How did you get into electronics/ engineering and when did you start?

In Manassas, Virginia, when I was much shorter than I am now, my dad had a basement full of electronics components. Back then it was tubes, resistors, capacitors, relays. Things like that. Big amp-sucking things like that. I found just a few TO-3 packaged transistors and quite variety of transistors looking as though they could have been the second or third production design ever. He also had books from a CIE home-study electronics course. I didn’t understand much of it at the time, but I paged through it as well as his CRC Handbook of Chemistry and Physics for entertainment. (I want a tally. How many of you have read the HCP for fun?)

Between then and my high school electronics classes, I tore apart or attempted to repair a lot of old transistor radios, tube radios and a few television sets. I started my own endeavors building electromagnets and crystal radios. I moved on to transistors and then digital logic in high school where I won the state VICA electronics competition and built a COSMAC Super Elf computer (which I still have). It came with 256 bytes of static RAM, six 7-segment LEDs for address and data display and a hex keypad. I later added an expansion with 4K, two S100 bus slots and a few other goodies.

If you don’t remember, or weren’t born back then, the Elf was built around the RCA CDP1802 processor. It had 16 16 bit registers, which was very cool. It didn’t have a built in subroutine instruction which was not so cool. An SoS version of the processor was used in the Galileo Jupiter probe.

I started on an EE degree, but at some point dropped out of that path to write software. I’m not sure that was the best move because I’ve always missed hardware but it was kind of a lemming move. Everyone was doing it. Well, not everyone. Actually, probably no one in Kelso, where I lived. But I heard about a lot of people doing it. I did some early business software in PASCAL, C and dBase (II and III). I did projects in FORTRAN and interface projects in assembly as well.

The currents of economy led me further away from direct technical work, but I’ve never quite abandoned it. I’ve managed to keep soldering iron burns on my fingers and memory leaks in my brain.

What are your favorite hardware tools that you use?

I’m pretty much a minimalist where hardware tools are concerned. I don’t have a scope at the moment which limits the types of project I can put together. I also don’t have much of a head for analog, which also limits the type of projects I can deal with. I guess there are a lot more hardware tools on my wish list than there are in my favorites list.

What are your favorite software tools that you use?

I use Eagle CAD and MPLAB – pretty standard stuff in the after-hours set. I’ve also used PCB123 a bit. I’m working on my first Atmel project at the moment so I’ve introduced myself to GCC. The proliferation of inexpensive or free tools like those have done a great deal to revitalize independent innovation. Years ago, an independent inventor had access to virtually the same set of design tools that professionals did: pencil, paper, black layout tape. When CAD finally came in to solid being and chip pin-counts went way up, breaking in with some new electronic invention became a lot more difficult.

Today, with low-cost, high quality CAD and compilers around, the barrier to entry has dropped considerably. Sure, the pro tools are more feature rich and things, but the other ones are good enough that budding entrepreneurs can design and build some very cool stuff.

What is the hardest/trickiest bug you have ever fixed?

Rather than being difficult, many of my projects seem to become somewhat a comedy of errors. They usually work in the end, but getting there is often an adventure. Take my new Geiger counter project. I could have purchased one of Jeff Keyzer’s kits, but it’s open source and I work for Screaming Circuits, an SMT proto assembly house, so what choice did I have? I had to re-lay it out with SMT parts. Shouldn’t have been a problem. I do layout all the time. But it’s the little things that get you.

I have a PIC programmer, but nothing for the AVR and this design uses an AVR. I managed to find a really simple programmer design online requiring just a few wires and resistors. Astoundingly, I couldn’t find a DB25, so I had to run to Fry’s to get one. Hard to believe I didn’t have one in my garage. I probably do, but it’s likely just buried deep. Add in some jumper wires and old solderless protoboard, purchased from RadioShack just shy of a century ago, and I was in business. Almost.

I recently upgraded a few of my computer internals and, with the parallel port programmer ready, realized that I don’t have a parallel port any more. I found an old Celeron motherboard vintage about 2003 buried in the garage. It does have a parallel port and with enough other junk, I had it working with XP installed and ready to run in an hour or so. I installed PonyProg2000 and copied the .hex file over. I did have one more small detour during that last bit. The old motherboard has but two USB ports, both taken by KB and mouse. I unplugged the KB for the flash drive and then copied it all over and programmed my Geiger and swapped back to the KB.

I don’t know if it works though. I haven’t found anything radioactive and the particular Geiger tubes I purchased are apparently the least performing of the available types.

What is on your bookshelf?

Virtually everything Tolkien (I read the Silmarillion back when the first edition was published), A good set of Dilbert and The Far Side (favorite is Seymore Frishberg, Accountant of the Wild Frontier), The Best of OMNI Science Fiction No. 2 (I wish Omni was still around), Expanded Universe by Heinlein, Rocks and Minerals by Richard M. Pearl, Space Time and Spacetime by Lawrance Sklar, 33 Hiking Trails in the Southern Washington Cascades by Don & Roberta Lowe, Great Dialogs of Plato by Plato (who else?), The C Programming Language by Kernighan and Ritchie (an oldie but still a very goody), Six Easy Pieces by Feynman, A Flying Ace Needs a Lot of Root beer by Charles M. Shulz (quafffing root beer with Bill Mauldin), The Gulag Archipelago by Solzhenitsyn (holy mackerel, this is a depressing book), Snow Crash by Neal Stephenson (I thought he tried too hard to be “cyberish”), Webster’s Instant Word Guide (I needed this before the Internet), The Peace War by Vernor Vinge (he’s not a prolific writer, but a very engaging story teller), Tex Johnston Jet-Age Test Pilot by A. M. “Tex” Johnson with Charles Barton, From Quarks to the Cosmos by Leon Lederman and David Schramm, QED The Strange Theory of Light and Matter by Feynmam, “Surley You’re Joking, Mr Feynman” by Feynman, Russian for Beginners by Charles Duff / Dmitri Makaroff, Backpacking One Step At a Time by Harvey Manning, Boeing Trivia by Carl Cleveland, Microcomputer Primer by Mitchel Waite and Michael Pardee, How to Program Microcomputers by William Barden Jr. (those are a couple of really old micro books) , The Franco Prussian War by Michael Howard (good read on a subject otherwise not well covered), The 9/11 Commission Report, The Cosmic Code by Heinz R. Pagels (before Hawking’s books, this is one of the better cosmology overviews for non-cosmologists), The Elements of Grammar by Margaret Shertzer (again, needed pre-Internet), Golden Apples of the Sun by Ray Bradbury (his best anthology, in mho), The Franco Prussian War by Michael Howard (good read on a subject not well covered), The Savage Wars of Peace by Max Boot (US military history without all of the stuff you’ve already heard about), Jonathan Livingston Seagull by Richard Bach,… Oh. Wait. You didn’t mean literally? Never mind.

Do you have any tricks up your sleeve?

I don’t know if I’d call it a trick, but I seem to be obsessed with symmetry and visual organization on the PCBs I design. There’s an old phrase that says that if an airplane looks good, it’ll fly good. I think I subconsciously follow that same rule for layout even though I know it’s not really true in either case.

Not having easy access to a scope, I really have to think things through both during design and layout. That makes a lot of my work almost theoretical in practice, but so far I’ve done okay. I really do need to pick up a scope.

What has been your favorite project?

My favorite would have to be my 60 Amp dual motor driver system. It’s modular with a simple bus connecting a power unit, MCU unit and two drive boards. I went through about ten iterations of the project to teach myself as much as possible about MOSFETs. I’ve got it driving a pair of 850 Watt motors right now. I still need to improve the current limiting. That and build a chassis for something. I’m not very talented in the mechanical arts.

Years ago, I dabbled with Op Amps and built a device for audibly detecting termites and carpenter ants. It was pretty sensitive but I managed to keep the noise level down enough so I could hear the little critters chewing. Very strange sound. I should have recorded it.

Do you have any note-worthy engineering experiences?

Well, there was the time that I went to discharge the HV lead on a TV set and realized that I had left the thing plugged in. Oops. That was dramatic. But that was quite a few years ago so I’m not sure how relevant it is.

What are you currently working on?

A couple of things. I have the fifth incarnation of my PIC-based robot board. It has eight GPIO, each with dedicated power and ground in the common RC servo pin out, I2C, USB, 20MHz crystal, one push button switch, ICSP and a dual 1A motor driver chip.

I’m also near finished laying out a PIC 28 pin eval board. I use 28 pin PICs in most of my projects. Most of them are pretty close in pin-out, but not quite close enough. Going between PIC16F and PIC18F parts, I can usually use a PCB that I already have, sometimes with just a few jumpers. But the dsPIC30s are just too far off. I’ve loaded the board up with jumpers that will allow me to switch between a number of 16,18 and 30s and evaluate RS232, USB, I2C, PWM, etc for the various chips.

My third project is more of just a layout job. I’m taking the Open Source Geiger counter design by Jeff Keyser of Mighty Ohm and have re-laid it out with SMT parts. I also added a few things like a power on LED indicator and a MAX232 type RS232 transceiver. I’ll undoubtedly shorten the battery life with those two mods, but hopefully not too much. If I get it to work as well as his does, I’ll publish all of the design files in some open source type thing.

What direction do you see your business heading in the next few years?

In my day job, at Screaming Circuits, I see a lot more smaller-componentry on the way. Most of the newer functionality chips (Bluetooth, ZigBee), especially in the RF area, don’t and won’t come out in thru-hole packages. Some only seem to come out in QFN or BGA packages. That makes it much more difficult to hand build something using these new technologies. BGAs are showing up in 0.4mm pitch and I’ve heard rumors of 0.3mm on the way. Rules change at those sizes. Escape routing on a big BGA can be difficult enough as it is. Try it at 0.4mm pitch or smaller. Try and find a PCB house that can fabricate a PCB for you with 3mil or small trace and space, not to mention blind & buried vias or filled and plated vias.

What challenges do you foresee in our industry?

To answer this one, I think we can look to the past. In the mid 1970’s entrepreneurial innovation exploded. The barriers to entry were low and everything was new and exciting. Another decade and the technology changed to the point where it took big money to develop a product. Right now, barriers to entry are low and innovation is everywhere at all sizes.

It will be very different in a decade. Components are on the verge of getting so small that home brew will become very difficult again. It’ll take a lot of money to get something going. It may not seem possible, but it will happen. Chip vendors used to make a wide variety of packages – thru hole down to the small SMTs. That rarely happens with new silicon. Costs dictate that one package is developed. The big money is in mobile electronics, so the smallest possible package is chosen. We’ll see pin pitch a fraction of what it is now. There will be direct copper to copper bonding – no solder on the smallest parts.

What’s it all about?

Carburetors man. That’s what life is all about

In this economy, is engineering a good idea?

I’d have to say “yes.” The winds of need shift constantly. Different careers come and go in demand. Today, it looks like a Wall Street career is the way to go, but in a few years, it might be teaching. A few years after that, engineering. If you love the thought of engineering, do it. Don’t be discouraged by the now.

It’s not a good idea to base such a life-critical decision on what is essentially fad mentality. To be an engineer is to create or to refine. Society will always need that. Us as individuals need a creative outlet. It keeps the brain fresh. And, creativity is not reserved for artists. Engineering is very demanding of creativity.

A few years ago, I heard a lot of predictions that software as a career in this country was dead. It was all going to be sent off shore. A lot was, but try to hire a good software developer now. The need is higher than the supply and salaries are very good. That being said, there probably are some careers without much hope of coming back. Like whaling, for example.

What would be a better career choice, digital or analog? What about software?

The lines have blurred. Most of the opportunities do not lie in pure digital or pure analog. Think mixed signal. Anything embedded likely has some analog. Clock speeds are getting high enough that it may as well be analog. To keep employable, you’d be best to get familiar with both. You can focus on one or the other, but don’t go exclusive. Throw in some software too. It helps to be able to put some code into your microcontroller or write up some python test scripts. Again, the most employable engineers will likely have a specialty, but will also be proficient in digital, analog and software.