For novel ideas about building embedded systems (both hardware and firmware), join the 27,000+ engineers who subscribe to The Embedded Muse, a free biweekly newsletter. The Muse has no hype and no vendor PR. It takes just a few seconds (just enter your email, which is shared with absolutely no one) to subscribe.
By Jack Ganssle
In 1998 an anti-shoplifting device reset a man's pacemaker. He collapsed, but a nearby nurse's quick application of CPR saved his life.
Recently various sources (http://news.com.com/2100-1041_3-5415719.html) reported that a malfunctioning flat screen TV emitted radiation on the 121.5 MHz band, one reserved for emergency radio beacons. Law enforcement zeroed in on the location of the false distress call and descended on the college student's room. He was ordered to leave the set off.
Old timers will remember AM radio transmissions being obliterated every time a car with a poorly-maintained ignition system drove by.
Last month the FCC deregulated broadband over power lines (BPL), a fascinating technology that promises to put high speed Internet access at every household outlet. Unfortunately the licenses permit operation between 2 and 80 MHz, a huge swatch of spectrum currently occupied by numerous other services. Ham radio and cellular advocates are up in arms, concerned that BPL will generate tremendous amounts of interference. It seems to me that BPL violates the basic tenet of good RF engineering: constrain unwanted radiation using twisted pair or coax. Power lines do flip over each other once or twice per mile, but only to limit the 5 million meter wavelength of 60Hz radiation. At HF frequencies the wires are very effective antennas.
See http://www.ntia.doc.gov/osmhome/allochrt.pdf for a chart that dramatically illustrates the jam-packed spectrum.
The anti-BPL community complains that there's really no need for this technology. DSL and broadband over cable are everywhere today. Here in Baltimore I pretty consistently get 3 MB/sec downloads via Comcast's cable network, for a pretty reasonable (considering just how much I use this service) $70 or so per month.
For ethical reasons we engineers have a responsibility to limit unwanted RF emissions. When electronic products interfere with each other everyone loses. Hospitals, for instance, are radio wastelands where too many pagers, cell phones, and embedded products all emit so much noxious RF that many systems just don't work properly.
The FCC and European CE regulations mandate certain limits for radio interference. Design a product that breaks the rules and you'll face heavy fines. The law requires that an independent lab certifies requires most products.
Long ago I merely scraped through the two electromagnetics courses all EEs were required to take. The teacher was quite proud of the "elegant" del and curl notation of Maxwell's laws. Though the equations sure were pretty to look at they never made much sense to me. But I was planning to be a digital engineer so figured electromagnetics would be as useful to my career as celestial mechanics.
Then I got a job calculating satellite orbits on a 360/95. And computer speeds climbed from 1 MHz to nearly infinity. Today's sub-nanosecond switching speeds means even a 4 MHz microcontroller spews excessive levels of RFI.
As concerned citizens we have a responsibility to conserve the environment - in this case the scarce natural resource of the radio spectrum. We embedded developers are custodians of this precious resource. Follow the FCC and CE rules, of course. But go a step further. Some CPUs have programmable slew rates for outputs. Use the slowest transitions your system will allow. Keep clock rates as low as possible. Use multi-layer PCBs to control emissions. Enter sleep mode rather than an idle loop. Use fiber rather than long copper cables.
We all want the benefits of wireless electronics. Carelessly designed spectrum-polluting embedded systems will make that dream much harder to achieve.
As will poor FCC decisions.