Interview with Caleb Fulton

Featured Engineer

Caleb Fulton

Caleb Fulton - Graduate Research Assistant, Purdue University

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

I essentially “came of age” over the same time period as the internet did, and I’ve been working on and around computers since I was about eight years old. I actually got my first job working in a small computer shop at age 14. My innate obsession with finding out how things work, increasing boredom with little programming projects, and perhaps even a bit of frustration with hardware (money)-limited PC gaming performance lead me slowly away from computer applications and software into the world of low-level electronics. I built a license plate cover with sequenced LEDs, garnering a bit of derision from my peers. I bought out all of the 2N3055 power transistors from every RadioShack in the greater Indianapolis area to support a high voltage generator based on an old TV’s flyback transformer. I even built a 14-foot potato gun with an electronically-actuated air valve. While my parents were understandably apprehensive about these last two “projects” of mine, they were very supportive when I decided to join my high school’s new FIRST Robotics team during my senior year. Here, I interest in electrical engineering grew through my experience developing low-level hardware-software interfaces to enable autonomous robot navigation. It’s safe to say that I was hooked on EE by the time I began my college career.

What are your favorite hardware tools that you use?

In a pinch, I don’t want to be without an RF “pigtail,” which is just a coax cable with a stripped center pin that’s also had its braided ground pulled to one side and soldered into probe-like stub. It’s essentially a 50-Ohm probe that you can squeeze into tight spaces whenever you have a cascade of several tightly-spaced RF blocks on a circuit board that can’t be easily separated or connectorized. When something is going wrong in a circuit like this (i.e. no signal getting through), the pigtail is useful to see where the signal stops, and, if you’re willing to tolerate the resulting impedance mismatch, you can make very rough “power level” measurements. When we hand-built the transceiver boards for the DAR (digital array radar) project, we could make comparative measurements between the 16 channels at identical points in the RF chains to pinpoint soldering errors or open circuits.

Other than the “pigtail,” I’d have to say that I tend to make good use of our real-time spectrum analyzer for pulsed measurements and our four-channel, high-speed scope for debugging mixed signal and low-frequency analog issues. Moving forward with my research on array calibration, a nice anechoic chamber will be critical; fortunately, I have identified several institutions willing to lend their support for this.

What are your favorite software tools that you use?

I love to use Matlab when I can, but for high frequency simulations I use Ansoft/Ansys HFSS.

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

There have been several vexing bugs in developing the DAR system from scratch, but these were mostly a matter of isolating the cause of the problem in a complex system. Perhaps the trickiest but I’ve solved occurred when I built my first PLL, which was at L-band and intended to be used in a GFSK transmitter. I was trying to optimize the phase noise by switching out loop filter components, and suddenly the phase noise jumped up roughly 10 dB with no apparent cause. After reflowing all of the PLL components (and even replacing a few), the problem still persisted. Finally, when I inspected it under the microscope, I noticed a lot of “no-clean” flux filling the gaps between all of the components. After cleaning up the flux, the phase noise went back down. It turns out that the flux was lowering the Q of the LC resonator in the VCO circuit, making the phase noise go up.

What is on your bookshelf?

I have a whole list of engineering books to read after I graduate: Ben Munk’s metamaterials and frequency-selective surfaces books, Robert Collin’s Field Theory of Guided Waves (which is particularly light reading), and several phased array books. I tend to refer to both R. C. Hansen’s and R. J. Mailloux’s phased array books quite often. For fun, I like to read speculative fiction, a (somewhat pretentious) form of sci-fi.

Do you have any tricks up your sleeve? (special way to analyze circuits, special process you use to make something, etc.)

I have a few soldering tricks that I’ve developed out of necessity, and I’ve gotten fairly good at removing copper from a bare PCB with a razor blade. My main method for solving circuit issues is to identify which of my assumptions is wrong. It almost always comes down to a process of elimination. When I am analyzing a complex system—say, a MIMO radar array—I try to strip away the layers of complexity and add them in only after rationalizing and understanding the simplified system. In the MIMO radar array, for example, I would start with only two elements in a noiseless environment, then I’d add in white Gaussian noise, then more elements, then more bandwidth to the signals, etc. This helps keep me from taking anything for granted because of too much abstraction in the system model.

What has been your favorite project?

The DAR project has, of course, provided a wide variety of interesting problems to solve, from the antenna all the way back to the Gigabit Ethernet link. I think my favorite project, though, would have to be that first PLL I built, since it was the first RF circuit that I took from simulation and layout to fabrication and testing. There is also something strangely satisfying about a phase-locked loop firing up successfully for the first time.

Do you have any note-worthy engineering experiences? (blowing up things, getting shocked, etc.)

I’ve had good luck since I’ve been at Purdue. I occasionally burn my hand on the soldering iron, smoke an IC, or unintentionally melt something, but no shocks or explosions. The story is a bit different for my earlier projects. The high voltage circuit I built in my basement got me shocked several times, but it was a low enough current because of the turns ratios that I was not injured. The potato gun I built was fairly safe, as it was powered by air pressure and not an explosive gas, but I did learn the hard way not to use old PVC cement for high pressure applications.

What are you currently working on?

We are designing the new digital backend for the DAR project, which will take advantage of the new class of “low-cost” FPGAs to enable much more interesting digital beamforming capabilities.

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

My business, right now, is to graduate on time! After that, I hope to extend the work I’ve done on phased array hardware into real-world applications.

What challenges do you foresee in our industry?

As far as the phased array industry is concerned, I think cost will continue to be a big challenge. These systems can have 10,000, 100,000, even 1,000,000 elements, and, with the cost of traditional T/R modules at around $1000 each, one can easily see the problem. Panelization, higher levels of integration, and better packaging techniques are becoming more and more common to lower this cost, but at the same time the demands on performance and functionality are increasing.

On a more personal note, as an American EE graduate student, I can’t help but notice that the number of American engineering students in who choose to pursue a higher degree is fairly small. I think a challenge, then, at least for country, is to find ways to encourage more young people to pursue advanced engineering degrees so that America can stay technologically competitive.