Interview with Charles J Gervasi

Featured Engineer

Charles J Gervasi

Charles J Gervasi- PE, Electronics Engineer - Four Lakes Technology, LLC

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

I started experimenting with electronics when I was around 9 years old, playing with children’s electronics kits. Some of the projects involved radio. I got my first commercial VHF/UHF receiver when I was 10 and my first shortwave receiver when I turned 11. Shortwave was really exciting in those days because it was harder to get different points of view. I can still remember an interview I heard on Radio Moscow with a child my age about sports and about his parents wanting him to stay away from kids who misbehave. I had always thought of the Soviets in connection with the cold war. That radio show was the first time I thought of someone my age living in the Soviet Union dealing with the normal problems of kids that age. As I improved my antenna and equipment, I was able to pull in weak stations including two-way communication that I suspected could be associated with military or spy work. I explored the HF spectrum one a few kHz at a time.

I got my amateur radio license when I was 15 years old. This was in 1990, during the peak of the sunspot cycle. A portion of the 10m band had been opened for voice communication to people with licenses that required the ability to receive Morse Code at only 5 words-per-minute. Prior to that, voice operation in all bands that supported long-distance communication required 13 words-per-minute. I made many contacts that sunspot cycle, and I was constantly tweaking my antenna trying to make it a little better.

Many years later, I worked as an engineer designing industrial radios. The work was often very similar to what I did as a hobby in high school. I frequently went on the roofs or hilltops and experimented with different antennas. I tried to understand the relationship between a radio’s performance in the lab versus its performance in a real-world environment.

What are your favorite hardware tools that you use?

The most important tools for hardware work are ones that allow you to rework a circuit easily. That means using precision tweezers and a closed-loop temperature controlled soldering iron with fine and very fine tips. I get a lot of use from my IR / hot-air rework station. The IR part heats the back side of the board to a point below the melting point of solder, so that the hot air gun on the top side melts the solder in the area directly under the hot air gun. This is an easier way to remove a large TSSOP or SOIC. IR and/or hot air are the only way to lift a BGA.

I spend a lot of time using my basic scope and multimeter. I have had good results with the MSO-19 USB scope for decoding serial data. It’s nice when I rent or borrow a nice scope like the Agilent 5000 series because it allows capturing scope readouts to a graphics file that can be easily e-mailed. I still use my old analog scope more, probably because I learned on analog scope and I don’t have to worry about digital artifacts.

What are your favorite software tools that you use?

I have good results with a variety of packet capture software for debugging digital radios.

For embedded design, I have found Keil Tools to be much easier than the integrated design environment tools than the typical free tools available from processor manufacturers.

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

Some of the trickiest bugs involved timing issues in wireless. Things could work well in the lab. The problem only shows up when you add the 5μs per mile propagation delay. There’s no easy way to simulate the delay, so you have to physically separate the radios. If they behave differently, you have to work out what is due to timing and what is due to interference, multipath, etc.

The hardest bugs to catch involve things you wrongly think you know. Once I had a pull-down resistor that was insufficient. I used a 4.75kΩ resistor, but the input was unusual and required less than 1kΩ . I knew that line was pulled down, so I dismissed any explanation that involved that line going high. Instead I came up with elaborate hypotheses such as an unknown problem intermittently and very briefly pulling down a supply or some obscure bug in the chip itself. It was humbling to realize the problem was just a resistor value.

What is on your bookshelf?

I keep all my textbooks because they remind me of what I was doing in classes and projects that I would otherwise forget. For example, even though I don’t care for Sedra’s and Smith’s Microelectronic Circuits book that much, I keep it because it reminds me of a class I took 15 years ago. Just looking at my notes in the margin helps me remember stuff from a time when I was learning the details of transistor circuits.

Here is a partial list of books I particularly enjoyed or refer to often.

The Innovator’s Solution – This is my favorite book on the business of innovation. Even engineers not interested in the business side should read something on the business side to understand the perspective of their clients. This amazing book gave me models to think about things like commoditization and outsourcing/insourcing. Without some models for business, trying to understand business decisions is like trying to intuit how a transistor will behave in an AC circuit without models for bias current and small-signal AC models.
The ARRL Handbook – A nice summary of all electrical engineering with a focus on radio.
The Art of Electronics – Another great summary of all electrical engineering, not as encyclopedic as the ARRL Handbook
C and the 8051 – This is an excellent book for a hardware engineer who does some software work. It explains it like an engineer with a whiteboard, not as a scholarly text. You can use the information for other processors, not just the 8051.
Next Generation Wireless LANs – Explains how the nuts and bolts of 802.11n work, including MIMO, without it seeming like they’re just spoon-feeding you the spec.
Handbook for Sound Engineers – A huge book on audio circuits, acoustics, and audio systems. It goes into great detail and is easy to read.

What has been your favorite project?

My favorite project from a technology standpoint was a software defined radio I was involved in designing. The technology worked, but our design never went to production. My favorite projects from a practical standpoint are those that went to widespread production and led me to positive interactions with sales people, support engineers, and end users.

Do you have any note-worthy engineering experiences?

I was working on a board whose ground was floating at a 280VAC with respect to earth ground. I was tweaking the design by making component changes. I had it very close to working perfectly. I was so focused on the changes that on one attempted component change, I neglected to turn off the power before touching my soldering iron to the board. It caused a loud boom and completely burnt up most of the board. I must have been stunned for a few seconds because it seemed like immediately after the mistake two people were next to be asking me if I was alright. The only injury, though, was to my pride.

What are you currently working on?

I am currently working on sensors and instrumentation boards. One of them has an embedded microcontroller to read ADCs connected to various sensor amplifiers and transmit the values to a computer.

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

Every project has some kind of embedded microcontroller. The tools for embedded microcontrollers are becoming much more user-friendly, so much so that even people who are not engineers or interested in electronics per se can do basic projects on them. I am trying to get fast at working with these tools so I can add processing capabilities without spending much time developing software.

What challenges do you foresee in our industry?

Electronics engineering is one of the best areas to work. The biggest challenge is being able to adapt.

Ten years ago I thought MIMO was just something invented because it was interesting to simulate in MATLAB and therefore topic for people wanting to publish transaction papers. Now MIMO 802.11n transceivers are inexpensive and easy to deploy.

Engineers work in a truly global world because it is so easy to transmit software and drawings long distances.

We have to find a niche where we can solve problems competitively with people in low-wage countries. This means we have to be excited and about getting things working and sometimes we have to work odd hours to have live collaboration with our colleagues around the world.

Engineers cannot wait for a government initiatives to improve our job prospects because what one country does has little impact on the globalized world of engineering.

That is a challenge for those who want a stable job. It’s an opportunity, though, to ignore that actions of people outside our control and focus on getting the projects on our own bench working. As long as we can do that effectively, there is no end to rewarding and lucrative work for engineers.