed in order to
drive the laser on each blaster. The Atmel
ATmega644 I/O pins (and microcontroller
I/O pins in general) have a limited amount of
current that they can source. In order to drive
the laser hard enough for it to be usable, a
buffer is needed. For our application, a MOSFET
is used to buffer the current needed to drive
our laser. The source of the MOSFET is tied
directly to ground and the drain is connected
to VCC through a 1-kΩ resistor. The gate of
the MOSFET is driven by the microcontroller’s
A.7 pin. Finally, the laser is connected to VCC
and the MOSFET drain. When the output on
A.7 is low, the MOSFET is off, leaving the laser
off. When the MOSFET gate is driven high,
the device turns on, effectively connecting
the negative terminal of the laser to ground.
This functionality provides the laser with the
current it needs, giving us maximum distance.
The sensor system as described would
detect laser beams directly incident on the
phototransistors very well, but would not be
able to detect shots pointed close by. At a
distance of 5 m, the spot size is approximately
6 mm in diameter. By adding a concave lens
in front of the beam, the diameter grows to a
much larger size a few meters away, allowing
us to only use a few phototransistors, rather
than potentially hundreds, to capture a shot.
We used foam core from a local crafts
store to house the lens/laser system, and
it worked quite effectively. They were both
circuitcellar.com
SOFTWARE
The overall software design can be
divided into the same general categories
as the hardware, as it serves to coordinate
concurrent operation of the wireless
communication, the LCD screen, response
to the shot detection, firing the blaster, and
the various buttons. Two separate c files are
needed, base_station.c, which controls the
base station, and blaster.c, which controls
both blasters. Each blaster is also provided
with a unique identifier, my_number, when
programmed.
These activities are managed by the Tiny
Real Time (TRT) kernel. The kernel allows us
to easily manage the main tasks and control
how often to run them, as well as to provide
certain tasks with a higher priority when
needed. Additionally, the LCD update is a
slow process and might otherwise create a
bottleneck on our system. By scheduling other
tasks at the same time as the LCD update, the
controller is able to continuously accomplish
work. The four tasks are: task_health(),
task_wireless(), task_lcd_update(),
ABOUT THE AUTHORS
Max Rothstein (mr579@cornell.edu) is pursuing a BS in Electrical and Computer Engineering at Cornell University. His main
interests are in computer architecture and
embedded software.
Richard Speranza (rjs483@cornell.edu) holds
a BS in Electrical and Computer Engineering
from Worcester Polytechnic Institute and is
pursuing an MS in Engineering at Cornell
University. His main interest is embedded
systems design.
and task_laser(). Additionally, external
interrupts are used to handle shot detection
and firing the blaster.
TASK_HEALTH() & SHOT
DETECTION
Shot detection is performed by an interrupt
service routine (ISR) by using an external
interrupt on PinD.3, which is connected to
the output of the shot detection comparator.
When the interrupt is triggered, the ISR reads
two other bits of information from the shot
detection hardware.
A blaster shot is detected by reading the
FEATURES
placed in the foam by making a small entry
cut, and then firmly pressing the object down
through the foam. Then, we peeled away the
foam covering the glass portion of the lens.
The final peripheral in our system is
Nordic Semiconductor nRF24L01+, a 2.4
GHz RF transceiver. The transceiver and
microcontroller communicate using SPI, an
implementation that already existed in the
drivers, which were borrowed from a previous
project.[1] We have a wireless module on both
blasters and the base station (refer to the
either schematic for pin connections). This
module is what gives us the ability to update
each player with information about the game.
The main challenges involving the wireless
radio were in software, and will be discussed
below.
23
24
PHOTO 2
Base station prototype
FEATURES
TASK_WIRELESS()
value of PinA.5, and a shot to the outer ring of
phototransistors on the detector is detected
by reading the value of PinA.6. Shots to
the blaster and outer ring are flagged as a
weak_hit, while an inner shot is flagged as
a strong_hit.
In task_health(), the values of
those two variables are checked and the
corresponding health update is made. To limit
the number of hits allowed per a given amount
of time, the task sleeps for a longer period
of time after detecting a hit. As in traditional
laser tag, this allows a player who has been
shot to retreat and reset after getting shot.
Also, it prevents small hardware oscillations
from affecting the software operation.
PHOTO 3
Blaster prototype
The wireless communication allows us to
provide the players with real time information
on the state of the game, providing health
information for the other players. Additionally,
it provides a way to configure the game from
the base station terminal on startup, and
send an alert when the game has finished.
The wireless communications are handled
by a state machine, which allows a back and
forth conversation between all blasters and
the base station. While the base station and
blaster have slightly different behaviors,
the state machine is almost identical from
a high level. The main difference is during
initialization state, which is called INIT_
GAME for both. In the blaster’s initialization
state, it waits on a packet containing the game
configurations, namely maximum health and
ammo. This packet has a special identifier to
indicate it is an initialization packet, rather
than an update packet. In the base station’s
initialization state, it first prompts the user
via the PuTTy terminal to respond to a series
of questions regarding game configurations.
Once these questions have been answered,
the base station sends packets to the
awaiting blasters. To ensure delivery without
acknowledgment, the base station transmits
this message for a couple of seconds before
transitioning to the next state. Once a blaster
is initialized, it also transitions to the next
state, INIT_RX.
In INIT_RX, the wireless module is set
to receiver mode, and then the controller
transitions immediately to POLL_UPDATE.
Here, the controller continuously checks to see
if a new packet is available in the receive FIFO.
That packet will contain the remaining hit
points and ammo for all players. Essentially,
this allows any player/base station to update
any player/base station on the current state
The new EAGLE has landed!
Version 7
now
available
For more information, visit www.cadsoftusa.com
26
PHOTO 4
Sensor vest prototype
FEATURES
required to be in range of the base station
at all times. In order to reconcile potential
differences between those update packets, a
value is only considered new if it is less than
the local value. This preserves the proper
state because health and total ammo can only
decrease over time.
If the base station receives an update, it
transitions to INIT_TX, where the mode of
the transceiver is changed to transmit. The
blaster only transitions to INIT_TX when
the player is shot, which is detected by an
external interrupt as discussed earlier. Both
transition immediately to SEND_UPDATE
after the initialization is complete. Here, the
update packet is formatted and broadcast
out. To simplify the communication, the
message is transmitted continuously for a set
duration without requesting a receive packet
acknowledgement. This allows the base
station to reach all blasters without needing
to spend time performing a hand shake with
each player. For the purposes of the game,
it is acceptable if any one controller misses
an update. After the update is sent, both
switch back to receive mode and wait on new
updates.
TASK_LCD_UPDATE()
of the game. For example, Player 1 could
send an update packet to Player 2. Player 2
could then update its local information and
send its own update packet that the base
station might receive. This extends the range
of play for the game because players are not
PROJECT FILES
for the Atmel ATmega8L AVR,” ISRN LUTFD2/
TFRT7608SE, Department of Automatic Control,
Lund Institute of Technology, 2004.
Nordic Semiconductor, “nRF24L01+: Single
Chip 2.4 GHz Transceiver,” Product Specification v1.0, 2008.
circuitcellar.com/ccmaterials
SOURCES
REFERENCE
[1] Lee, Y. H., “Design a Wireless Lab for ECE4760,” Cornell University, http://bit.ly/
USmWTb.
ATmega644A Microcontroller
Atmel Corp. | www.atmel.com
VLM-650-03-LPA Laser and LTR-301
Phototransistors
Digi-Key Corp. (distributor) | www.digikey.com
RESOURCES
Henriksson, D. and Cervin, A.,
“TinyRealTime—An EDF Kernel
Transceiver nRF24L01+ Module with Chip
Antenna
Nordic Semiconductor | www.nordicsemi.com
The LCD screen can either display
personal status info (i.e., its local health and
ammo values), the current blaster type, or
the scoreboard containing the global health
values. The state is controlled by clicking
the Change Screen button. This results in
incrementing the state variable to the next
possible screen. In order to prevent the screen
from flashing, it is only updated when new
information is available. Since updating the
LCD is a relatively slow process, local copies
of the game info are created and displayed,
which means the display may lag behind the
true values. The difference will only be on
the order a few milliseconds, so it will not
affect the game play. This task also handles
reload, which is done by polling PinA.1, which
is driven low by pushing the Reload button.
TASK_LASER() & FIRING SHOTS
This task implements the two types of
blasters available to each player: automatic
and single shot. Each player starts out using
their automatic blaster, and may choose to
switch between weapon types by clicking the
Change Weapon button.
When using the automatic blaster, the
player only needs to hold down the Fire Laser
button and the laser will continuously fire.
On alternate passes through this portion of
the loop, the output pin driving the laser is
turned on and off, its rate controlled by the
TRT kernel function trtSleepUntil(). The
circuitcellar.com
CHALLENGES
The main challenge was enabling wireless
communication. Code for the drivers was
borrowed from a previous project[1]; however,
the sample code does not implement a system
in which one controller takes on the role of both
transmitter and receiver. The function set_
mode() provided by the driver file wireless.c
is supposed to change between modes, but
this function alone does not work. In order
to switch between modes, we discovered it
was necessary to clear out the receiver buffer
by calling flush_RX_FIFO(), and then
setting the new mode. The transceiver does
not maintain the transmitter address while in
receive mode, so we needed to call set_TX_
addr() with the new address when switching
from Receiver to Transmitter mode.
RESULTS
We successfully created working prototypes
for one base station and two blasters, each
equipped with a sensor vest. Pictures of the
complete product are shown in Photos 2-4.
Wireless updates work seamlessly. The sensor
system is fairly accurate at detecting hits
from a reasonable range, and the remaining
features work as intended.
The main limitation on the current
prototype system comes down to the range
of play. The sensor vest can detect shots from
a maximum range of approximately 5 m. The
short range is due to both decreased intensity
of the laser beam, as well as low sensitivity
of the sensor system. More precise tuning of
the comparator’s hysteresis as well as noise
reduction from ambient light would help to
improve the range of the system. Additionally,
a production model with an additional focusing
convex lens would have greater range, since
the beam’s power density would not drop off
as quickly.
At a range of 3 m, our system is
approximately 92% reliable in logging hits.
This test was carried out by first fixing the
direction of the blaster to point directly at the
center of one phototransistor on the vest, and
then firing repeatedly while holding the vest
steady. Most importantly, no false positives
were detected while playing in a well lit room.
The range of the wireless transceivers in
open lab space is approximately 32 m. Due to
the method of sending updates, in which the
transmitter sends out packets continuously
for a couple of seconds, packets are reliably
received while in range. As long as the players
stay within 32 m of each other the game can
continue, regardless of how far they are away
from the base station, since the blasters can
maintain communications with each other.
The battery life of the blasters is sufficient
for a number of games. At steady state,
a blaster draws about 34 mA and jumps to
about 64 mA when firing the blaster. In the
worst case scenario, the blaster can fire for
50% of the time, putting the average current
draw at 49 mA. This translates to a lifetime
of 12.8 h based off of a 625 mAh lifetime of
the 9-V battery. However, a more reasonable
estimation is that the blaster is firing about
5% of the time, resulting in an average
current draw of 35.5 mA and a 17.6-h lifetime.
FUTURE UPDATES
We achieved the main goals as originally
outlined in our proposal. We now have a
working prototype to play a game of laser
tag. We borrowed lessons learned from all of
the previous ECE 4760 labs at Cornell, learned
some new ones, and successfully integrated
them into a single product.
Given the chance to do this project again,
we would have taken into consideration the
complexity of integrating many working
parts and given ourselves more time for
this portion, rather than assuming that they
would simply “work” when they were all put
together. Originally, we wanted to also have
sound effects to make exciting “laser blasting”
sounds, but time limitations prevented us
from finishing the task.
In the future, not only would we like to add
sound back in, but we would also like to build
more blasters to allow many more players.
Also, the lasers could be fired with a payload
that would allow the detector to determine
the originator of the shot. Perhaps most
importantly, expanded range would make the
game much more interesting to play. We see
plenty of opportunity to take our core system
and create an exciting and unique laser tag
game.
FEATURES
player’s ammo value is decremented on each
fire.
When using single shot, the trigger
detection makes user of an external interrupt.
The Fire Laser button is also wired to INT2
on PinB.2, which is triggered by a transition
from high to low. A flag, called trigger_
interrupt, is set in the ISR for the external
interrupt. The value of the flag is read within
the task and a shot is fired as a result. Button
clicks reliably result in only one laser fire
since the allowed time between shots, which
is controlled by the sleep time, is long enough
for any transient signals to die out. If the
blaster is disabled as a result of getting shot
directly in the blaster, then the task sleeps for
five seconds, rendering the blaster unusable.
27
•Spotlight
88 | October 2014 | www.elektor-magazine.com
advertorial
@electronica 2014
Made in Munich
Come MAKE it @electronica 2014
By Wisse Hettinga (Elektor Labs)
You’ve all heard of the maker revolution and seen cool things described
but how cool is it to MAKE some real electronics at the world’s largest
exhibition on real electronics? To have a place where you can relax,
charge your phone, e-gossip, have a coffee and touch base with real-world electronics. Where you can bring along anything you’d want
to unbox, (un)solder, show off, measure, check-4-specs, Arduino’d, or
Raspberry Pi’d. Feel free to come work with us the Elektor way!
Elektor designers and engineers are on standby to lend a helping hand and supply
everything you need to get some real electronics work done right there and then.
German, English, Dutch, Spanish, C++, spoken.
For this special occasion Elektor Labs are putting at your disposal: desk space, tools,
test & measurement equipment, a 3D printer, and free WiFi. Not forgetting mini workshops, techtalk, coffee (please donate), and plenty of power sockets to charge phones,
tablets, laptops, and gizmos.
Navigate to the Elektor Maker Space in Hall A6, Booth 380
Stay in touch with Elektor’s activities
register for our newsletter on www.elektor.com, right hand bottom corner.
IA
ONAL MED
I
T
A
N
WHO:
R
E
T
LEKTOR IN
YOU AND E
E
WHAT:
AKER SPAC
M
’S
R
O
T
K
ELE
WHEN:
4
11–14, 201
NOVEMBER
,
WHERE:
A, MUNICH
C
I
N
O
R
T
C
ELE
BOOTH 380
HALL A6 /
advertorial
www.elektor-magazine.com | October 2014 | 89
30
FEATURES
Land Mobile Radio
Wireless Control
For long-distance monitoring and control,
Land Mobile Radio (LMR) is another wireless
option to consider. You can build a wireless
interface that has eight output relays and
eight digital inputs that can be controlled
by the serial port on a mobile radio.
By Ken Merk
T
here was a time when controlling
something wirelessly was magical.
Let’s consider some examples.
I remember the first automatic garage
door opener on our street. People walking
down the sidewalk would stop and watch this
miraculous invention perform its duty.
PHOTO 1
Kenwood’s NX-210 NXDN 5-W portable radio is capable of analog and digital transmissions. The desired
function is highlighted using the Up and Down buttons on the front keypad and then sent by pressing the
side Push-to-Talk button.
Then there was the
remote control. I remember
the Christmas when my father
brought home a new Zenith black-and-white
television. My main interest was its remote
feature. Inside the handheld remote (which we
called the “clicker”) were four aluminum rods
cut to specific lengths. When you pressed one
of the four buttons, a spring-loaded hammer
would strike the rod emitting a sound. Each
rod emitted a different ultrasonic frequency
that would activate a specific function on
the TV. Of course, sometimes rattling a
ring of keys in front of the TV would cause
it to change channels or mute the volume.
Primitive as it may sound, it worked—and it
didn’t need batteries.
Next, I saw cordless phones come
on the scene. The early ones had long
telescopic antennas because of the low
operating frequency. Later, as technology
progressed, the frequencies increased,
with the antennas and the units themselves
becoming smaller.
Today, we have cell phones with Bluetooth
earpieces and wireless routers for our
Internet connections. Our computers have
wireless peripherals and we can place wireless
printers anywhere in the house. We can lock
car doors with a keychain transmitter. And
when we drive over a toll bridge, a tracking
system can detect us wirelessly. Most houses
circuitcellar.com
LAND MOBILE RADIO
The Land Mobile Radio (LMR) system I’ll
cover in this article probably isn’t familiar to
most of you. It’s been around for a long time.
Most public safety services such as police,
fire, and ambulance use it. You have probably
seen a police officer with a portable radio
strapped to his belt and a familiar speaker/
mic on his lapel. Forestry, railways, trucking
companies, and power companies also use
FEATURES
now have smart power meters that send
power consumption data by radio to your local
power company. The list can go on into the
industrial sector with remote control cranes,
trains, and other machinery.
With all these wireless products there
must be standard data protocols in place to
ensure compatibility. The most common are
all the cellular protocols, wireless Ethernet
802.11, Bluetooth, and ZigBee. Radios can
be narrowband or spread spectrum, hopping
or direct sequence, and usually low power
without the need for a license.
All the wireless control technologies I’ve
described thus far have one thing in common.
The range of these systems is relatively low
as is the RF power output of the radios, and
a radio license is not needed. The ISM radio
band from 902 MHz to 928 MHz usually allows
1 W maximum RF output. Most 2.4-GHz
802.11 radios are limited to 1 W (30 dBm)
of RF output power. Bluetooth and ZigBee
devices max out around 100 mW. Remember
that the type of antenna also plays a part
in the effective radiated power (ERP) output
of the systems. These higher-frequency
systems need a visual line of site between
the transmitting and receiving antennas for
maximum distance. This visual line of sight
also includes radio line of site where Fresnel
zones come into play.
31
PHOTO 2
Kenwood’s NX-700 NXDN 30-W mobile radio is used to receive status messages and control the relay I/O
board. The text associated with the status message is displayed upon arrival.
the service. The portable handheld radios are
usually 5 W with the mobile radios around 30
W and the base stations could be up to 100
W. With this kind of power, radio licenses are
needed and frequencies are assigned to each
user by a regulatory group, such as the US
Federal Communications Commission (FCC) or
Industry Canada.
Most LMR systems are in the VHF band
136 to 174 MHz or UHF band 400 to 470 MHz.
Radio to radio distances measured in miles
could be increased when working through a
repeater or trunking system. With a repeater
placed high on a building or mountain top,
you could cover an entire city.
The actual radio hardware used in this
project are Kenwood’s NX-210 portable (see
Photo 1) and NX-700 mobile (see Photo
2), which are capable of analog and digital
PHOTO 3
The back panel of the NX-700 mobile
radio has access to all the electrical
connections. The DB-25 connector
contains the RS-232 port which will
control the I/O board. The black and
red wires power the radio and the
female UHF connector on the far left
is the antenna input.
FEATURES
32
PHOTO 4
The relay I/O board has eight output relays and eight digital inputs. Terminal strips on the board make it
easy to wire in external devices. A test LED is connected to output relay #8 which lights when the relay is
energized.
modulation. For the most part all vendors
manufacturing 12.5-kHz narrowband FM
analog radios are compatible with each other,
but the digital spec is another story. A joint
effort by Icom and Kenwood developed the
digital technology Next Generation Digital
Narrowband (NXDN) for its line of LMR radios.
Another popular digital format is P25 which
PHOTO 5
This is the setup I used to build the
interface. Interconnects from the radio
and single board computer terminate
on a common perfboard which
gives access to all pin connections.
The ribbon cable terminated with a
DE-9 female connector plugs into a
computer's serial port for diagnostic
purposes. A 50-Ω dummy load
simulates an antenna connection.
is not compatible with NXDN. To maintain
product flexibility, Icom and Kenwood also
make a P25 line of radios.
There are a few modes in which the radios
can be programmed. Channels can be set
up as narrowband analog 12.5-kHz spacing,
narrowband digital, and very narrowband
(VNB) digital 6.25-kHz spacing. Due to the
very narrow bandwidth of 6.25 kHz, analog
transmissions will not work, so digital is the
only mode available.
The technology of two-way radios has
come a long way since their early beginnings.
Embedded microcontrollers now control
the radios so there is no need for crystal
elements, as all frequencies are synthesized
and all functions are programmed into the
radio using the vendor’s software. Kenwood
has given users access to the microcontroller
via a serial port on the back of the radio. An
ASCII-based protocol over this serial port is
used to control the radio which can mimic
someone controlling the radio from the front
panel buttons. Access to data received by
the radio can also be acquired through this
port. The DB-25 connector Photo 3 on the
back of the NX-700 mobile radio is where
this connection is made. A word of caution,
the DB-25 is not wired as a standard RS232 port so do not connect it directly to a
computer via a serial cable. This connector
not only contains a serial port for control and
programming, but also another port for GPS
connection, audio in, audio out, DC power
out, multiple programmable I/O ports, and
other functions.
FEATURES
34
TABLE 1
This is the status list that was
programmed into both radios. The
list shows the status number and the
corresponding text. You can customize
the text for you own application.
Instead of Relay8 ON you could
use Motor ON, Pump ON and name
whatever device you are controlling.
Number
Status
Status Name
1
31
Relay1 ON
2
32
Relay2 ON
3
33
Relay3 ON
4
34
Relay4 ON
5
35
Relay5 ON
6
36
Relay6 ON
7
37
Relay7 ON
8
38
Relay8 ON
9
41
Relay1 OFF
10
42
Relay2 OFF
11
43
Relay3 OFF
12
44
Relay4 OFF
13
45
Relay5 OFF
14
46
Relay6 OFF
15
47
Relay7 OFF
16
48
Relay8 OFF
17
50
ALL OFF
18
51
ALL ON
19
61
Input Status
20
62
Output Status
21
63
Ping
WIRELESS I/O INTERFACE
In this article, I’ll describe how to build
an interface that has eight output relays and
eight digital inputs which can be controlled
by the serial port on the NX-700 mobile radio.
Anyone with a properly programmed NX-210
portable radio can now control the eight output
relays and monitor the eight digital inputs
wirelessly. This will work with narrowband
analog, narrowband digital and VNB digital
as well as trunking. In Analog mode, the
signaling will be MSK, which Kenwood calls
Fleetsync, and the Digital mode will be NXDN,
which is four-Level FSK. Fleetsync over the air
speed can be 1,200 or 2,400 bps, but 1,200 is
usually used for better reliability. Narrowband
digital NXDN can achieve 9,600 bps and VNB
6.25-kHz bandwidth 4,800 bps.
For this project, I used JK Microsystems’s
Flashlite embedded DOS 186 single-board
computer (SBC) together with its eightchannel relay I/O board (see Photo 4). The I/O
board sits on top of the SBC using standoffs
and connected together with a ribbon cable
(see Photo 5). It has eight normally open
SPST relays with a continuous current rating
of 5 A and eight optocouplers with an input
voltage ranging from 3 to 24 V DC or AC.
The Flashlite 186 SBC has a 33-MHz
80186-compatible processor that enables you
to write code on your PC in the language of
your choice and then upload it to the on-board
flash drive. A debug serial port is available
that was used to interact with the program
during development. The on-board switching
power supply accepts an input voltage from 7
to 34 VDC and supplies a 5 VDC output at 300
mA for the user to power external devices—
which in this case is the relay I/O board.
The software for this project was written
in Forth. Using Forth enabled me to get the
hardware up and running very quickly. A
language called NEXX was written first (in
Forth) and then used as my development and
debugging environment until the program
was working to my satisfaction. NEXX was
used to interact with Kenwood’s PC protocol
embedded in their radio. The program was
then configured to run automatically on
power-up.
SENDING STATUS MESSAGES
FIGURE 1
The schematic shows the radio, Flashlite SBC, and relay I/O board, which are the three main components.
There is a minimum of wiring involved to get the interface up and running. The LED which simulates your
device to be controlled is powered when relay 8 energizes. To verify power is applied to the LED, input 8
monitors this voltage for positive feedback.
Kenwood radios can be configured to send
out “status” messages. A status message is
basically a number which is transmitted out
in the data stream. This number has a text
string associated with it in the transmitting
and receiving radio. Only the number is sent,
but the associated text will show up on the LCD
screen of both radios. For analog channels,
status numbers are two digits and range from
10 to 99. Status numbers 10 to 79 are general-
circuitcellar.com
: Match?
:
nxdn.buf 14 + 3 search.for.string ;
purpose and free to use. Status numbers 80 to
99 are reserved for special functions, such as
remote kill/stun, horn alert, and emergency
man down. Kill/stun will disable the radio or
inhibit from transmitting remotely. Digital
channel status numbers range from 1 to
207 with 224 to 226 reserved. Some special
functions are user programmable and can be
mapped to any status number.
A list of status numbers and their
associated text (16 chars max) is programmed
into both the NX-210 hand-held portable radio
and the NX-700 mobile radio (see Table 1). The
list in the NX-210 portable radio will create
an up/down scrolling menu where the desired
text can be highlighted with the front panel
keypad and then transmitted by pressing
the Push-to-Talk (PTT) button. It’s hard for
humans to remember a series of numbers so
we usually associate the status number with
text, which in this case will be the function we
want to activate. The status number is sent
to the receiving radio in a special data packet
with attributes to distinguish it as a status
message. The receiving radio takes the status
number and displays the corresponding text
from its programmed list. If you do not enter
any text in the status list, the functions will
still work but only the status numbers will be
displayed.
BREAK
BREAK
BREAK
BREAK
BREAK
BREAK
BREAK
BREAK
BREAK
BREAK
BREAK
BREAK
BREAK
BREAK
BREAK
BREAK
FEATURES
DO.NXDN.STATUS
.” NXDN Status “
buf.pointer 15 - 3 =
IF
nxdn.buf 14 + 3 type
“ 031” Match? IF Relay1 >on .” Relay1 ON” drop
“ 041” Match? IF Relay1 >off .” Relay1 OFF” drop
“ 050” Match? IF ALL.OFF .” All OFF” drop BREAK
“ 051” Match? IF ALL.ON
.” All ON” drop BREAK
“ 061” Match? IF Opto.status drop BREAK
“ 062” Match? IF Relay.status drop BREAK
“ 063” Match? IF Ping
drop BREAK
THEN ;
LISTING 1
This is a portion of Forth code running
on the Flashlite 186 SBC. The heart
of the code is a case statement that
compares inbound status message
numbers to an onboard function list.
When a match occurs the function is
executed.
35
Need a Free RTOS for
Your DIY Project?
Use the best. SMX® is now freely available for non-commercial projects.
• Learn multitasking and develop
your own applications
• Free for non-commercial usage
• No time limit, No sales contact
Y O U R
mdi_learningkits_CCI_081014.indd 1
R T O S
P A R T N E R
• Extensive operating examples and
excellent manuals
• Work with low-cost boards and free
IAR tools
www.smxrtos.com/lk
8/15/14 8:54 AM
FEATURES
36
PHOTO 6
This is a constellation diagram showing a strong NXDN digital data stream. A correct symbol received with
no error is indicated by a dot within the circle of its quadrant.
The receiving radio also gives you access
to the status message data packet through
the RS-232 port on the back of the radio. This
is monitored by the SBC and the program
will pick out the status number and activate
the desired function on the relay I/O board.
For diagnostic purposes, the received data
packet is also sent out over a second RS-232
port to a laptop running a terminal program,
such as Procomm or Hyperterm. The packet
is displayed in Hex with corresponding ASCII
values in a sort of protocol analyzer style.
STATUS MESSAGE FUNCTIONS
Table 1 shows the status list that is
programmed into both the transmitting and
receiving radios. The list’s function is to
display the text associated with the status
number that is sent or received. The computer
board’s job is to pick out the status number
from the received data packet and perform
the desired function.
You can see from the list that status
numbers 31 to 38 will turn on relays 1 to 8 and
status numbers 41 to 48 will turn off relays 1
to 8. Status 50 will turn on all eight relays and
51 will turn off all eight relays. Status 61 will
send back the status of the eight optocoupler
inputs as an 8-bit binary number. A display
showing IN->01000001 indicates inputs 1 and
7 are high. Status 62 works in a similar way by
sending back the current status of the relays.
A display showing OUT->00001100 indicates
relays 3 and 4 are energized. The indications
sent back are actually text messages that are
only sent to the radio that requested it. Each
radio has a unique ID number that’s pulled
out of the received data packet. This ID is
used to target only the requesting radio.
Status 63 is the Ping function, which is
used to see if the receiving radio is within
radio range. A successful Ping will result in
“Ping->OK” being displayed.
The NXDN protocol has built-in feedback to
indicate a successful data link. When a status
message is sent from the portable radio, the
display shows << SEND DATA >> as the data
is sent. The receiving radio sends back an ACK
to the requesting radio and << COMPLETE >>
appears on the display. If the sending radio
does not get an ACK, it tries three times and
then displays << NO REPLY >>. This verifies
the radio link, but is the relay contact feeding
power to your device? The relay coil could be
energized, but if the contacts are faulty the
device will not turn on. To verify power to
your remote device, the eight digital inputs
on the relay board could be connected to the
devices and then an input status could be
requested to verify power. This would give
you positive feedback on the field conditions.
The schematic in Figure 1 shows an example
on how this could be done.
HARDWARE & FORTH CODE
Photo 5 shows the setup I used to build
and troubleshoot the radio interface. The
Kenwood NX-700 radio’s DB-25 connector
(Photo 3) is terminated on a perfboard
mounted on standoffs. Header J10 on the
Flashlite SBC, which contains serial port 0, is
also terminated on the perfboard. This makes
it easy to connect both serial ports together
as indicated in the schematic (see Figure 1).
Both ports use true RS-232 levels so they
are simply connected together. Header J8
on the SBC is the diagnostic serial port and
is terminated with a DE-9 female connector
as seen in Photo 5. The small pushbutton
ABOUT THE AUTHOR
Ken Merk (krem@telus.net) is an electronics technologist with a degree from The British Columbia Institute of Technology (BCIT). He currently works for Canadian Pacific Railway, where he is
involved with radio-controlled locomotives and wireless data systems.
DESIGN YOUR SOLUTION TODAY
CALL 480-837-5200
Touch Panel Computers
TS-7250-V2 Embedded Board
Panel Mount or Fully Enclosed
High Performance & Industrial Grade
Series start at
Pricing starts at
Qty 100
Qty 100
$369
$165
$409
$199
Qty 1
Qty 1
Shown w/
optional
micro
SD Card
Features can include:
5, 7, and 10 Inch Touchscreens
Fanless Operation from -20 ºC to 70 ºC
Up to 1 GHz ARM CPU, 2 GB RAM, 4 GB eMMC Flash
Features:
Ethernet, USB, DIO, CAN, RS-232, Modbus, SPI
Optional Cellular, WiFi and XBee Radios
Supports Android & Linux with Fast Boot Times
Modbus Peripherals
Standalone Signal Processing
TS-1400
Up to 1 GHz ARM CPU
2x USB Host
512 MB RAM
1x USB Device
2 GB eMMC Flash
6x Serial Ports
2x SD Card Socket
75x DIO, 1x CAN
2x 10/100 Ethernet
1x PC/104 Connector
Benefits:
TS-1420
Hardware Flexibility with On-board FPGA
Several control I/O interfaces
Launches your application in under a second
Easy development w/ Debian and Linux 2.6
GI
NA
L
RI
Support every
step of the way
with open
source vision
O
N
-40 ºC to 85 ºC Industrial temperature range
PE
Embedded
systems that
are built to
endure
High Data Reliability with SLC eMMC Flash
O
GE
D
We’ve never
discontinued a
product in 30
years
12x 8 A Power Switches
Output Power Monitor
6-Bit PWM Capable
RU
G
LO
NG
LI
FE
6x 24 VDC Auto Relays
2x MODBUS-TW RJ45
Rugged Connectors
Unique embedded
solutions add
value for our
customers
FEATURES
38
PHOTO 7
This constellation diagram shows a very weak NXDN signal, which is dominated by noise. All dots not within
the circles indicate errors and corrupt data.
mounted on the perfboard is connected to
one of the I/O pins on the radio. The pin is
programmed to be an external push to talk
button to key up the radio for test purposes.
An external power supply of 12 VDC is used to
power the SBC and the radio.
Listing 1 shows the heart of the code
running on the Flashlite SBC. Here we see a
long case statement that matches received
status numbers to the desired function. The
Forth word Match? searches the serial input
buffer nxdn.buf for a status number and
compares it to the list in the case statement
using search.for.string. If there is a
match the function is executed, the name of
the function and status number is sent out the
diagnostic serial port and the case statement
exited using BREAK. This is repeated anytime
new data enters the serial input buffer.
Sources
Flashlite 186 SBC
JK Microsystems | www.jkmicro.com
NX-210 Portable radio and NX-700 mobile radio
Kenwood USA Copr. | www.kenwoodusa.com
F-PC Forth 3.6
Tom Zimmer, Forth Interest Group | www.
forth.org
Normally, the radios described in this
article are used for voice transmissions, but
there are many other useful features that
should be considered. The NX-700 mobile
radio can be configured to run in Transparent
mode where it basically becomes a data radio.
If you need to send RS-232 data wirelessly
from one point to another, this could work for
you. Input your RS-232 data into the serial
port of the radio (DB-25) where it is stored in
a text buffer. While the buffer contains data,
the radio will automatically key up and send
the data to the receiving radio where it will be
available on its serial port. When the buffer
empties, the radio will stop transmitting and
wait for more data.
The DB-25 connector on the radio also
contains programmable I/O pins. I explained
how a pushbutton switch was connected
to one of these I/O pins and programmed
to be an external push to talk switch. You
can configure this button to do many other
functions. If you only need to control one
relay remotely, you can do so with just the
radios. There’s no need for any computer
interface. The I/O pin can be programmed to
send a status message when the pushbutton
is pressed and send another when the button
is released. On the receiving radio, an I/O
pin can be configured to drive a relay. When
the radio receives the button-pressed status
message, it will energize the relay. When it
receives the button-released status message,
it will turn off the relay.
I already explained how status messages
can be used to activate functions remotely.
There are only so many status numbers
available to use. To have an unlimited number
of controllable functions, you can use text
messages in place of status messages. The
Kenwood NX-210 portable radio is capable of
sending text messages entered on the front
keypad. A specific text string can be sent and
compared at the receiving end to activate a
function. If you send “Function1” as a text
message, it will turn on the test LED on the
Flashlite SBC. Function2 will turn off the LED.
I have included these in my code just for
testing and demonstrating purposes only.
To really grasp the capabilities of this
interface it is necessary to become familiar
with the operations and programming of
Kenwood’s NXDN line of radios. This is all
available in their programming and user
manuals available from any authorized
Kenwood dealer.
FOUR-LEVEL FSK
NXDN is a digital over the air protocol
that takes voice and implements a vocorder
(AMBE+2) to digitize it into a data stream.
Forward error correction is introduced and
circuitcellar.com
Instead of using audio tones, NXDN
shifts the carrier frequency of the radio
transmitter to create symbols or transitions
and each symbol carries 2 bits of information.
A deviation of +1,050 Hz represents binary
01, +350 Hz binary 00, –350 Hz binary 10,
and –1,050 Hz binary 11. Photo 6 shows a
NXDN data stream being received by a piece
of test equipment and showing the results
in a graphical form called a constellation
pattern. You can see four quadrants which
represents four of the symbols I just
described. Each quadrant is labelled with the
frequency deviation of each symbol. When a
carrier deviation is detected, it is plotted in
one of the quadrants as a dot. If the dot
is within the circle seen in each quadrant,
it is within spec and considered good data.
Photo 6 shows a strong signal and no errors
in the data stream. Photo 7 shows a weak
signal where noise is introduced to create
errors which show up as dots not within the
circles.
This is a little insight into the technology
used in the NXDN protocol. You do not need to
know all the details to set up the radio
interface I described in this article. But it is
something to think about.
Wide Temp. Panel PC
PPC-090T
l Vortex86DX2 933MHz Fanless
l Up to 1GB of Onboard Memory
l Low Power Consumption
l 2 RS232/422/485 serial ports
l 3 USB 2.0 Host Ports
l 10/100 BaseT Ethernet
l PS/2 KB port, Audio Out
l Compact Flash Slot / SD Slot / SATA
l 9 inch 1024 x 600 WSVGA TFT LCD
l Resistive Touch Screen
l DC-IN +12VDC / +24VDC
l IP65 Front Panel
l Wi-Fi (Optional)
l Giga Ethernet (Optional)
l Wide Temperature (Optional)
3.xx KERNEL
The PPC-090T comes ready to run with the Operating System installed on flash
disk. Apply power and watch the Linux X-Windows desktop user interface appear
on the vivid color LCD. Interact with the PPC-090T using the responsive integrated
touchscreen. Everything works out of the box, allowing you to concentrate on your
application rather than building and configuring device drivers. Just Write-It and
Run-It... Pricing starts at $495 for Qty 1.
http://www.emacinc.com/sales/cc9
Since 1985
OVER
29
YEARS OF
SINGLE BOARD
SOLUTIONS
EQUIPMENT MONITOR AND CONTROL
Phone: (618) 529-4525 · Fax: (618) 457-0110 · www.emacinc.com
FEATURES
this data stream is used to modulate the
transmitter using four-level frequency shift
keying (FSK). The receiver takes this data
stream and converts it back to audio as voice.
Early phone modems used FSK to send data
over phone lines to access other computers.
A certain audio frequency tone was sent to
represent a binary 1 (mark) and another
frequency to represent a binary 0 (space).
The tones of the FSK are sometimes called
symbols. The transition from one tone to the
other represents one bit of data, which is 1
baud. The rate at which the transitions occur
is called the baud rate. Back then the baud
rates were slow, so a 2,400-baud modem
produced 2400 bps. To increase the data rate,
you would increase the rate of transitions, but
this only worked up to a certain point. You
can only transition so fast until you exceed
the bandwidth of the phone line or radio. If
you could send 2 bits per transition, your bits
per second would increase and the baud rate
would stay the same. Four-level FSK achieves
this by sending 2 bits per symbol (transition).
When you do this, the terms “baud rate” and
“bps” are no longer interchangeable. We
could have a 1,200-baud modem sending out
9,600-bps data streams.
39
40
FEATURES
AC Monitor
Want a simple, straightforward way to determine voltage
values? With the circuit described in this article, you can
quickly tell if voltage is low, nominal, or high. The circuit uses
an LM3914 and a 10-segment LED bar to display a range of
voltage values.
By Larry Cicchinelli (US)
H
ave you ever wanted a simple visual
indication of a voltage value? For instance,
wouldn’t it be useful to be able to tell, with
only a quick glance, if a voltage is low,
nominal, or high? The circuit I’ll describe in
this article may be just what you are looking
for (see Figure 1)! By selecting the appropriate
resistor values, you can monitor just about
any voltage and voltage range and show its
value with a nine- or 10-segment bar graph.
The circuit shows you how to use a well-known
IC, the LM3914, with a 10 segment LED bar to
display a range of voltage values.
DISPLAY DRIVERS
The National Semiconductor (now Texas
Instruments) LM3914 is one of a family
of three display drivers in which the only
difference among them is how the internal
voltage divider is configured. Each IC has a
10-tap voltage divider with a comparator on
each tap. The LM3914 is linear, the LM3915 is
logarithmic with 3-dB steps, and the LM3916
is also logarithmic but scaled like a VU meter.
The units can be operated in either of two
modes: Dot mode, where only a single LED
is on, or Bar mode, where all lower value
LEDs are also on. The indicated website has
complete specs as well as some application
notes on the device.
The drivers can be used individually or
cascaded in order to create a display with a
larger range. In an article titled “Calibrated
Decibel Meter Design” (Circuit Cellar 236,
2010), I explained how I used two display
drivers (LM3915 and LM3916) in cascade for
a multirange audio level meter. The circuit I
describe in this article uses a single driver to
show whether or not the monitored voltage is
near its nominal value.
CIRCUIT OVERVIEW
This specific implementation is used to
monitor the line voltage coming into my home.
I use a 9-V AC wall wart to step the voltage
down to a more manageable level, followed
by a 3:1 voltage divider. I used a Variac to
characterize the wall wart and the rectifier
circuit. The results are shown in Table 1. The
wall wart I used is rated at 9 VAC and 300 mA.
Its open circuit voltage is about 16 V (with
120-V input) and it drops off significantly with
even a light load. I ensured a constant load by
selecting the LM3914’s Dot mode. One of the
circuit’s limitations is that voltages beyond
the range limits will not be displayed.
The LM3914 has a built-in 1.2-V reference
that can be used to develop the comparator
references. It is configured using the same
method as the popular LM117 series of
AN ELEKTOR INTERNATIONAL MEDIA PROMOTION
Get Started with Advanced
Control Robotics
By C. J. Abate (Content Director, EIM)
ADVANCED CONTROL ROBOTICS HANNO SANDER
I met Hanno Sander in 2008 at
the Embedded Systems Conference in
San Jose, CA. At the time, Sander was
at Parallax’s booth demonstrating a
Propeller-based, two-wheeled balancing robot. When I saw his interesting
balance bot design and his engaging
way of explaining design to interested
engineers, I knew that Sander would
be an excellent resource for future
Circuit Cellar content. I was right.
Several months after
the conference, we
an article
ADVANCED CONTROL published
he wrote about the
balancing robot project (Circuit Cellar
March 2009). Today,
HANNO SANDER
Sander runs OneRobot with the aim of
“building high-quality, affordable products by pushing offthe-shelf components
to their limits.”1
ROBOTICS
THE FUTURE IS NOW
When it comes to
robotics, the future
is now. With the ever-increasing
demand for robotics
a p p lic a tio n s—f r o m
home control sysTitle: Advanced Control Robotics
tems to animatronAuthor: Hanno Sander
ic toys to unmanned
Publisher: Elektor/Circuit Cellar
planet rovers—it’s an
Year: 2014
exciting time to be a
Buy: www.elektor.com/advancedroboticist,
whether
control-robotics
you’re a weekend DIYer, a computer science student, or a
professional engineer.
It
doesn’t
matter
whether
you’re building a line-following robot toy or tasked with designing
a mobile system for an extraterrestrial exploratory mission: the
more you know about advanced
robotics technologies, the more
you’ll succeed at your workbench.
Advanced Control Robotics is intended to help roboticists of various
skill levels take their designs to the
PARALLAX PROPELLER
ADVERTISEMENT
next level with microcontrollers and
the know-how to implement them
effectively.
THEORY & BEST PRACTICES
Advanced Control Robotics simplifies the theory and best practices of advanced robot technologies.
You’re taught basic embedded design theory and presented handy
code samples, essential schematics,
and valuable design tips (from construction to debugging).
LEARN THEN DESIGN
The principles described and topics presented in Advanced Control
Robotics are immediately applicable. With the book at your side,
you’ll be innovating in no time.
Sander covers:
• Control Robotics: robot actions,
servos, and stepper motors
• Embedded Technology: microcontrollers and peripherals
• Programming Languages: machine level (Assembly), low level (C/BASIC/Spin), and human
(12Blocks)
• Control Structures: functions,
state machines, multiprocessors,
and events
• Visual Debugging: LED/speaker/
gauges, PC-based development
environments, and test instruments
• Output: sounds and synthesized
speech
• Sensors: compass, encoder, tilt,
proximity, artificial markers, and
audio
• Control Loop Algorithms: digital
control, PID, and fuzzy logic
• Communication Technologies: infrared, sound, and XML-RPC over
HTTP
• Projects: line following with vision and pattern tracking
Are you ready to start learning
and innovating? Order Advanced
Control Robotics today.
1OneRobot.org/about
FEATURES
42
FIGURE 1
R3, R5, and R7 are optional. You can use them to modify the adjustment range of the potentiometers. I used
the values in this schematic for my AC monitor.
VIN (VRMS)
VOUT(meas)
VOUT(calc)
% Error
100
10.33
10.25
-0.75
105
10.90
10.79
-0.97
110
11.20
11.34
1.21
115
11.80
11.88
0.65
120
12.35
12.42
0.55
125
12.94
12.96
0.15
130
13.55
13.50
-0.36
135
14.11
14.04
-0.48
TABLE 1
This table shows the characteristics of the wall wart I used to drive my monitor. The measurements were
taken while it was driving the circuit. The VOUT(meas) values are the same voltages that are applied to the 9
VAC voltage on H1 in Figure 1. The 9 V is not fixed; it changes with the wall voltage.
voltage regulators. However, another of the
LM3914’s features is that the LED current is
approximately 10 times that of the REF OUT
pin current. Also, note that the resistance of
the voltage divider chain in the IC can vary
from 8 to 17 kΩ. These two characteristics
could have made it more difficult to set up the
voltages I wanted to use for my upper and
lower references. Because of this, I decided to
use an external voltage reference of 5 V. The
voltage divider with its two pots (R4 and R6)
allow for setting the upper and lower limits
SOURCES
DipTrace Schematic and PCB software
www.DipTrace.com
National Instruments (now Texas Instruments)
LM3914
Texas Instruments | www.ti.com/product/
lm3914#relprod
quite easily. The low values of the divider
resistors were chosen in order to ensure that
the voltage set points are relatively unaffected
by the divider chain in the LM3914. I chose to
use a 5-V regulator so that the unit can be
powered with a voltage as low as 6.5 V. The
LM3914 will operate with a supply as low as 3
V, but the upper voltage reference must be at
least 1.5 V lower than the supply.
Note that Figure 1 indicates that R3, R5,
and R7 are optional. Depending on your specific
application, you may want to use them to modify
the adjustment range of the potentiometers.
The values indicated in the schematic are those
that I used for my AC monitor.
There is an optional resistor (R9) that can
be installed on the lowest LED of the LED bar,
which enables it to be used as a power on
indicator. If you do install R9, then the LED
will always be on, but it will get brighter if
the applied voltage is within the range of the
lowest comparator.
There is an optional jumper (Jmp1) that
enables you to change the operation from
Dot mode to Bar mode. For this type of
application, I believe that Dot mode is more
appropriate. The 1.2-kΩ resistor (R8) on the
REF_OUT pin “programs” the LED current to
approximately 10 mA.
Both potentiometers specified in the bill
of materials are single-turn and adjustable
from the top. I happened to have some small
multiturn pots in my parts bin, which I used.
The pin spacing of the parts I used easily fit
into the pattern for the single turn pots.
PRINTED CIRCUIT BOARDS
I designed two PCBs for this circuit:
one with all through-hole components and
one with SMD resistors and capacitors. The
“mostly” SMD board still uses the DIP versions
of both the LM3914 and LED array. It also
uses the through-hole part for the voltage
regulator because the SMD part actually
takes more board space. I did the SMD board
mainly to see if it would yield a smaller PCB.
As it turned out, both boards are exactly the
same size. I had to add some ground vias
on the SMD board because I could not run
traces between components. I probably could
have made the SMD board smaller by putting
components on both sides of the board, but I
wanted to have all the traces for the two DIPs
on the bottom so that I could etch my own
board and not have to worry about not having
plated through holes.
I calibrated my unit so that LED 1 turns
on solid at 108 VAC and LED 10 is on at 135
VAC. This yields a span of about 3 V per LED
and with LED 5 turning on at 120 VAC. You can
see in Photo 1 that LED 6 is on solid and that
LED 7 is starting to come on. This indicates
circuitcellar.com
PHOTO 1
LED 6 is on solid and that LED 7 is
starting to come on, which indicates
that the voltage is a little higher than
123 VAC.
that the voltage is a little higher than 123
VAC, probably about 124 VAC, which is what I
measured with my DVM.
KEY COMPONENTS
FEATURES
All of the parts listed on the BOM are from
DigiKey except H1. If you send a check with
the order you will not have to pay shipping
charges, but be sure to include whatever sales
taxes are appropriate. The BOM does not list
an enclosure. I used one from my stock of
parts. It looks like it is a small aluminum
project enclosure from Radio Shack, which I
spray painted dark blue.
H1 and its associated parts are from
Jameco. I use these headers and associated
parts, for the interconnections on most of
my projects so I have quite a few on hand.
For this project, it is probably not worthwhile
to purchase the parts necessary for this
connector. I would just solder leads from
whatever voltage source you are monitoring
to the pads for this header.
Schematics and PCB layout files for both
versions are available for download from my
website, www.qsl.net/k3pto/. I use DipTrace
(www.DipTrace.com) for all of my designs.
There is a fully functional free version
available if you want to try it.
ABOUT THE AUTHOR
Larry Cicchinelli holds a BSEE from the Drexel Institute of Technology and an
MSES from Pennsylvania State University. He was technical support manager
at ZWorld, Rabbit Semiconductor, then Digi International (Rabbit Brand) from
2000 to 2012. From 1967 to 2000, Larry worked for Ford Motor Company. He
has been licensed as K3PTO since 1961. Circuit Cellar published Larry’s book,
Assembly Language Essentials, in 2011 (www.cc-webshop.com).
RS485/422/232/TTL
ASC24T
IBS485HV
RS232<=>RS485 ATE Converter
5 Port Isolated
RS485 Repeater
$349.00
$45.00 board only
Enclosur es, Cables,
Power Supplies and Other
Accessories
INTRODUCING THE SMFCOMX!
• Converters
• Repeaters
• Multi-Repeaters
• Hubs
• Fiber Optics
• Isolators
43
• Extended Distance Units
• Serial to Digital I/O
• Large Multi-Drop Networks
• Custom Units &
Smart Units
• Industrial, 3.0 KV Isolation
Call the RS485 Wizards
513-874-4796
www.rs485.com
44
THE CONSUMMATE ENGINEER
COLUMNS
The Not So Humble Capacitor
Capacitors are important passive, two-terminal devices in electrical
circuits. The more you know about have these components the better
you’ll do when designing electrical systems. This article details common
capacitor types and covers basic capacitance-related equations.
By George Novacek (Canada)
G
erman Ewald Georg von Kleist invented
the capacitor in 1745. The Dutch engineer
Pieter van Musschenbroek improved it a year
later. The capacitors were called Leyden Jars
after the University of Leiden. In this article I’ll
reintroduce you to the capacitor by detailing
common capacitor types and covering basic
capacitance-related equations.
CAP BASICS
In most electronic circuits, capacitors
are just as abundant as resistors. Capacitors
are passive, two-terminal devices. Their
fundamental principle of operation is the
storage of energy in an electric field. In terms
of their construction, they comprise two
conductive plates with an insulator between
them. The insulating dielectric material—
which includes air—affects many parameters,
such as the capacity with respect to the
plates’ areas and distance, temperature
characteristics, working voltage, size, and
so forth. Figure 1 shows commonly used
capacitor types’ schematic diagram symbols.
When there is a potential difference (i.e.,
voltage between the capacitor plates), an
electric field develops across the dielectric
with positive and negative charge collected
on the respective plates. No DC current flows
through a fully charged, ideal capacitor. Such
a capacitor is described by its capacitance: C=
Q/V. The unit of capacitance is one farad (F),
which is one Coulomb (the unit of electrical
charge) per volt. Figure 2 shows a capacitor
being charged by a DC current.
The capacitor voltage V(t) increases
exponentially—as is expressed in the following
equation—until it reaches the battery voltage
V BAT less one electron.
V (t ) = Vbat (1 − e
−
t
RC
)
The product of R times C is referred to
as time constant RC, which is used when
calculating the charge and discharge voltage
with respect to time, such as in timers or
filters.
In AC circuits, the capacitor behaves as a
reactance, where “i” is an imaginary unit with
the property of i2 = –1.
Z =−
i
2π fC
It indicates that the AC voltage lags the AC
current by 90°. In electrical engineering, “j” is
often used instead of “i” to prevent confusion
with the sign for immediate current. RC filters
behave according to the ratios of resistive to
reactive impedances.
Circuits where both a capacitance and an
inductance are present act as resonant (tank)
circuitcellar.com
circuits, resonating at a frequency where both
the capacitive and the inductive impedances
are equal. The resonant frequency is:
f =
FIGURE 1
Fixed, polarized and variable capacitor
schematic symbols
+
1
2π LC
Unlike resistors, capacitors connected in
parallel add their capacitances:
Fixed
Polarized
Ctotal = C1 + C2 + C3 + ... + Cn
Variable
FIGURE 2
Charging capacitor by DC current
R
Ctotal
=
1
1
1
1
+
+
+ ... +
C1 C2 C3
Cn
To calculate the resulting capacitance
for two series capacitors ( just like for two
parallel resistors), you can use a simplified
formula:
Ctotal
C1× C 2
=
C1 + C 2
ESR VS. ESL
Capacitors store or release energy.
There is no energy dissipated in ideal
capacitors. Therefore, no heat is generated.
V(t)
C
VBAT
COLUMNS
In series, the inverse capacitances are
added:
1
45
Unfortunately, no capacitor is perfect due to
parasitic properties such as leakage of the
dielectric, leads resistance, inductance, and
so forth. Consequently, there is some energy
dissipation as one would see with resistors.
Specifications of polarized capacitors state the
equivalent series resistance (ESR). This is not
an actual resistance, but has the same effect,
adding a real component to the capacitor’s
TABLE 1
Standard capacitor values
pF
pF
pF
nF
nF
nF
µF
µF
µF
G
G
1.0
1.1
1.2
1.3
1.5
1.6
1.8
2.0
2.2
2.4
2.7
3.0
3.3
3.6
3.9
4.3
4.7
5.1
5.6
6.2
6.8
7.5
8.2
9.1
10
11
12
13
15
16
18
20
22
24
27
30
33
36
39
43
47
51
56
62
68
75
82
91
100
110
120
130
150
160
180
200
220
240
270
300
330
360
390
430
470
510
560
620
680
750
820
910
1.0
1.1
1.2
1.3
1.5
1.6
1.8
2.0
2.2
2.4
2.7
3.0
3.3
3.6
3.9
4.3
4.7
5.1
5.6
6.2
6.8
7.5
8.2
9.1
10
100
1.0
10
100
1.0
10
15
150
1.5
15
150
1.5
15
22
220
2.2
22
220
2.2
22
33
33
3.3
33
330
3.3
33
47
470
4.7
47
470
4.7
47
68
680
6.8
68
680
6.8
68
F
Capacitances
from about
0.047F up to
about 10,000F
exist
46
COLUMNS
PHOTO 1
Different types of capacitors
impedance. In applications such as switching
regulators, low ESR is critical for efficient
operation. Capacitor leads add parasitic
inductance to the characteristics. This is
called equivalent series inductance (ESL). The
ESL becomes important at high frequencies
as it can cause unwanted resonances.
UNITS
ABOUT THE AUTHOR
George Novacek is a
professional engineer
with a degree in Cybernetics and Closed-Loop
Control. Now retired, he
was most recently president of a multinational
manufacturer for embedded control systems
for aerospace applications. George wrote 26
feature articles for Circuit Cellar between 1999
and 2004. Contact him
at gnovacek@nexicom.
net with “Circuit Cellar”in the subject line.
The unit of capacitance is one farad, which
is an extremely large capacitance. Only in
recent years have super-capacitors (also called
ultracapacitors) with capacitances on the order
of several farads become available at reasonable
prices. The capacitance values commonly used
range from 1 pF (picofarad = 10 –12 F) to several
millifarads (1 mF = 10 –3 F). A millifarad is
commonly designated as 1 G. Strictly speaking,
this is incorrect. Historically, schematic
diagrams showed capacitor values without a
unit. The convention was that such values were
automatically considered in microfarads (1 µF =
10 –6 F). And since 1 mF = 1,000 µF, those values
were called gigafarads (G).
CAP NOMENCLATURE
This brings us to the nomenclature.
In North America capacitor values were
understood to be in microfarads, unless a
unit was specified. This was generally for
capacitances from 1,000 pF (0.001 µF) and
larger. Below 1,000 pF, just as in Europe, a
“p” or “pF” (meaning picofarad) is used. From
10 –9 to 10 –6 F, the unit nanofarad (nF, or just
n or sometimes k, which means thousand
picofarads) is used. Some prefer to designate
RESOURCES
Cardwell Condenser, http://cardwellcondenser.
com/.
N. Gruending, “Varactor Diodes, Weird Component #6,” Circuit Cellar 287, 2014.
G. Novacek, “Impedance Matching,” Circuit
Cellar 281, 2013.
values between 100 n and 1 µF as a fraction of
a microfarad. A microfarad (µF, uF, or u or M)
is 10 –6 F to 10 –3 F. Above 10 –3 F, we use G and
super-capacitors are in Farads (F).
It should also be noted that, similar to
resistor values, European designations do not
always use a decimal point. A 2.2 µF capacitor
could be identified as 2.2 in an American style
diagram, while in an European diagram it could
also be 2M2 or 2u2. A 0.022 capacitor would
be designated as 22n, while 0.0022 would be
2n2. Over the years, I’ve seen a trend away
from the American MIL style designations
towards the European style. Table 1 shows
generally available capacitor values.
CAP STYLES
There are three fundamental styles of
capacitors: fixed, polarized, and variable.
Let’s cover each one.
Fixed capacitors are found in electronic
circuits in great numbers as by-pass
capacitors, members of RC networks in
timers and filters, and so on. Every integrated
circuit (IC) should have at least one bypass
capacitor. The energy they store filters out
digital circuits’ switching transients, bypasses
power supply noise to ground, and suppresses
ground loops. Other than for bypassing, there
is not much use for capacitors in digital
circuits, but they are essential as storage
devices in power supplies. In analog circuits
capacitors are instrumental in bandwidthaltering filters, transient protection, timing
circuits, decoupling and blocking DC current
while allowing AC current to pass.
Polarized (i.e., electrolytic) capacitors are
mainly used in power supplies because of their
large capacitance-versus-volume ratio. The
most common are aluminum and tantalum
capacitors. Storage capacitors, especially
in switching power supplies, need low ESR
to minimize the losses at high switching
frequencies. In 60- or 50-Hz rectifiers, the
ESR is generally not critical.
A common task is to determine the size
of a storage capacitor to maintain a system
operation during a power interrupt. The
allowed voltage drop during the interrupt
is small compared with the full operating
voltage, so the differential equation can be
linearized and the needed capacitance is C =
(I × t)/V. Note that I is the average current
drain in amps, V is the allowed voltage drop
in volts, and t is the maximum duration of the
power interrupt in seconds.
Variable capacitors fall into two categories.
Trimmer capacitors, analogous to resistor
trim potentiometers, are shown in Photo 1.
The two capacitors in the center use air for
the dielectric. Trimmers with solid dielectrics
(the yellow stubby cylinder) are smaller,
circuitcellar.com
VARICAPS
Varicaps
are
diodes
that
exhibit
capacitance dependent on the magnitude of
their negative bias. The major limitation of
varicaps is that the radio frequency (RF) signal
they are tuning into resonance must be small
compared to the bias to minimize modulating
of the set capacitance. This can be to some
COLUMNS
usually less expensive and have lower quality
factor Q. Typically, trimmers come in the 1 to
50 pF range.
Variable capacitors, such as the one shown
in Photo 2, are no longer as prevalent as they
once used to be. Often used in radios, these
tuning capacitors comprised several separate
capacitors on a single shaft. The capacitor in
Photo 2 has two sections for the AM band and
four for the FM band.
These capacitors were bulky, heavy and
costly, although cheap units made with a solid
dielectric were available too. Worse, they
couldn’t be electrically adjusted, short of using
a complicated electromechanical subsystem.
This was a fundamental requirement for
scanners and synthesized digital tuners
found in essentially every modern receiver.
This dilemma has been solved by the arrival
of varactor diodes, also known as varicaps.
degree alleviated by using two varicaps backto-back, but when a large voltage is needed
across the tank circuit, it is the old technology
we must revert to. Variable capacitance
principle is used in numerous sensors, but
this is a subject for another article.
I hope you found this review of capacitors
useful as I have. The more you know about
how they work, the better you’ll implement
them in your future designs.
PHOTO 2
Air tuning capacitor
The Premier Business Development
Event for the Global Robotics Industry
October 15-17, 2014
Meet industry leaders, investors and new customers who
can accelerate your business at this one-of-a-kind event.
Hynes Convention Center
Boston, MA
Join over 1,200 industry leaders for:
Register Today!
robobusiness.com
800-305-0634
SAVE OVER 10%
when you register
by October 14
Use Code RB2003
•
•
•
•
2 Intensive Workshops: Industrial Robotics & Startups
32 Targeted Conference Sessions
7 Visionary Keynotes from Global Industry Leaders
Extensive Networking Opportunities
Is your time in Boston limited?
Expo Pass now available for October 16 only.
Visit robobusiness.com for details.
Founding Sponsor
Media & Association Partners
47
Platinum Sponsors
Gold Sponsor
Premier Analyst Sponsor
48
EMBEDDED IN THIN SLICES
COLUMNS
Linux System
Configuration (Part 3)
Build Your Linux System
This is the final installment in a multipart
article series dealing with configuring Linux for
embedded systems. In this article we will look
at some of the ways available to you to build the
embedded Linux system you need.
By Bob Japenga (US)
F
all is in the air and it is the time of the year
I love best—especially in New England.
Contrast that with building a new version of
embedded Linux for a project. As a manager,
I cringe every time one of my teammates tells
me that they need to build (for the first time)
a new version of Linux.
There were times when we designed a
board from a reference design and built
everything from a ready-made distribution
from the chip maker. That was easy. A
couple of quick cookbook-style steps here
and there. Perhaps configuring a new flash
memory chip or RAM chip. Compile. Build.
Done. Simple. But then, at some point later,
we discovered that our version of Linux didn’t
have the driver we needed for some piece of
hardware or it didn’t support some software
functionality. This recently happened to one
of our customers who delayed implementing
a software function until late in the project.
When we were asked to implement it, they
already had special drivers that only worked
with their version of Linux to interface to
their proprietary hardware. But they needed
a streaming video USB gadget that wasn’t in
their version of Linux. We had a dilemma. We
had to either bring all of the proprietary code
over to a new version of Linux or back-port
the code from a later version of Linux into
the older version. It was time for the antacid.
There was no easy answer. We knew cost
overrun and schedule slips were imminent.
This can happen, and it is painful. Very often,
the chip manufacturers don’t mainstream their
changes (i.e., put them into a stable release at
kernel.org). That can be understandable in the
fast-paced world of hardware development. If
asked, they will say that they haven’t done it
yet. If you have to bring that code into a new
version of the kernel, you will need to bring
all of their changes that they made to get the
reference design running into the new kernel.
One time we used a reference design and very
late in the life cycle discovered a bug in the
C library. The manufacturer not only didn’t
mainstream their changes, they couldn’t find
the source code they used for the C library.
What a nightmare.
These days we are smarter and things are a
little easier. Thanks in part to OpenEmbedded,
Yocto, and Ångström. This month I’ll cover
each of these. We have used all three. There
are other options, like Linux From Scratch,
but since we are only examining it in thin
slices, I am only going to cover the slices we
have actually used.
OPENEMBEDDED
In Part 2 of this article series, we discussed
the concept of a Linux distribution. These are
things like Ubunto, Fedora or, in the embedded
circuitcellar.com
BITBAKE
In the past, developers have used makefiles
to build the Linux kernel and the associated
applications. The kernel was configured as
discussed in my June 2014 article, in which
I described using script files to tie the whole
process together. We used this method on
many projects early in our deployment of
Linux. Once in place, it worked okay. The
problem came when we had to make changes
to the version—of either the kernel or its
associated libraries or the applications.
Another problem had to do with the
tool chain. The tool chain needs to be part
of the build process during development. A
different tool chain can result in your source
code not building or building properly. Only
once have we used the same build process
from one Linux project to another. And that
was because the hardware was identical. This
is a significant problem. The OpenEmbedded
organization was formed to help us with that.
What OpenEmbedded brought to the table was
a new tool and specified a new language to
define the architecture. That tool was BitBake.
Think of BitBake as make on steroids.
The inputs to BitBake are called “Recipes.”
BitBake takes recipes and metadata to
create a complete Linux distribution. It is
not necessary to have a build script (i.e.,
write code) to create most distributions. The
language is powerful enough to do everything
you need. And if it has shortcomings (they all
do), it does allow you to use Python to create
more complex operations. It is flexible enough
to handle multiple hardware architectures
and helps you to manage and create multiple
releases for those targets.
METADATA
So what is this metadata that BitBake
uses? Think about what you need to know
to create the Linux kernel. You need a
configuration file. You need to know where
your tool chain is to compile and build the
kernel. You need to know where the source
files are located. Ideally, they should be
available on the Internet. You need to know
complier options. And you need to know a
series of dependencies (i.e., if you change
this file, other files need to change). Building
a package has all of the same requirements.
All of this gets wrapped up in what is called
the metadata.
RECIPES
Okay, you are saying: “This is sounding a
little cheesy.” Baking. Recipes. What are these
recipes? If BitBake is make on steroids, recipes
are makefiles on steroids. A given recipe tells
BitBake exactly the who, what, where, and
how for building a particular package. Who
is the name of the package. What is the type
of package. Where is the location. And then
How to build it. The end result is a file system
image and a kernel image that you can put on
your target system.
YOCTO
Wow! That sounds great. But what’s
lacking? Why did Yocto need to come along?
Basically, Yocto picks up where OpenEmbedded
left off. More heavily funded, it accomplishes
all of the same goals that OpenEmbedded set
out to attain. But Yocto relies heavily on two
things from OpenEmbedded: BitBake and the
metadata. OpenEmbedded remains a separate
organization from Yocto and is concentrating
on making BitBake better and providing
metadata for a host of architectures. Yocto
is concentrating more on the tools and the
board support packages (BSP) for a wide
variety of architectures and processor boards.
For example, OpenEmbedded maintains the
metadata for something like the BeagleBoard
and the Raspberry Pi boards, but the Yocto
project is the place to get the BSP.
Another way to look at Yocto is to think
of it as a place where silicon providers like
Atmel, Texas Instruments, and Freescale can
create a common framework for the release
of Linux for their new processor or core. For
us as users, that is a win-win situation.
Another benefit for developers occurs
when we choose to use a commercially
supported release of Linux. Perhaps we
need to meet the FDA requirements for
Software of Unknown Provence (or Pedigree)
(SOUP). Yocto helps because all of the major
vendors are members of Yocto. So, if you
do all of your development on a particular
Yocto version and then need the kind of
support that these commercial packages
COLUMNS
world, Ångström. A distribution is a complete
set of tools, the kernel, the packages, and
the build instructions to create a version
of Linux. OpenEmbedded provides you the
framework to enable you to create your own
distribution. From the name, you will gather
that it specializes in creating an embedded
version of Linux, which is very different than
starting with a distribution like Ubunto.
The problem for us as designers is how to
create a distribution for our project. Over the
years, we have seen a number of attempts to
help us. The OpenEmbedded project was a big
step forward in this process. I won’t bore you
with the history of how this evolved, but the
major contribution that the OpenEmbedded
organization provided to our community was
the development of BitBake and the structure
for metadata used to define your system.
Let’s look briefly at each of these.
49
50
reference design or an SBC that is supported
by Ångström, you will probably have the
most tools and flexibility for maintaining your
design throughout its lifecycle.
ABOUT THE AUTHOR
COLUMNS
Bob Japenga has been designing embedded
systems since 1973. In 1988, along with his
best friend, he started MicroTools, which specializes in creating a variety of real-time embedded systems. With a combined embedded
systems experience base of more than 200
years, they love to tackle impossible problems
together. Bob has been awarded 11 patents in
many areas of embedded systems and motion
control. You can reach him at rjapenga@microtoolsinc.com.
STICK WITH LINUX
provide, you can seamlessly switch. This is
not true if you use just any old distribution.
We found this on one project when we
evaluated going to a commercially supported
version of Linux from Wind River. We were
evaluating this because of the processes and
documentation that they provided for FDA
approval of a medical product. Going with the
commercially supported release was going
to save us considerable time and money
for documentation. The problem was that
we used a version of the kernel from Texas
Instruments that came with our development
kit and that version was not supported by
Yocto. Creating a new distribution for the
closest version supported by Wind River
would probably cost more than the benefit
received by having the documentation and
support provided by Wind River.
ÅNGSTRÖM
If all of these names haven’t gotten your
head spinning, let’s just talk about one more.
Ångström heavily relies on OpenEmbedded
and Yocto to create a distribution specifically
targeted for embedded systems. It is highly
scalable. It can run on a device with as little
as 4 MB up or on a system with several
terabytes. But of course the question is,
what can it do with only 4 MB? There are a
host of off-the-shelf boards that it supports
including the popular BeagleBoard and Beagle
Bone open source hardware platforms as well
as the Raspberry Pi. So, if you are using a
Where are we to begin as system
designers? We are offered a plethora of
choices. Let’s consider them in order of
decreasing complexity resulting in decreased
cost and time: we can certainly build Linux
and the subsequent applications from kernel.
org; we can use a tool like “Build Linux from
Scratch”; we can use a board-support package
provided by a hardware vendor that is not part
of Open Embedded, or from one that is part
of Open Embedded; we can build our design
using a project already supported by Yocto;
or we can start with an existing distribution
(Ångström, or Ubunto, or Debian) and go on
from there. Our experience is that the cost
savings across the life cycle are significantly
larger the later we go in that list of options.
A few years back, hardware venders of x86
SBCs could tout simplicity by providing an SD
card or Compact flash with Ubunto or Debian
on it. Everything was there for you. You could
concentrate on your killer application.
Today, with Ångström, you can have
that for a wide range of processors and for
specifically an embedded system. If you
are rolling your own hardware based on
a reference design, make sure the silicon
provider is using OpenEmbedded and ideally
has created a Yocto project.
We have come a long way in developing
easier ways to create, modify, and maintain
embedded Linux systems. We still have a long
way to go to achieve the full potential that
Linux offers. Sometimes, the very issues I
covered in this article make me want to stop
using Linux. But then where else would I go?
Hopefully, this article series has helped you
better understand how to configure the
kernel, select and build the applications, and
finally put together a distribution for your
next embedded Linux project. Of course, only
in thin slices.
RESOURCES
Ångström, www.angstrom-distribution.org
Open Embedded, www.openembedded.org BeagleBone, http://beagleboard.org/bone Ubuntu, www.ubuntu.com
BeagleBoard, http://beagleboard.org
Wind River Linux, www.windriver.com/products/linux.html Fedora, https://fedoraproject.org
Yocto, www.yoctoproject.org Linux from Scratch, www.linuxfromscratch.org
Raspberry Pi, www.raspberrypi.org
CC Vault
Unlock the power of embedded design.
This pocket-sized vault comes fully loaded with every issue
of Circuit Cellar magazine and serves as an unparalleled
resource for embedded hardware and software design tips,
schematics, and source code.
From green energy design to ’Net-enabled devices,
maximizing power to minimizing footprint, CC Vault*
contains all the trade secrets you need to become a better,
more educated electronics engineer.
BONUS! Build your archive by downloading your latest-issue PDFs straight to
the drive! Personalize your CC Vault by adding Elektor or audioXpress issue archives,
available as an add-on during time of purchase, or your very own project files.
*CC Vault is a 16-GB USB drive.
Order yours today! cc-webshop.com
52
RETRONICS
Elektor ComputerScope (1986)
FEATURES
Goodbye Green-Only ‘Scope CRT
Almost 30 years ago, a trainee at
Elektor Labs designed an ambitious test
instrument like an oscilloscope with PC
support.
By Jan Buiting (Editor in Chief, Elektor,
The Netherlands)
B
ack in 1986 at the crowded Elektor
Electronics
laboratory,
somebody
approved the idea proposed by a young
Dutch BSc graduate of Electronics to make an
oscilloscope add-on for personal computers. I
had been appointed as an assistant editor just
months before, and being able to wield an
electric typewriter and an English Dictionary,
I was asked to write a deft letter to ITT
Semiconductors of Freiburg, Germany, asking
for a few samples of their latest UVC3101
device.
I was told it was a “high-speed” ADC and
DAC in one chip and “difficult to get.” I got a
favorable reply by telephone from Reinhard
W. Preuss, the most officious and gentlemanly
PR and Press Officer I have come across in
the electronics scene these past 25 years. A
few days later the ICs arrived, complete with
engineering notes. The lot allowed Rene van
Linden to start designing his ComputerScope
project and hopefully complete his BSc. thesis.
TIME PERSPECTIVE
In those days it was hot to make
computers do things other than the
usual spreadsheeting, PacMan’ing, word
processing and ASCII art doodling. It was
the exciting era of the BBC Micro, the Acorn
Electron, Philips MSX and the first IBM PC
clones starting to appear in households
and in electronics engineers’ workshops.
It was a time also when computers were
totally accessible in terms of hardware
and software. As long as a computer had a
port of some sort using an easy to obtain
connector, engineers, British school children
and Elektor trainees got into developing
peripherals the computer manufacturers
could not have foreseen. There was a lot
of POKE(16xxx) going on at that time. In
the case of the Elektor ComputerScope,
the idea was to make a personal computer
handle the visualization of measured data
on its monitor with composite video input.
Paradoxically, that monitor had a CRT just
like the traditional oscilloscope, only larger
and more colorful, so the crux or cost
advantage of the computerized ‘scope over,
say, a refurbished Tek I find hard to grasp.
In good Elektor tradition for large projects
from the lab, the magazine article consisted of
circuitcellar.com
53
two instalments. The first in September 1986
gave the circuit diagram, the basic method
of operation and some screendumps showing
not too complex waveforms. In the October
1986 edition the construction, adjustments
and software aspects were discussed, and
the project was carried on the front cover of
the magazine (see Photo 1). Note the visually
implied hierarchy of two home computers
pictured: the BBC Micro majestically in front
and a Commodore C64, well, also visible.
As a peripheral the ComputerScope
requires a fast and wide highway into the
computer’s memory and display systems.
However, these are digital, while the ‘scope
probe will pick up analog signals mostly, so
ADC-ing is called for. The port traffic is not
one-way either, as the ‘scope input settings
including the AC-DC selector and attenuator/
amplifier need to be controlled from the
computer. A couple of Clare DIP relays were
used for the purpose. There are no controls
or knobs on the Elektor ComputerScope to
fiddle with, hence the instrument looks
insipid in its grey case except perhaps for
the stylized S.
The crux of the circuit is found in the “fast”
ADC and DAC inside the UVC3101, and a small
memory section based on Inmos IMS1420
RAMs. These parts were rare and expensive
at the time—fortunately some advertisers in
the magazine started to supply kits of the
project.
The ADC and RAM section of the schematic
is reproduced in Figure 1. The DAC (not
visible here) was used to add an offset to the
input signal if required. The datawidth from
the ADC was 7 bits; bit 8 was the trigger
indicator. The memory area actually used for
data buffering was 512 bytes. The RAM section
had its own address counter built from TTL
ICs from the 74HC series (no microcontroller
in sight). The board was packed with 74HC
ICs for that matter, and a few straggling 74LS
and CD4000 ones, since a mass of control and
timing functions had to be catered for—like
the complete timebase. I counted 24 logic
ICs (see Photo 2), which would be labeled a
component grave these days.
The lot was powered by a separate supply
board type EPS 9968-5.
A small modification involving a slide
switch on the back panel (see Photo 3)
FEATURES
THE DESIGN
allowed the instrument to be used with IBM
(clone) PCs also.
Those of you yearning for the complete
article from 1986 are advised that the Elektor
1980’s DVD is about to be published; watch
www.elektor.com or your weekly E-Post
Newsletter.
THOSE LISTINGS
If you ordered the ComputerScope board
# 86083 through the Double Dutch named
Elektuur Print Service (EPS) it came with
ABOUT THE AUTHOR
Jan Buiting is Editor in Chief of Elektor magazine. His “Retronics” articles have been published regularly since 2004. Contributions, suggestions, and requests are welcome. Email
editor@elektor.com.
PHOTO 1
The ComputerScope project appeared
on the cover of the October 1986
eidtion of Elektor magazine.
54
FEATURES
FIGURE 1
The ComputerScope project appeared
on the cover of the October 1986
eidtion of Elektor magazine.
program listings on paper for the Acorn
Electron, the BBC Micro, the Commodore C64,
and MSX machines. I remember stacks of
this paperware on the shelves in our stores,
alongside the boards, and supplications on
my desk from readers all over the globe to
send them the listings on disk, or without
the board! To help owners of “other makes
of computer” to compile their own program
for the ComputerScope, a pitiful flow diagram
was printed at the end of the October 1986
instalment.
PHOTO 2
Take a look inside
LOVE STORY
Now compare the names of our two
editorial secretaries duly printed in the
October 1986 edition with those in November
1986—look on page 13. One surname
changed. Besides passing his BSc. Electronic
Engineering exam with good notes for the
ComputerScope project, lab trainee R (for
Rene) v(an) Linden, a silent, almost invisible
worker, in the course of 1985 also managed to
more than befriend “our” cherished secretary
W (for Wilma) Wijnen.
circuitcellar.com
55
PHOTO 3
A slide switch on the back panel
enabled the system to be used with
IBM PCs.
FEATURES
DUMPSTER STORY
Elektor moved offices in 2006 and in
an orchestrated attempt to “shake off the
dark past” the technical staff and editors
got strong advice to clear their stables and
make a fresh start at Elektor Castle. One
Friday afternoon, CEO attending, the old
lab space in Beek (Holland) was cleared out
rigorously by quasi volunteers, with a ton of
stuff disappearing in an x cubic feet dumpster
carefully parked on the driveway right under
the lab windows. There was beer and pizza
to celebrate the almost vacant building, and
reward the workers.
Of the twenty or so lab projects and
prototypes I was able to safeguard during
Operation Augias, most—I admit—were of the
‘finished’ type, meaning there was a case
around them. I managed to keep these out of
harm’s way and was happy to see that a few
colleagues did the same for sentimental
reasons. Today the small booty of that
boisterous Friday afternoon is incorporated in
the Retronics Collection, which is an “asset to
the company” under the current management,
and virtualized at www.elektor-labs.com/attic.
Look at the item on the shelf straight above the
Mugen Hybrid Amp—that’s the final resting
place of the 1986 Elektor ComputerScope.
TRACE 32 ®
Editor’s Note: This article first appeared in Elektor,
June 2014
advertisement_arm-overview.indd 1
03.06.2014 11:48:51
BOARDS, BOOKS, DVDs AND MORE AT WWW.ELEKTOR.COM/STORE
Elektor.STORE
The world of electronics
at your fingertips!
NE
!
W
Theory and best practices
Advanced Control
Robotics
It doesn’t matter if you’re building a line-following robot toy or tasked with designing a mobile
system for an extraterrestrial exploratory
mission: the more you know about advanced
robotics technologies, the better you’ll fare at
your workbench. Hanno Sander’s Advanced
Control Robotics simplifies the theory and best
practices of advanced robot technologies.
You’re taught basic embedded design theory
and presented handy code samples, essential
schematics, and valuable design tips (from construction to debugging).
160 pages • ISBN 978-0-963013-33-0 • $54
ore
This book and m
are available at
m/store
www.elektor.co
Books
Fun to build and use projects!
Create 30 PIC Microcontroller Projects
with Flowcode 6
The RPi in Control Applications
This book covers the use of Flowcode® version 6, a state-of-the-art, all-graphical based code
Raspberry Pi
Hardware Projects
development tool, for the purpose of developing PIC microcontroller applications at speed and
This book is about the Raspberry Pi computer
with unprecedented ease. Without exception, the 30 projects in the book are fun to build and
and its use in control applications. Dogan
use. A secret doorbell, a youth deterrent, GPS tracking, persistence of vision (POV), and an
Ibrahim explains in simple terms, with exam-
Internet webserver are just a few examples of projects in the book waiting to be explored and
ples, how to configure the RPi, how to install
mastered. This makes the publication a perfect source of projects constantly challenging your
and use the Linux operating system, how to
hardware and software skills as you progress, resulting in advanced microcontroller applica-
write programs using the Python program -
tions you can be proud of. All sources referred in the book are available for free download, in-
ming language and how to develop hardware
cluding the support software.
based projects.
232 pages • ISBN 978-1-907920-30-1 • $48
290 pages • ISBN 978-1-907920-29-5 • $54
Prices and item descriptions subject to change. E. & O.E
Books
Bestseller!
Learning to Fly With Eagle
Eagle V6
Getting Started Guide
110 Elektor Editions, Over 2500 Articles
DVD Elektor
2000 through 2009
Android Breakout Board
This DVD-ROM contains all circuits and pro-
The FTDI FT311D is a flexible bridge that
The book is intended for anyone who wants an
jects published in Elektor magazine’s year
can interface your circuit to an Android
introduction to the capabilities of the CadSoft’s
volumes 2000 through 2009. The 2500+
smartphone or tablet. This Elektor Android
Eagle PCB design software package. The reader
articles are ordered chronologically by release
Breakout Board offers options for seven
may be a novice at PCB design or a professional
date (month/year), and arranged in alphabet-
digital outputs, four PWM outputs, asynchro-
wanting to learn about Eagle , with the intention
ical order. A global index allows you to search
nous serial and I2C and SPI interfaces. The
of migrating from another CAD package. After
specific content across the whole DVD. Every
board is compatible with Android 3.1 (Ho-
reading this book while practicing some of the
article is printable using a simple print func-
neycomb) or higher (Android Open Accessory
examples, and completing the projects, you
tion. This DVD is packed with ideas, circuits
Mode should be supported).
should feel confident about taking on more
and projects that are ideal for any electronics
challenging endeavors. This book is supplied
enthusiast, student or professional, regardless
with a free copy of Eagle on CD-ROM for
of whether they are at home or elsewhere.
MS Windows, Linux and Mac.
Ready-built module
Art..# 130516-91 • $41
ISBN 978-1-907920-28-8 • $121
208 pages • ISBN 978-1-907920-20-2 • $47
Elektor is more
than just your favorite
electronics magazine.
It’s your one-stop shop
for Elektor Books,
CDs, DVDs,
New!
Kits & Modules
The Ultimate Guide!
The LTspice IV Simulator
www.elektor.com/store
IO-Warrior Expansion
Board
Don’t throw out your old PCs and notebooks
Simulator describes the operation of the pro-
or leave them gathering dust in the basement!
gram, all available commands, the various
They can be a useful resource: by adding this
editors, dealing with SPICE models, the use of
universal interface card an old PC can be pres-
non-linear components and more. This book is
sed into service as a measurement and control
more than just a manual. It also offers a variety
hub. An IO-Warrior module on the I/F board
of tips, methods and examples, all carefully
illustrated using almost 500 drawings, diagrams and screenshots on high-quality paper.
The book is designed so that it is suitable for
both beginner and veteran SPICE users.
744 pages • ISBN 978-3-89929-258-9 • $67
takes care of USB communication, and source
Elektor US
USA
Phone: 860.289.0800
Fax:
860.461.0450
E-mail: order@elektor.com
code is available that works with the free
version of Visual Studio.
Ready-built module
Art.# 130006-91 • $54
Kits & Modules
In a sturdy, hard-cover format, The LTspice IV
and much more!
58
THE DARKER SIDE
COLUMNS
RF Connectors
Tips for Taking
Care of Connections
Never take RF connectors for granted. Why? An RF connector can be a delicate
assembly of small mechanical parts with strict tolerances. This article details how
to properly select and care for RF connectors.
By Robert Lacoste
W
elcome back, dear reader. In the past
few years, I wrote several columns
dealing with radio frequency. When working
on RF projects, you must connect subsystems
together at a given time. You might also have to
use RF test equipment, like a spectrum analyzer.
Let’s assume that you have an RF transmitter—
for example, a Wi-Fi gizmo—and you need to
measure its output power. The first thing to
check is impedance matching. You know that the
output impedance of your transmitter is 50 Ω.
You also know that the input impedance of your
spectrum analyzer is 50 Ω, so you just have to
get a test cable with a characteristic impedance
of 50 Ω to interconnect them. If you have any
doubt about a cable’s characteristic impedance,
refer back my column on microstrip lines in
which I explained the concept (“Microstrip
Techniques,” Circuit Cellar 223, 2009).
So it’s easy, right? But, wait. This test cable
will need to have the proper mating connectors
at both ends. There are a hundred types. How
can you select the proper ones? Moreover, what
are the mandatory precautions required when
using RF connectors for any precise connection?
Is a connector as transparent to the signal as it
looks like? Take a seat. I’ll address these topics
in this article. And you may be surprised by
some of the answers.
WHAT A CHOICE
PHOTO 1
Here you see (from left to right) the ubiquitous BNC, SMA, and, lastly, the terrible RP-SMA. You can see their
male and female variants.
I will start our journey with a tour of the
RF connectors you will most likely encounter
(see Photo 1). Firstly, consider the ubiquitous
Bayonet Neill-Concelman (BNC) connector,
which is present on nearly all oscilloscopes and
signal generators. It is easy to use thanks to
circuitcellar.com
COLUMNS
its two bayonet fingers and quarter-turn lock.
Although BNC connectors are affordable and
easy to use, they’re barely usable above a
couple of gigahertz, as I will explain later. Old
timers may also recall that BNC connectors
were largely used for network connections,
when 10B2 Ethernet was the norm. A caution
here: the BNC connector exists both in 50- and
75-Ω variants, the latter being devoted to video
applications. And no clear indication enables
you to differentiate them as they physically
mate. The only solution is to store them apart,
or maybe to check which cable type is connected
to them, but that might not be easy.
For more professional designs or for slightly
higher frequencies, the most common RF
connector is the sub-miniature A (SMA), which
you can see in Photo 1. SMA is a precision 50Ω, screw-type connector developed in the 1960s
by Bendix Scintilla Corp. SMA can be used up
to 18 GHz thanks to their precision engineering
and polytetrafluoroethylene (PTFE) insulator.
Some manufacturers even offer them up to
26 GHz. Are there some variants? You bet so.
For high-performance instrumentation, you
may encounter the so-called 3.5- and 2.92mm connectors. These ones are physically very
similar to SMA; however, they provide much
higher performances, up to 46 GHz or so, thanks
to their stricter tolerances and air-insulator.
Theoretically, they can be freely mated with
SMA connectors but with great care. More on
that later.
SMA is great, but then someone invented the
reverse-polarity SMA, or RP-SMA (see Photo 1). It
is the same connector as the SMA, but the male
and female screw and pin pairs are permuted.
These RP-SMA are frequent on Wi-Fi endpoints
and antennas. I hate them, just because there
is now a 50% probability to use the wrong
variant between SMA and RP-SMA. Depending
on whether or not you are lucky, mating the
wrong pair will result in either an invisible open
connection or a pair of damaged connectors. It
seems that RP-SMA was invented to make the
lives of Wi-Fi users more difficult—namely, to
make the lives of Wi-Fi users more difficult. The
idea was to forbid them to replace the supplied
antenna with a higher gain one, thanks to the
use of an exotic connector. This helped to avoid
issues with the FCC regulations. But now this
makes everydody’s life more difficult!
What about larger connectors? Photo 2
shows you the Amphenol Precision Connectors
7-mm (APC-7) high-performance connector.
This costly device is a clever sexless connector
that was used on all high-end RF test equipment
some years ago with up to more than 20 GHz of
exceptional repeatability. Even if they provided
an incredibly low VSWR, they are now replaced
by higher performance and more compact
precision connectors. Anyway, I love them.
59
PHOTO 2
Here I show (from left to right) the precision APC7 in male and female configurations, N male and female, and
UHF male and female. Bottom row includes F and Belling-Lee TV connectors.
The N connector is for lower-end applications
(see Photo 2). (It was invented in the 1940s by
Paul Neill. Do you remember the “N” of BNC
? That’s the same guy. Two very common
connectors have his initial.) The first versions
were usable up to 1 GHz, but they quickly
were improved for operation up to 18 GHz
or so by Hewlett-Packard. These heavy-duty
connectors are able to support a significant
PHOTO 3
Take a look (from left to right) at SMB, SMC, MMCX, and UFL miniature connectors (each in male and female
variants)
COLUMNS
60
PHOTO 4
A torque wrench is mandatory for ensuring the long life of a connector, especially when you’re working with
an expensive high-performance connector.
amount of power, up to 500 W at 2 GHz. The
N connector is also available in 50- and 75-Ω
versions (even though I haven’t worked with the
latter). I added some 75-Ω components to Photo
2—like the so-called “UHF,” “F,” and “BellingLee” TV connectors—just so you could clearly
differentiate them from 50- Ω lab connectors.
What’s the trend in the RF connector’s
arena? Miniaturization, of course. They are
few N or BNC connectors in mobile phones
(even though my first GSM phone had an SMA
connector for its external rubber antenna).
Have a look at Photo 3 for some more compact
varieties, starting with the SMB and SMC. These
two are as old as the SMA (from the 1960s),
but they’re significantly smaller with a jack
diameter of 3.7 mm. Both are identical except
for the locking mechanism: snap-on for SMB
and screw-on for SMC. As a consequence, the
SMC provides better performances than SMB.
They are respectively usable up to 4 and 10
GHz. The micro-miniature coaxial (MMCX)
is smaller. You can find them on Wi-Fi cards,
GPS receivers, and PDA antenna connectors.
Developed in the 1990s, MMCX is usable up to
6 GHz, which is not bad for a snap-on, 2.4-mm
connector. But engineers never stop, and one
of the smallest I’m using regularly is the UFL
connector (see see the rightmost component in
Photo 3). Designed as far as I know by Hirose,
the UFL is also a 6-GHz, snap-on connector with
a diameter of 2 mm and a mated height of only
2 mm. Of course, the connecting cable must be
ABOUT THE AUTHOR
Rob er t L ac os te li ve s i n Fr anc e,
n e a r Pa r i s. H e h a s 2 5 ye a r s o f e xp e r i e n c e i n e m b e d d e d sys t e m s,
analog designs, and wireless telecommunications. A prize winner in more than 15 international design contests, in 2003 he started his consulting company, ALCIOM, to share
his passion for innovative mixed-signal designs. His book (Robert Lacoste’s The Darker Side) was published by Elsevier/Newnes
in 2009. You
can reach
him at rlacoste@alciom.com.
Don’t forget
to put “darker side” in
the subject
line to byp ass sp am
filters.
PHOTO 5
This is my HP8510C vectorial network analyzer. It’s nice piece of equipment.
circuitcellar.com
as small. Typically, these UFL connectors are
bought clamped on a 0.8-mm coaxial cable.
I have only scratched the surface of the
long list of RF connectors. But I guess I have
presented the key ones. If you are looking for
more exotic variants, or if you have any doubt
about any particular variation, I recommend
reading Pasternack’s “Connector Identifier,”
which is listed in the Resources section of this
article.
61
a)
CONNECTOR CARE
COLUMNS
Experience shows that you must always keep
two simple rules in mind: RF connectors are
fragile, and checking them with an ohmmeter
definitively isn’t safe. I’ll elaborate on the
latter in a minute, but let’s stay for a while on
the “fragile” side.
An RF connector, and especially a small
one, is a delicate assembly of small mechanical
parts with strict tolerances. If you aren’t
careful, there’s a high probability that its
performances will be quite far from what you
might expect.
Firstly, keep in mind that a connector has
a given lifespan, and it could be short. For
example, SMA connectors are rated for up
to 500 mating cycles. That could be fine, but
it could also be orders of magnitude lower
than what you would need for benchtop test
equipment. So, it must be anticipated at the
design phase when selecting the connector.
Another very common problem is that
connectors get stuck. Many people use pliers
to screw or unscrew stuck connectors. Never
ever do that! Pliers can apply high torque on
the delicate parts and can easily damage them.
Moreover, with pliers, you can’t ensure that the
torque is applied perpendicularly to their axis.
Large connectors, like N, are designed to be
screwed by hand. For smaller ones, like SMA,
use a proper torque wrench (see Photo 4). If
you don’t have one, use your fingers. If you
can’t, use a properly sized wrench, but do so
with great care.
Another way to destroy a SMA connector
is to spin the mating pin inside its female
counterpart. Always mate the two connectors
with a straigth movement, and then screw
them without turning their assemblies—just
the nut!
A last word on the SMA. As I explained,
low-cost SMA and precision 3.5 mm/2.92 mm
connectors can be physically mated together.
However, it generally isn’t a good idea. A
precision connector can cost several hundred
dollars, so it would be a bad idea to mate it with
a low-cost connector that could be damaged
(due to its long life). If you need such an adaptor,
buy a high-quality SMA/SMA connector, screw
it in with care on the precision connector,
and leave it. Of course, the performance will
b)
PHOTO 6
a—The first test was with two SMA-to-SMA high-quality adapters. b—The extra losses are minimal, close
to 0.2 dB in high frequencies. The return loss is better than 20 dB (horizontal: 45 MHz to 10 GHz; vertical :
1 dB/div for S21 and 5 dB/div for S11).
be reduced to the performance of a standard
SMA. (Refer to Microwave101.com—which is
listed in the Resources section of this article—
for addition information on topics such as
advice on cleaning connectors.)
LET’S EXPERIMENT
It is great to know that connectors are
sometimes problematic. Experimenting can
62
COLUMNS
a)
b)
PHOTO 7
a—I tested a pair of standard BNC-to-SMA adapters. b—Significant losses are visible from 5 GHz, with no
less than 3 dB losses at 10 GHz!
ial Connectors—1.9 mm or 2.4 mm Mated Height,” www.hirose.co.jp/cataloge_hp/
e32119372.pdf.
Microwaves101.com, “Connector Care,” www.
microwaves101.com/encyclopedia/connectorcare.cfm
RESOURCES
Amphenol RF, “Precision
Connector Series,” www.
amphenolrf.com/products/
precision.asp?N=0&sid=4A4D4A002301E17F&.
Hirose Electric Co., “Ultra
Small Surface Mount Coax-
———, “Microwave Connectors,” www.microwaves101.com/encyclopedia/connectors.cfm
Pasternack, “Connector Identifier,” www.pasternack.com/images/pdf/7%20Connector%20
Identifier.pdf.
SOURCES
HP8510C Network analyzer (now obsolete)
Hewlett-Packard | www.hp.com
help you to be convinced.
How can you measure the actual behavior
of a connector? Apply an RF signal to it, check
the signal at the other end, and determine
whether or not it is modified. Or, even better,
insert the connector in the middle of a 50-Ω
line and check if there the connector had
caused any signal reflection, which would
indicate an impedance mismatch. (I’ll explain
the advantage of this method in a minute.) You
can make measurements with a standalone RF
generator. But the good news is that there is a
dedicated test setup for such measurements—
the network analyzer.
Basically, a network analyzer includes
three subsystems: a tunable RF generator, a
measurement bridge to differentiate direct
and reflected signals, and a sensitive receiver.
Networks analyzers come in two flavors: scalar
(measuring only the changes in a signal’s
amplitude) and vectorial (measuring both
magnitude and phase). I wrote about such
equipment some time ago in my 2010 article,
“A Tour of the Lab (Part 2): Frequency Domain”
(Circuit Cellar 243). Photo 5 shows the nice
HP8510C vectorial network analyzer (VNA) in
my company’s lab.
Let’s cover how I tested an SMA connector.
First, I calibrated the VNA using a high-quality
SMA/SMA jumper cable. I then inserted a pair
of standard SMA/SMA adapters (SMA male/male
and SMA female/female, respectively). Photo 6a
shows the setup. How did the two adapters
impact the signal transmission ? Look at Photo
6b. The top blue line is the transmission loss
from 45 MHz up to 10 GHz, with a 1-dB/division
scale. The loss was very small, lower than 0.2
dB or so. The yellow curve on the bottom shows
the reflection coefficient with a 5-dB/division
scale. This reflection coefficient is, in decibels,
the power of the reflected signal level divided
by the power of the incident signal. It was lower
than 20 dB, which indicated good performance.
This proved that high-quality SMA connectors
perform well at least up to 10 GHz.
Now what about BNC? I replaced the two
SMA/SMA adaptors with BNC/SMA adapters, as
you can see in Photo 7a. What happened ? Look
at Photo 7b for the result. Now the attenuation
starts to climb at a frequency of 5 GHz and is
as high as 3 dB at 10GHz. This means that only
half the available power goes through such
a BNC connector! The reflection curve is also
interesting. Firstly, you can see that the first
small degradations of the signal at 5 GHz are
easier to see on the reflected measurement
that on the transmission behavior. Secondly, at
10 GHz, the measured reflection coefficient is
10 dB. Remember that only 50% of the signal
was transmitted. We now know that –10 dB
(meaning 10%) of the signal power is reflected
to the source. That implies 50% transmitted
circuitcellar.com
b)
a)
THE FINAL EXPERIMENT
My last experiment was a real heartbreaker.
I took a good SMA/N adapter and measured it. I
then voluntarily damaged it with a screwdriver.
I gently pushed aside one of the four central
spring contacts by 0.5 mm or so (see Photo 9a).
The damage was minimal, and the remaining
three contacts were untouched. Was the
connector’s performance impacted? You bet, as
Photo 9b proves. No significant change in low
frequencies, but no less than 2 dB extra loss at
10 GHz!
The final step in the experiment? I tossed
the connector in a garbage bin. The destructive
test cost me about €20. But it was worth it to
help me convince you that connectors are
members of a very delicate species. You must
respect them, take care of them, and select
them with great caution. In fact, you must love
them. If not, soon or later, they’ll take their
revenge and show you their darker side.
a)
PHOTO 9
a—I tested a slighly degraded N connector. b—I got 2 dB extra losses at 10 GHz!
b)
COLUMNS
PHOTO 8
a—I tested a 10-cm UFL jumper. b—I zoomed in on the behavior of the UFL jumper after
several cycles. You can see 0.2 to 0.3 dB variations.
and 10% reflected makes 60%. Where is other
40%? It’s dissipated inside the connector
assembly, as heat. This is not noticeable with
the couple of milliwatts generated by a VNA,
but the same test with a 100-W source may be
funny.
In another experiment I used an ultraminiature UFL connector (see Photo 8a).
Rather than measuring its insertion loss, I
used it as a reference and then performed
numerous insertion/removal cycles prior to a
new measurement. Photo 8b shows the result.
Variations of up to 0.2 dB were measured. Such
a change in signal level could be negligible—or
not, depending on your application. Just keep
in mind that it could happen!
63
64
FROM THE BENCH
Embedded Voice
Recognition (Part 1)
COLUMNS
Recognition as a Peripheral
With the right parts and a little knowhow, you
can add effective speech and voice recognition
capabilities to almost any application. This series
serves as a proper introduction to the basics of
embedded voice recognition.
By Jeff Bachiochi (US)
S
martphones have put so much power
in our hands that we tend to take some
functions for granted. Just look at Siri, Apple’s
voice command application. She requires a
ton of high-level processing to interpret your
requests. All this is best handled in the cloud,
and thanks to the Internet, this gives the
impression that your phone is truly “smart,”
when in reality it’s just gathering and playing
back audio. Dictation applications have been
available to your PC for years. Chances are,
however, that you are not using one. It seems
incorrect interpretation often introduces an
increased necessity for editing, which defeats
32-KHz Crystal oscillator
or RC Oscillator
(RC on chip)
Timers
(4)
FIGURE 1
NLP-5x family features a 16-bit DSP
with chip debug, 128-KB OTP code
space, three 16-bit ADC channels, two
microphone pre-amps, two 16-bit DAC
channels, direct speaker drive, 24-KB
RAM, five timers plus an RTC, and
LCD control. Additional peripherals
include motor control logic, USB1.1,
an SPI, a UART, an IR interface, an
analog comparator, power-on reset,
brown-out reset, and 79 configurable
general-purpose I/O lines.
Microphone
Two
Pre-amps
with gain
control
(one line
input)
Regulator
Stereo
16-bit DAC
DAC Out
PWM
Three channel
16-bit ADC
Comparators
(2)
P.O.R
and B.O.R
Watchdog
timer
Clock main
USB, SPI,
UART and I2S
Interfaces
General-purpose I/O
2 × 16-bit parts
1 × 8-bit port (40 I/O)
NLP-5X
General-purpose
16-bit DSP
Internal code
space
(128-KB OTP)
22-KB Data SRAM
2-KB Code SRAM
LCD, motor and
IR control blocks
External memory interface
16-bit data, 23-bit address
Speaker out
circuitcellar.com
3.58-MHz Oscillator
or research
ADC
Microphone
Pre-amp
and
gain
control
Timers (3)
Watchdog
timer
Digital
filters
Audio
wakeup
DAC
DAC Out
PWM
RSC
General-purpose
microcontroller
AGC
Comparators
(4 input)
General-purpose I/O
3 × 8-bit ports (24 I/O)
its purpose as a labor-saving application.
These applications require a level of contextual
understanding that exceeds their abilities.
Because
of
high
computational
requirements, we’ve not seen this technology
trickle down into microcontroller applications.
However, the technology can be offloaded and
exploited as a peripheral device. Sensory,
Inc. designs highly accurate, low-cost voice
recognition and synthesis ICs. Its audio and
speech technologies are directed at four
areas: speech recognition, audio, interactive
and robotics, and voice recognition for
Bluetooth products. The recognition-related
areas are covered by two devices. The NLP-5x
uses a 16-bit DSP (see Figure 1). The RSC-4x
is an 8-bit general-purpose microcontroller
(see Figure 2).
Sensory has a complete line of development
and evaluation tools that gives you control
over high-performance speech I/O features,
including noise-robust speaker-independent
and speaker-dependent recognition of many
languages, speech synthesis and sound effects
using Sensory SX synthesis technology, MIDIcompatible music synthesis, voice record
and playback, audio wake-up, and touch
tone (DTMF) output. These audio and speech
technologies can reside together with the
application program creating a true singlechip application. Once written, the application
will reside in the processor’s ROM and can be
reproduced in a 100-pin LQFP package or in
unpackaged die.
The development costs can put this out of
reach for many designers who just want to
add some of this capability without having to
design, debug, and program for low-volume
quantities. Fortunately, thanks to TIGAL KG
from Austria, we have an alternative that
makes use of Sensory’s technology presenting
it as a speech recognition and audio output
module. The EasyVR 2.0 is designed to add
Speaker out
Internal ROM space
128K Bytes
FIGURE 2
The RSC-4x family features an 8-bit
microcontroller, a 16-bit ADC, a 10bit DAC, a microphone pre-amp, an
independent programmable digital
filter engine, 4.8-KB RAM, five timers,
twin-DMA, a vector math accelerator,
a multiplier, 8-bit data on-chip
storage for SD, SV, Templates, Analog
Comparator Unit, 24 configurable I/O,
and a nested interrupt structure.
4.8K SRAM
Vector accelerator
with twin DMA
COLUMNS
Low battery
detection
32-kHz Oscillator(optional)
65
External memory interface
8-bit data, 20-bit address
cost-effective speech and voice recognition
capabilities to almost any application. The
module includes a set of built-in speaker
independent (SI) commands for ready-torun basic controls. It supports up to 32 userdefined speaker dependent (SD) triggers or
commands, as well as voice passwords. SD
commands can support any language. This
ability is supported via separate module and
development boards that won’t set you back
more than about $60.
VEEAR (VR) MODULE
The EasyVR module in Photo 1 measures
less than 1″ × 2″ inches and is manufactured
using mostly SMT parts. Separate pin headers
on each of its four corners allow it to easily
be piggybacked onto your application PCB.
Two two-pin headers offer support for the
included microphone and the user supplied
PHOTO 1
The EasyVR module takes up little
space and will piggyback snuggly onto
your application PCB using 2-mm stick
headers in all four corners. It has a
voltage input range of 3.3 to 5.5 V and
accepts simple ASC commands over
its TX/RX serial port.
66
COLUMNS
PHOTO 2
The VR Commander application aids the user in adding 32 of
their own speaker dependent commands to its preprogrammed
speaker independent library of 28 commands. Optional software
from Sensory, Inc. can be purchased if speaker-independent
commands are required for your application.
Command
Arguments
ASCII
Byte
Value
1 = 115200
2 = 57600
3 = 38400
6 = 19200
12 = 9600
Hex
‘a’
0x61
Baud
‘b’
0x62
none
‘c’
0x63
Index
‘d’
‘e’
0x64
0x65
Description
Baudrate
0 = trigger
1-15 = generic
16 = password
Index
0 = trigger
1-15 = generic
16 = password
Index
0 = trigger
1-15 = generic
16 = password
Position
0-31
none
‘h’
0x68
Index
0-31
-1 = SI grammar
count
Length
4
8
IndexH
0
0-7
IndexL
0-15
0-31
DelayH
0-15
0-31
DelayL
0-15
0-31
Length
4
8
Level
0-2
TimeoutH
0-15
0-31
TimeoutL
0-15
0-31
Reply
STS_SUCCESS
Break
STS_SUCCESS
STS_INTERR
Get count for SD
STS_COUNT
Activate SD/SV
recognition
STS_RESULT
STS_SIMILAR
STS_TIMEOUT
STS_ERROR
Erase training of SD/SV
command
STS_SUCCESS
Read wave table data
STS_TABLE_SX
STS_OUT_OF_MEM
Read custom and builtin grammar data
STS_GRAMMAR
STS_COUNT
Send a SonicNet token
Index=
(ValueH*32)+ValueL
Delay=
((ValueH*32)+ValueL)
* 28uS
STS_SUCCESS
Receive a SonicNet token
Timeout=
((ValueH*32)+ValueL)
* 28uS
STS_TOKEN
STS_TIMEOUT
‘i’
0x69
WordSet
0-3
Activate SI recognition
from specified word set
STS_SIMILAR
STS_TIMEOUT
STS_ERROR
‘k’
0x6B
Level
0-4
Set SI knob to specified
level
STS_SUCCESS
Language
0 = English
1 = Italian
2 = Japanese
3 = German
4 = Spanish
5 = French
Set SI language
STS_SUCCESS
‘l’
0x6C
TABLE 1
If you use the VR Commander to add and train commands,
you won’t need to use most of the Commands in this list. They
are available if you want to program the module from your
application. I preprogrammed my library using the Commander.
Thus, my application requires only three commands. (continued
on following page)
8-Ω speaker (when sound output is required).
The remaining pin headers are split between a
five-pin control header and a four-pin power/
communication port. The module will run
on 3.3 or 5 V with TTL UART communication
defaulting to 9,600 bps. This header and the
microphone is all that is required to use a
configured module. The EasyVR development
kit PCB has a USB port that allows the module
to be configured and programmed from a
PC running the free EasyVR Commander
configuration software. For you Arduino buffs
out there, there is an alternative development
board, EasyVR Shield.
Photo 2 shows a list in the left-hand
panel of the EasyVR Commander application
that includes potential selections. These are
identified under the description column as
Trigger, Group, Password, SoundTable, and
Wordset.
Note that Wordset 1 is highlighted and the
contents of this group are shown in the right
panel. There are eight (0–7) predefined SI
words in this group. Each Wordset contains
a number of related predefined words. By
segregating words into separate groups, you
can have better control over which words will
be active at any point in time.
The Trigger group includes one predefined
SI word, “robot.” Since the Trigger and
Wordset groups are predefined with specific
speaker-independent vocabularies, a user
circuitcellar.com
SI VS. SD
To use the EasyVR module effectively, you
need to understand the difference between
speaker-independent and speaker dependent
recognition. You may have noticed that people
visiting from outside your area seem to have
a distinct accent, drawl, brogue, or twang
that makes it difficult to understand them at
times even though they are speaking in your
native language. Unless you train your ears
to this difference, it can sometimes seem as
an entirely different language. Our vocal tract
is the cavity where larynx vibrations and air
exhalations are filtered into unique sounds.
The International Phonetic Association (IPA)
has attempted to document these into a
phonetic alphabet.
To communicate, we string together
different combinations of these sound bits
to produce an audible representation of a
particular word. Simply put, we can define
‘m’
‘n’
‘o’
‘p’
0x6D
0x6E
0x6F
0x70
‘q’
0x71
‘r’
0x72
‘s’
‘t’
‘u’
‘v’
‘w’
‘x’
‘y’
0x73
none
0x76
0x77
0x78
0x79
STS_MASK
Label SD/SV command
STS_SUCCESS
Set recognition timeout
STS_SUCCESS
Read SD/SV command
data (label and training)
STS_DATA
Configure, query or
modify general purpose
I/O pins
STS_SUCCESS for
Pinmode 0-1
STS_PIN for
Pinmode 2-4
Index
0 = trigger
1-15 = generic
16 = password
Position
0-31
Length
0-31
Text
Upper Case Letters
Timeout
-1 = default
0 = infinite
1-31 = seconds
Index
0 = trigger
1-15 = generic
16 = password
Position
0-31
Pinnumber
1-3
Pinmode
0 = output low
1 = output high
2 = input
3 = input strong
4 = input weak
Confirmation
‘R’
Reset all commands and
groups
STS_SUCCESS
Mode
0 = character only
1 = whistle or
character
2 = loud sound or
character
3-5 double clap (with
varying sensitivity) or
character
6-8 = triple clap
(with varying
sensitivity) or
character
Go to the specified
power-down mode
STS_SUCCESS
STS_AWAKEN
Index
0 = trigger
1-15 = generic
16 = password
Train specified SD/SV
command
Position
0-31
STS_SUCCESS
STS_RESULT
STS_SIMILAR
STS_TIMEOUT
STS_ERROR
Index
0 = trigger
1-15 = generic
16 = password
Position
0-31
Strictness
1 = easy
2 = default
5 = hard
IndexH
0 -31
IndexL
0-31
Volume
0 = min
15= full
31 = 2X
Constant
-1
Index
0-9 = digits
10 = ‘*’ key
11 = ‘#’
12-15 = ‘A’ to ‘D’
-1 = dial tone
Duration
0-31 in 40ms units
for keys and seconds
for the dial tone
0x74
0x75
Request bit-mask of
non-empty groups
none
Delay
0-10 = ms
11-19 = 20-100 ms
20-28 = 200-1000
ms
Remove SD/SV
command
STS_SUCCESS
Set SD level
STS_SUCCESS
Wave table entry
playback
Index=
(ValueH*32)+ValueL
1-1023 = sound index
0 = built-in “beep”
STS_SUCCESS
STS_ERROR
Play a DTMF key tone or
dial tone
STS_SUCCESS
Request firmware
identification
STS_ID
Set transmit delay
STS_SUCCESS
COLUMNS
can’t alter them.
The SoundTable selection, which I’ll
describe later, is for audio out. Group 1–15
and Password 16 are for you to add and train
up to 32 of your own SD words in any of up to
15 groups (similar to SI Wordsets).
The Password group can be used to hold
special words that require a higher level of
correlation before achieving a match. You
might want to implement this as a sort of
biometric voiceprint for security purposes.
While the EasyVR Commander handles all
the communication with the EasyVR module
for the initial setup, you will need to be
familiar with at a few commands to make use
of the module. Table 1 lists these commands.
Table 2 lists responses.
All data passed between the EasyVR
module and the host (be it the PC or your
microcontroller) is in ASCII readable form.
Commands and Responses begin with a
lowercase letter. All data within Arguments
are also printable characters. All data values
are in the range of –1 through 31. This means
that you must add an offset of 0x41 to each
value to make it printable (see Table 3). Not
only does this enable all the communications
to be seen (as printable characters) on
a terminal, but you can directly enter
commands via a keyboard since there are no
control characters (binary data).
Many applications won’t need the majority
of these commands, particularly when they
have to do with configuration of the module,
including adding speaker-dependent word
groups and training these words. Your
applications might only need to select a group
to listening for and to find out which word
was spoken. Let’s begin by looking at the
difference between SI and SD groups.
67
68
align correctly; otherwise, the least bit of
difference will prevent a perfect match. If the
rules are relaxed somewhat, we might be able
to determine a match even when the two have
slight differences. Further rule relaxation will
make it easier to find a match, but it also
COLUMNS
recognition as dissecting audio into its basic
phonetic components in order to identify
the word it represents. With a large enough
dictionary, we might be able to not only
determine the word, but also the language
represented. This only works when the stars
TABLE 2
For every command issued, the module
responds with status information
(except for infinite timeout used with
the trigger group). Each command list
in Table 1 lists the status responses
expected. This list contains all of the
possible responses.
Response
ASCII
Hex
Data
Name
Byte
Description
Value
‘c’
0x63
STS_COUNT
Count
0-31 = command/grammar count
-1 = 32 commands/grammars
Count of commands or
total number of SI grammars
‘d’
0x64
STS_DATA
Training
-1=empty
1-6 = training count
+8 = SD/SV conflict
+16 = SI conflict
Provide command data
Position
0-31
(only meaningful when trained)
STS_TABLE_SX
Length
0-13
Text
Upper Case Letters
IndexH
0-31
IndexL
0-31
Length
0-31 (Length of Text)
Provide sound table data
Index= (ValueH*32)+ValueL
Text
Upper Case Letters
‘e’
0x65
STS_ERROR
Error
03h = too noisy
04h = spoke too soft
05h = spoke too loud
06h = spoke too soon
07h = too complex
11h = recognition failed
12h = recognition result doubtful
13h = recognition result maybe
14h = invalid SD/SV command
17h = bad pattern durations
4Ah = bad release number
4Eh = bad data in speech file
80h = not found
Signal recognition error
‘f’
0x66
STS_TOKEN
IndexH
0
0-7
IndexL
0-15
0-31
Detected a SonicNet token
Index= (ValueH*32)+ValueL
‘i’
0x69
STS_INTERR
none
‘k’
0x6B
STS_MASK
Mask
Interrupted recognition
‘m’
0x6D
STS_OUT_OF_MEM
none
‘o’
0x6F
STS_STATUS
none
‘p’
0x70
STS_PIN
State
0 = low
1=high
Provide pin input status
‘r’
0x72
STS_RESULT
Position
0-31
Recognized SD/SV command or
Training similar to SD/SV command
‘s’
0x73
STS_SUCESS
Wordindex
0-31
Recognized SI word or Training
similar to SI word
‘t’
0x74
STS_TIMEOUT
none
‘v’
0x76
STS_INVALID
‘w’
0x77
STS_AWAKEN
‘x’
0x78
STS_ID
Version
‘z’
0x7A
STS_GRAMMER
Flags
16 = trigger
0 = command
Count
0-31 Number of commands
Length
0-31 (Length of Text)
Text
Upper Case Letters
4-bit values that form 32-bit
mask
(LSB first)
Mask of non-empty groups
Memory error (no more room for
commands or sound table not
present)
OK or no errors status
Timeout expired
Invalid command or argument
none
Wake-up
(back from power-down mode)
Provide firmware identification
Length and Text repeated for all Commands
Provide custom grammar data
When it
comes to
robotics,
the future
is now!
From home control systems to
animatronic toys to unmanned
rovers, it’s an exciting time to
be a roboticist. Advanced Control Robotics
simplifies the theory and best practices of
advanced robot technologies, making
it ideal reading for beginners and experts
alike. You’ll gain superior knowledge of
embedded design theory by way of
handy code samples, essential schematics,
and valuable design tips.
With this book, you’ll learn about:
• Communication Technologies
• Control Robotics
• Embedded Technology
• Programming Language
• Visual Debugging... and more
Get it today at ccwebshop.com.
COLUMNS
70
TABLE 3
All values between –1 and 31 have an
offset of 65 (decimal) added to them so
they become displayable ASCII. Values
larger than 31 require two 5-bit values
to be sent so the maximum value is
10 bits.
ASCII
@
A
B
...
Z
^
[
\
]
_
`
HEX
0x40
0x 41
0x 42
...
0x5A
0x5B
0x5C
0x5D
0x5E
0x 5F
0x 60
Decimal
64
65
66
…
89
90
91
92
93
94
95
Value
-1
0
1
...
25
26
27
28
29
30
31
TABLE 4
Speaker-independent
vocabularies
are created by first breaking
down each word into its phonetic
makeup. Phonetic recordings allow
each command to be more generic
and more apt to match numerous
speakers.
Speaker-dependent
vocabularies are recorded directly by
a single speaker, making them less
generic and therefore more difficult to
match numerous speakers.
Speaker
Best for
recognition
Training
Independent Any speaker
No training
TABLE 5
The EasyVR module is preprogrammed
with a 28-word SI vocabulary. If your
application can get by using just
these, then you can proceed directly
to integrating the module in to
your application. You can purchase
optional Sensory, Inc. software if your
application requires SI commands.
Dependent
Single speaker Needs training
increases the chance of making an erroneous
match. We must consider the probabilities for
both correct and incorrect recognition.
We have the wonderful ability to tweak
our interpretation of a specific word based
on context of the word within a phase or
sentence. At this point, this is beyond most
recognition algorithms. And so we are left
with some compromise. If general rules are
being used, we can use the term “speaker
independent,” as recognition will require
some relaxation of the rules to take into
account differences in pronunciation, as I
mentioned earlier. However, if we were to
substitute a speaker-specific set in place of the
generalize set for comparison, that speaker’s
pronunciation would better match without
the need for such a large relaxation of the
rules. This is considered speaker-dependent
recognition that might have advantages for
you over speaker independent recognition
(see Table 4).
It is important to note that using speakerdependent recognition does not necessarily
mean it will not recognize the words correctly
when spoken by someone other than the
trainer. This is in a large part based on
both the speaker and the words chosen for
inclusion in any specific group.
Group Offset
Trigger
Word set 1
Word set 2
0
Robot
Word set 3
Action
Left
Zero
1
Move
Right
One
2
Turn
Up
Two
3
Run
Down
Three
4
Look
Forward
Four
5
Attack
Backward
Five
6
Stop
Six
7
Hello
Seven
8
Eight
9
Nine
10
Ten
SELECT THE LIBRARY
The key to optimum recognition is to
select the library carefully. It’s pretty simple.
If two words sound similar, they will most
likely cause recognition issues. Besides this
obvious concern, the framework lessens
some potential shortcomings. By framework,
I mean how the recognition engine works.
Suppose you happen to be engaged in a
conversation in a noisy area. The recognition
engine can’t know the difference between
extraneous noise and you trying to get
its attention. So, it would naturally try to
make sense out of what it hears. At some
point, it will hear something it thinks is a
command and this will initiate an action in
your application. Hopefully, this isn’t a launch
missile command. The framework requires
that a particular trigger word be recognized
before it will look for any further command
input. Additional commands must follow the
trigger word within a timeout period or else
the system will revert back to listening for
the trigger.
We’ve
already
discussed
speakerdependent
and
speaker-independent
recognition. The EasyVR module has 28 SI
words that you can add to your setup. For
the best possible results, organize these
into smaller unique groups of words. For
instance, let’s say your application requires
verbal commands to a robot. A command
will consist of one or more words that may
consist of some action. Depending on the
action, there may be a direction involved and
potentially an amount. Table 5 shows the
SI dictionary contained in a default EasyVR
module. Not only are they available right out
of the box, but you also can select from a
number of languages (e.g., English, Italian,
Japanese, German, Spanish, and French).
Your application would begin by looking
for the trigger word “Robot” (offset 0 in
group trigger). The module will respond
with some status. Unless you get a response
that indicates you have recognition success
of offset 0 (ASCII A), you would continue to
look at the trigger group. This low-level loop
assures no other matching will take place
until after the trigger word Robot has been
recognized. This reduces the possibility of
normal background noise from causing a
false positive. You can then switch to Wordset
1 and look for recognition. Again, status
will indicate success or failure and you can
Join The Elektor Community
Become a GREEN Member Now!
Your GREEN Membership
Comprises:
• 10 Editions of Elektor magazine in digital
format (pdf)
• Direct access to Elektor.LABS; our virtual,
online laboratory
• Direct access to Elektor.MAGAZINE; our
online archive for members
• A minimum of 10% discount on all
products in Elektor.STORE
• Elektor.POST newsletter sent to your email
account each week
• 25 Extra Elektor projects per year (through
Elektor.POST)
• Exclusive GREEN Membership card containing a state-of-the-art Mifare Ultralight
RFID/NFC chip usable with NFC-compatible
smartphones
EXCLUSIVE OFFER
FREE E-BOOK on AVR/Software
Defined Radio with an Elektor
GREEN Membership!*
Order Today at
www.elektor.com/membership
* Available through www.elektor-magazine.com after you have received your magazine download login details.
Take Out Your GREEN Membership Now at www.elektor.com/green-membership
Connect with us!
www.facebook.com/elektorim
www.twitter.com/elektor
COLUMNS
72
PHOTO 3
I wanted to create a PC application that could interface directly with the module’s serial interface. I added
a serial-to-USB interface to translate the module’s TTL serial-to-USB for my input to my PC. This simple
shape-drawing application enables me to change shapes, colors, sizes, and positions from voice commands.
decide what to do next. Perhaps recognition
will indicate recognition of offset 6. You
might want to instruct your robot to halt
all movement upon recognition of the Stop
command.
Note that Robot and Stop are not the same
as Robot Stop. The two single commands
are treated as separate entities. The system
must recognize the word Robot before the
wordset can be changed to look for the word
Stop. However, you may use RobotStop as an
SD word. Let’s take a closer look at SD word
groups.
MY OWN THING
You’re likely thinking, “This is pretty cool,
but this vocabulary doesn’t quite cover my
application.” This is where things begin to get
interesting. As I noted earlier, there is a 32word (or phrase) limit on what you can add.
Liberty Basic
Shoptalk Systems | www.libertybasic.com
SOURCES
EasyVR Module and EasyVR
development kit
TIGAL KG | www.veear.eu
RSC-4x Speech recognition and synthesis
microcontrollers
Sensory, Inc. | www.sensory.com
But that’s sufficient for many applications. I
just covered how SI word groups are set up to
give a flow to what is being recognized.
Let’s consider an example in which I used
the module to provide serial input to a PC (in
this case, as input to a PC application). For
improved recognition, I divided the commands
into a logical grouping. I defined five groups:
trigger, function, color, shape, and value.
The trigger group uses an infinite timeout.
It contains a single word that I can use as
a wake up and begin a round of recognition.
Earlier, I mentioned the word “Robot” as a
predefined SI trigger. But for this example,
“change” is my SD trigger word. If the module
is continually listening, it’s possible to hear
something that might be recognized as a
legal word, and that’s why it’s best to begin
with a trigger group of one word. In fact,
you can put the module to sleep and have it
wake up when instructed or when it hears
a specific sound, whistle, or clap. For this
application, I begin recognition in the trigger
group and switch to group 1 (function) after
a successful recognition of the trigger. The
module always returns status information
when it hears some input. As I noted earlier,
the status depends on the issued command.
For most of my application, I’m issuing a “d”
command for various groups. Refer back to
Table 1 and Table 2. This command has four
possible responses. Unless I receive the “r”
response, I want to indicate an error. I use
the predefined “beep” from the “sound table”
group. This is a sound file that I can play with
the “wAA” command.
The function group has seven words: Color,
Shape, Height, Width, Horizontal, Vertical,
and Clear (see Photo 3). When any word
except Clear is recognized, I change active
group to one of the three groups. This can
be either a group containing colors, shapes,
or values. The first two are straightforward
and require a single word to complete
the recognition cycle and return to group
0, looking for the trigger word. The third
contains numbers 0–9 and an additional word
to exit the routine. Since the value can have
one or more digits, I need a way to indicate
that I’m through entering digits. So, I added
the word “done” to this list. This same list is
used for functions, height, width, horizontal,
and vertical. The value is initialized as zero,
and this number is multiplied by 10 before a
recognized digit is added. By speaking “one,”
“zero,” “zero,” “done,” I create a value of
“100.” Once the word “done” is recognized,
the value is complete and can be assigned to
some function.
Once you’ve decided how your commands
will be grouped, you can create and train
these using the EasyVR Commander
circuitcellar.com
SPEAK ‘N DRAW
Ready for an application? I connected the
EasyVR module directly to a PC. My language
of choice for quick PC applications is Liberty
Basic, which has some defined drawing
commands that enable you to draw a shape
ABOUT THE AUTHOR
Jeff Bachiochi (pronounced BAH-key-AHkey ) h a s b e e n w r i t i n g f o r C i rc u i t C e l lar since 1988. His background includes
product design and manufacturing. You
can reach him at jeff.bachiochi@imagine
thatnow.com or at www.imaginethatnow.com.
on a graphic area. A shape is drawn using an
active color from a predefined palette. Each
shape can be drawn as an outline or filled.
There are height and width dimensions as
well as a starting position on the screen.
I started by writing the application to take
random selections from each of the necessary
parameters and draw the selection to the
display every 100 ms. The selected values are
displayed at the bottom of the screen for each
figure drawn.
Once the application was filling the screen
with random colors and shapes, I added some
individual check boxes to each function that
would halt the random selections for that
function. If a box is unchecked, that function
will remain at its present selection (i.e., every
object “red”).
By adding a COMM routine, I can then
send commands to and receive status from
the EasyVR module connected to the PC. I
added three additional text boxes along the
bottom of the screen. These will display the
command sent, the status received, and a
prompt for the user. Since the protocol uses
only displayable ASCII commands, I can see
the communication between the application
and the EasyVR module.
Any time a function is recognized, the
application will deselect that function. From
that point on, you will need to repeat that
function to change the parameter being
used to display an object. Photo 3 shows
this application once the shape and color
commands have been recognized, as noted
by their checkboxes deselected. The “Clear”
command can be issued to clear the display
and set all of the checkboxes.
START AN APP
You now have sufficient information to
begin using an EasyVR module, or at least to
start thinking about how to use one. Next
month, I’ll explain how to add sound files to
the sound table. This will give you access to
the sound effects, speech, or music files you
would like to include in an application.
Depending on the compression used, you’ll
have room for several minutes’ worth of
sound!
COLUMNS
application. When the application is started,
you must connect to an EasyVR module either
using the development kit or your own serial
interface circuitry (discussed in the manual).
One advantage of the development kit is that
it auto-magically handles the logic to the
*RST and *XM configuration inputs to put the
module in the correct mode (Programming
vs. Run). Once connected, the Commander
application is trading commands with the
module.
Start by adding any words to a group by
selecting the group and adding some text
to identify it using either the edit menu or
the taskbar commands for insert, add, and
remove commands. Once these are set
up, the command window lists them and
indicates that none of the commands have
been trained. Next, select a command and hit
the train command. This will present you with
the opportunity to get ready and prepare an
appropriately quiet environment (e.g., lower
the volume of any background music). You’ll
be prompted to press a key when ready and
speak the word within 5 s. This is repeated a
second time in order to compare the second
pattern to the first. If the two match, the
training for this word is complete. If not, you
can try again.
After all words are trained, you can use the
test group command to see how the module
will react to a spoken word. You should see a
match when a word from the group is spoken.
You can also play around with trying to fool
the recognition system by speaking other
words or changing an inflection or accent. If
the module doesn’t find a match, you’ll get an
error message as opposed to recognition of a
word from the list.
After all of your groups are trained and
tested, your module is ready for use (adding
sound output will be discussed later). As you
add commands to a group and train them,
the module will automatically update. Once
satisfied, you can disconnect and move on to
implementing the module in your application.
I found that the EasyVR development
kit board requires DTR to be high or else
commands aren’t passed to the module. The
COMM function wouldn’t allow me to set the
DTR line, so I was forced to use my application
PCB to test the module with a PC application.
That wasn’t a big deal because I was going
to use it on the PCB. However, you should be
aware of this gotcha.
73
74
CROSSWORD
OCTOBER 2014
The answers will be available at circuitcellar.com/crossword.
1
2
3
4
5
6
7
8
9
10
11
TESTS & CHALLENGES
12
13
14
15
16
17
18
19
EclipseCrossword.com
ACROSS
4. Phase-locked oscillator
8. Single-ended
9. Unwanted capacitance
10. Early capacitor [two words]
12. Conducting fluid
14. One circuit’s signal creates an unwanted effect on another
16. Absorbs sound or electromagnetic wave reflections
17. Used in bipolar power-supply transformers to improve output
voltage symmetry
18. Semiconductor diode [two words]
19. Frequency change that occurs when emitter and receiver move in
unison [two words]
DOWN
1. A resistor that draws the critical amount of load current
2. Detects magnetic anomalies
3. Two frequencies combine to produce new ones
5. Place components directly on PCBs [two words]
6. Garbage or gibberish
7. Measures small voltages
11. Passes current in only one direction
13. Sharp, flexible wire that connects to a semiconductor
crystal’s surface [two words]
15. Cold-cathode-glow discharge diode
circuitcellar.com
75
What’s your EQ? The answers are posted at www.circuitcellar.com/
category/test-your-eq/. You can contact the quizmasters at
eq@circuitcellar.com.
ANSWER 1
An R-C snubber is the series combination of a resistor and a
capacitor that is placed in parallel with a switching element that
controls the power to an inductive load in order to safely absorb the
energy of switching transients.
The problem is that a load that has an inductive component will
produce a brief very high-voltage “spike” when the current through it
is interrupted quickly. This spike can cause semiconductor devices to
break down or even mechanical contacts to arc over, reducing their
lifetime. The snubber absorbs the energy of the spike and dissipates
it as heat, without ever allowing the voltage to rise too high.
ANSWER 2
To pick the resistor value, you first need to know what the
maximum voltage you want to allow is. For example, if you have a
MOSFET that has a drain-to-source breakdown rating of 400 V, you
might choose to limit the snubber voltage to 200 V. Call this V MAX .
Next, you need to know the maximum current that will be flowing
through the load (and the switching element). Call this IMAX . At the
instant the switching element opens, this current will be flowing
through the resistor, and this will determine the initial voltage that
appears across the switching element. Therefore pick the resistance:
R = V MAX /IMAX .
ANSWER 3
Contributed by David Tweed
combine the twe equations and then solve for C:
0.5 V MAX 2C = 0.5 IMAX 2L LOAD
C = (IMAX 2 /V MAX 2)L LOAD
This value will actually be somewhat conservative, because
some of the initial energy of the inductance will be dissipated
in the resistor during the initial transient, before it even gets
to the capacitor. After that, the inductance and the capacitor
will behave as a series-resonant circuit, with the current
oscillating back and forth until all of the energy is gone.
ANSWER 4
When switching DC, the snubber absorbs the energy
stored in the load’s inductance, and after a while, no
current flows and the capacitor is charged to the supply
voltage. However, when switching AC, the snubber has a
finite impedance at the AC frequency, which means that
it “leaks” a certain amount of current even when the
main switching element is open. While this may or may
not cause a problem for the load (usually not), there is
also the issue of the continuous power being dissipated
in the snubber resistor. The resistor must be rated to
withstand this leakage power in addition to the energy of
the switching events.
Contributed by David Tweed
TESTS & CHALLENGES
Picking the capacitor can be more tricky. The key concept is that
you need to pick a capacitor that can absorb the energy stored in
the inductance of the load while keeping its terminal voltage under
VMAX. Since loads don’t often specify their values of inductance, this
may require some experimentation. Let’s call the load inductance
LLOAD. The energy that it stores at the maximum current is: E = 0.5
IMAX 2L LOAD.The energy that a capacitor stores is: E = 0.5 V 2C.
So, if we say that we want the capacitor to store the same energy
that’s in the inductance when its terminal voltage is at V MAX , we can
TEST YOUR EQ
Sign up today and SAVE 50% • Sign up today an
Sign up today and SAVE 50% • Sign
Now offering student
SUBSCRIPTIONS!
When textbooks just aren’t enough, supplement
your study supplies with a subscription to Circuit
Cellar. From programming to soldering, robotics
to Internet and connectivity, Circuit Cellar delivers
the critical analysis you require to thrive and excel
in your electronics engineering courses.
Sign up today and
Sign up today and SA
Sign up today and SAVE 50%
www.circuitcellar.com/subscription
Sign
and
0%
76
CC SHOP
311 Circuits
311 Circuits
311 Circuits – is the twelfth book in Elektor’s celebrated ‘300’
series. An immense source of inspiration for all electronics enthusiasts and professionals, this book deserves a place not far
from the workbench.
311 Circuits
311 Circuits – contains circuits, design ideas, tips and tricks
from all areas of electronics: audio & video, computers & microcontrollers, radio, hobby & modelling, home & garden, power
supplies & batteries, test & measurement, software, not forgetting
a section ‘miscellaneous’ for everything that doesn’t fit in one of
the other categories.
1 CC VAULT
CC Vault is a pocket-sized USB that
comes fully loaded with every issue of
Circuit Cellar magazine! This comprehensive archive provides an unparalleled
Elektor International Media
amount of embedded
hardware and
www.elektor.com
software design tips, schematics, and
source code. CC Vault contains all the
trade secrets you need to become a better, more educated electronics engineer!
311 Circuits – presents complete solutions for numerous problems, as well as starting points for your own creations.
311 Circuits – has been compiled from the 2009, 2010 and
2011 ‘Summer Circuits’ double editions of Elektor magazine.
311 Circuits – is mostly based on readers’ contributions, supplemented by circuits engineered and developed in the Elektor
Labs.
1
ISBN 978-1-907920-08-0
7
Cover_311schaltungen_GB_rz.indd 1
22.09.2011 11:04:25 Uhr
Item #: CCVAULT
4 311 CIRCUITS
2 CC 2013 CD
2
2013 was an exciting year for
electronics engineers! The continued
success of open-source solutions,
Internet of Things (IoT) revolutions,
and green-energy consciousness has
changed the face of embedded design
indefinitely. In Circuit Cellar’s 2013
archive CD, you can find all of these
hot topics and gain insight into how
experts, as well as your peers, are
putting the newest technologies to the
test. You’ll have access to all articles,
schematics, and source code published
from January to December 2013.
Item #: CD-018-CC2013
3 MICROPROCESSOR
DESIGN USING VERILOG HDL
After years of experience, Monte
Dalrymple has compiled his knowledge
of designing embedded architecture
and
microprocessors
into
one
comprehensive guide for electronics
engineers. Microprocessor Design
Using Verilog HDL provides you with
microarchitecture, writing in Verilog,
Verilog HDL review, and coding style
that enables you to depict, simulate,
and synthesize an electronic design on
your own.
3
Author: Monte Dalrymple
Item #: CC-BK-9780963013354
An immense source of inspiration
for all electronics enthusiasts and professionals, 311 Circuits deserves a place
on your bookshelf! This book includes
tips in all areas of electronics: audio
and video, computers and microcontrollers, power supplies and batteries,
test and measurement, and more. The
12th book in Elektor’s celebrated 300
series, it presents complete solutions
for numerous problems and distinct
starting points for your DIY projects.
Author: Elektor
Item #: BK-ELNL-978-1-907920-08-0
circuitcellar.com
CC SHOP
Clemens Valens
Clemens Valens
Stefan Schwark
Mastering Microcontrollers
Mastering Microcontrollers
Stefan Schwark
Helped By Arduino
Android Apps
programming step-by-step
+ Eclipse development platform
+ JAVA object oriented programming
www.elektor.com/android
Mastering
Microcontrollers
Helped By Arduino
Helped By Arduino
Android: programming step-by-step
Clemens
Valens
www.elektor.com/arduino
5
26.11.12 15:38
www.elektor.com/arduino
5 ANDROID APPS:
PROGRAMMING
STEP-BY-STEP
6 MASTERING
MICROCONTROLLERS:
HELPED BY ARDUINO
Many smartphones and tablet computers are powered by an Android OS.
These portable devices’ speed and
computing power enable them to run
applications that would have previously required a desktop PC or custom-designed hardware. Android Apps
introduces you to the programming
required to design apps for Android
devices. Operating the Android system
is explained step-by-step to show how
personal applications can be easily
programmed.
Arduino boards have become hugely
successful. They are simple to use and
inexpensive. Mastering Microcontrollers
will teach you how to program microcontrollers and help you turn theory into
practice using an Arduino programming
environment. Become a master today!
Author: Clemens Valens
Item #: BK-ELNL-978-1-907920-23-3
Author: Stefan Schwark
Item #: BK-ELNL-978-1-907920-15-8
8 LINEAR AUDIO SERIES
Linear Audio is a series of bookzines
full of unique content you won’t find
anywhere else. Each book offers
everything from tutorials to circuit and
system design, to test reports and book
reviews. Why wait? Read, learn, and do
it yourself! See website for all seven
editions.
Author: Jan Didden
8
7
6
7 ADUC841 MICROCONTROLLER DESIGN MANUAL
This book presents a comprehensive
guide to designing and programming
with the Analog Devices ADuC841 microcontroller and other microcontrollers in the 8051 family. It includes a
set of introductory labs that detail how
to use these microcontrollers’ most
standard features, and includes a set
of more advanced labs, many of which
make use of features available only on
the ADuC841 microcontroller.
The more advanced labs include
several projects that introduce you to
ADCs, DACs, and their applications.
Other projects demonstrate some of
the many ways you can use a microcontroller to solve practical problems. The
Keil μVision4 IDE is introduced early on,
and it is used throughout the book. This
book is perfect for a university classroom setting or for independent study.
Author: Shlomo Engelberg
Item #: CC-BK-9780963013347
Further information and ordering
www.cc-webshop.com
CONTACT US:
Circuit Cellar, Inc.
USA
Phone: 860.289.0800
Fax: 860.461.0450
E-mail: custservice@circuitcellar.com
77
78
IDEA BOX
the directory of
PRODUCTS & SERVICES
AD FORMAT:
Advertisers must furnish digital files that meet our specifications
(circuitcellar.com/mediakit).
ALL TEXT AND OTHER ELEMENTS MUST FIT WITHIN A 2” x 3” FORMAT.
E-mail adcopy@circuitcellar.com with your file.
The Vendor Directory at
circuitcellar.com/vendor
is your guide to a variety of
engineering products and
services.
For current rates, deadlines, and more information contact Peter Wostrel
at 978.281.7708 or circuitcellar@smmarketing.us.
$20 for 5PCBs 2 layer,4x4inch, FR4(RoHS),0.063”,1oz, 2LPI,Green,1SK,Lead free HASL Standard PCB: Promotion code: FPGA Boards
from JAPAN
SAVING COST=TIME with
readily
eadily available FPGA boards
ra
Alte
CC14061 PCB & PCBA Small to Mass QTY
INSTANT QUOTE AT:
www.myropcb.com
OR CALL:
1-888-PCB-MYRO
XCM Series
Kintex-7, Spartan-6, Virtex-5
and other many FPGA boards
are available
x
ACM Series
Xilin
Cyclone V, Arria II,
MAX II and other
many FPGA boards
are available
Basic and simple features, single power operation
Over 100 varieties of FPGA/CPLD boards are available
PLCC68 Series
Designed for 68-pin PLCC socket
Very small size (25.3 x 25.3 [mm])
50 I/Os (External clock inputs available)
3.3V single power supply operation
www.hdl.co.jp/CC/
circuitcellar.com
Join The
Save 10%
Online Web-Order Code:
C143009
Become
a member of
Valid thru October 1, 2014
HRXL-MaxSonar®-WR™
- High noise tolerance
- IP67 rated
- 1 mm resolution
- Multi-Sensor operation
- Calibrated beam pattern
- Starting at $109.95
XL-MaxSonar®-EZ™
- Great for UAV’s and robotics
- Incredible noise
immunity
- Small in size
- 1cm resolution
- Automatic calibration
- Starting at $39.95
Circuit Cellar !
Phone: 218-454-0766
Email: sales@maxbotix.com
www.maxbotix.com
TRi
Programmable Logic Controllers
Powerful & Easy Ladder
+BASIC Programming
Ethernet integrated
MODBUS TCP/IP
DI/Os & AI/Os integrated
Project articles and
design applications
Embedded industry
news and announcements
12 issues annually
CC.Post e-newsletter
OEM Prices as low as $119
for full-feature Nano-10 PLC
tel : 1 877 TRI-PLCS
web : www.triplc.com/cci.htm
TRIANGLE
RESEARCH
INTERNATIONAL
microEngineering Labs, Inc.
www.melabs.com
888-316-1753
Programmers for Microchip PIC® Microcontrollers
PC-Tethered USB Model (shown):
•Standalone software
•Command-line operation
•Hide GUI for automated use
•Override configuration with drop-downs
Stand-Alone Field Programmer:
Starting at $79.95
circuitcellar.com/
subscriptions
•Power from target device or adapter
•Program file stored on SD-CARD
•Programming options stored in file
•Single-button operation
Program in-circuit or use adapters for unmounted chips.
Zero-Insertion-Force Adapters available for DIP, SOIC, SSOP, TQFP, and more.
PIC is a registered trademark of Microchip Technology Inc. in the USA and other countries.
79
80
The Future of TemperatureCompensated Crystal Oscillators
TECH THE FUTURE
By John Esterline
M
John Esterline is the CEO of
Esterline Research and Design, LLC,
(www.esterlineresearch.com)
a Pennsylvania based start-up
company. John holds an MEngEE
and a BSEE from Pennsylvania
State University. His research
interests focus on temperature
compensation algorithms for the
improvement of embedded time
bases. John is the inventor on
two US patents (US8188800 B2,
US8525607 B2), and the inventor
of one patent pending (US
13/570,563). Esterline Research
and Design, LLC offers consulting
services in frequency control, test
and automation and other subject
matter in addition to its RF testing
products. You can contact him at
jesterline@esterlineresearch.com.
ost modern digital and analog electronic
devices require a time base to perform their
intended function. Found in everything from cell
phones to smart munitions, quartz crystal oscillators
are widely used in many embedded applications.
Quartz resonators’ high Q, excellent temperature
performance, and superior long-term aging makes
them the clear resonator of choice for many
applications. The frequency versus temperature
performance of a discrete LC oscillator will be on the
order of several hundred parts per million (ppm) per
°C, where a crystal oscillator (XO) will have roughly
±30 ppm over the entire industrial temperature
range (–40 to +85°C). While being superior to a
discrete oscillator, this temperature stability is not
nearly sufficient for many modern applications.
The temperature-compensated crystal oscillator
(TCXO) employs the use of an open loop compensation
circuit to create a correction voltage to reduce the
inherent frequency versus temperature characteristic
of the crystal. The crystals used in TCXOs have
frequency versus temperature characteristics that
approximate a third-order polynomial, as seen in the
nearby figure.
The early designs for TCXOs employed a network
of thermistors and resistors to create a correction
voltage. By using thermistors with different slopes
and properly selecting the fixed value resistors, the
correction voltage can be made to have a shape factor
matched to the crystal’s frequency versus temperature
performance. The correction voltage is applied to a
varactor in the feedback path of the TCXO. This change
in capacitance in the feedback path alters the tuning
of the oscillator, thus changing the output frequency
and compensating it for temperature effects.
Thermistor/Resistor network TCXOs can achieve
frequency versus temperature stabilities of around ±1
ppm over the industrial temperature range; however,
they are limited in their curve-fitting capabilities
because of the nature of using discrete thermistors
and resistors. Thermistor/resistor network TCXOs
are still found in specialized environments including
AT cut frequency versus temperature characteristics
satellite and other space applications where modern
solid-state devices do not have the radiation hardness
to survive. Most TCXOs manufactured today utilize an
ASIC which contains the oscillator circuit and a thirdor fifth-order polynomial voltage generator. The
polynomial generator is an analog output voltage but
also has digital registers for setting the coefficients
of the polynomial. The newest generations of TCXO
ASICs can provide temperature performances of ±0.1
ppm over the industrial temperature range. This is
a 10-fold improvement over what is obtainable with
a traditional thermistor/resistor network TCXOs and
also has the advantage of a much smaller footprint
(5 mm × 3.2 mm).
Some high-precision applications require
frequency versus temperature stabilities better than
±0.1 ppm. To meet these challenging specifications
a different methodology is implemented. An ovencontrolled crystal oscillator (OCXO) uses a heater
circuit and thermal insulation to keep the crystal
at an elevated temperature (≈15°C above the
upper operating temperature limit). By controlling
the crystal’s temperature and keeping it nearly
constant, the frequency deviation due to ambient
temperature changes is vastly reduced. OCXOs can
achieve frequency versus temperature stabilities of
±0.005 ppm. This improved performance comes at
the cost of a larger footprint and increased power
consumption. The TCXO’s performance limit of ±0.1
ppm is due to several factors. First, the resonators
are not perfect. Their frequency versus temperature
stability approximates a third-order polynomial;
however, higher order effects are present. Secondly,
the polynomial generator is nonideal and induces
some higher order artifacts, leaving the user with
residuals of ±0.1 ppm. A new methodology which
uses an artificial neural network (ANN) to create the
correction voltage has recently been demonstrated.
The ANN is superior in that the neural network is not
inherently shape limited like a third-order polynomial.
If enough data is presented to the ANN, it can “learn”
the crystal’s temperature performance shape and
correct for it. This new methodology has been shown
to provide ±0.01 ppm frequency versus temperature
stability over the industrial range. The ANN algorithm
can achieve OCXO temperature performance in a
much smaller footprint, and without the need for the
power-hungry oven.
The evolution of quartz crystal time bases over
the last 70 years has seen the frequency versus
temperature stability improve by a factor of several
thousand. As our need for more stable oscillators in
smaller packages with less power consumption
grows, the development of better compensation
schemes is paramount. The ANN demonstrates a
technology that has much potential. Its ability to
adapt and change its shape factor makes it ideal for
complex compensation problems.
embedded
engineering electronics
discuss
engineering
audio business
networkingmedia
community
system
data
projects
information
product news
talk
social media
mobile
contests
design tips tutorial software
engage
tools
Want to talk to us directly?
Share your interests and opinions!
Check out our New Social Media
Outlets for direct engagement!
circuit cellar / audioxpress / elektor
MULTIPLE
PCBA
LINES
NO MORE WAITING
ACROSS THE GLOBE!
Imagineering utilizes the Jet Printer Technology which has been proven
successful in five continents. This breakthrough innovation allows us to achieve
high-precision solder deposits at speeds of more than one million dots per hour.
High-Speed Non-Contact Jetting Nozzle • Complete Volume Control • 100% Software Driven
More Components, More Possibilities • High-Performance Platform
www.PCBnet.com
(847) 806-0003 • sales@PCBnet.com

SIGNALsolutions - simpson.net.au

Transcription

Similar documents

KIRCHOFF`S VOLTAGE LAW: EXAMPLE 1

Low Cost Pure Sine Wave Solar Inverter Circuit

Thevenin`s and Norton`s Equivalent Circuit

TDA2040 - Micropik

Lecture 20

CJ125 V3.1 Documentation _ENG

TB-CDP02-TX/RX

Experiment 17: Kirchhoff`s Laws for Circuits