Interview with Anthony Esposito

Featured Engineer

Anthony Esposito

Anthony Esposito - Owner; Avatar Engineering Corporation

Avatar Engineering Corporation, http://www.avatarengineer.com

When did you first get interested in electronics and engineering?

My start was early in grammar school, I was experimenting with diodes and transistors after my oldest brother took me shopping at a surplus electrical store. The discovery of NPN and PNP transistors and (SCR) latching circuits prompted my building of swimming pool alarms and touch controlled lamps, then later building a radio to fit into a cigarette lighter impressed all the kids at school back in the 60’s. Later, when I was twelve, I earned money repairing tape recorders and radios for the local Radio Shack franchise.

What was your first experience as an engineer?

Back in the 1970’s, I began as an engineer at Mattel Toys in California developing those original hand-held digital games such as football, basketball, soccer. All written in assembly code, they used 4 bit NMOS processors with less than 128bytes of memory, a few LEDs, a Piezo speaker, and they ate batteries for breakfast. The most important experience I had was my training in cost engineering. That taught me early on about the nuances of product development by realizing all the cost issues from materials, labor, overheads, as well as the far reaching implications of off shore manufacturing and distribution channels. As a project engineer, the making of a $30 electronic toy was; “how to make this function with only $1.50 in electronics”?

After playing with toys, I was hired as senior engineer at Hamilton Standard Space Division in Connecticut working on the Shuttle Enterprise and the environmental systems. That’s when I got into power conversion.

What are your favorite hardware tools that you use?

My old HP41 calculator and my Extech FLIR camera. Both are very quick to use and extremely convenient. Working in power conversion requires a FLIR camera or you’ll miss something you didn’t calculate for.

What are your favorite software tools that you use?

Spreadsheets and Spice. I also have preferences of CAE and CAD tools (I think I’ve used them all!), but they all have short-comings that are too frustrating to recommend. It seems as though application software engineers have some distorted view of reality so that the software products can never satisfy the user. I think it takes a good hardware engineer to make good software.

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

There were many times I created “Tiger Teams” to recover external projects from failing, but one of the many “bug” fixes I remember was well appreciated.

While employed by APC as senior staff, I was performing as technical liaison to a UPS company in Denmark to transition 5KVA and 10KVA products to Ireland. While in Denmark in early 2000, the technical management revealed they had a design problem that had festered for 8 years. Power devices would fail mysteriously under any load condition. I reviewed their schematics and understood their techniques for floating IGBT gate drive and suspected their method of pulse latching the on-off states was prone to false triggering. After 2 days of testing and probing, I found the sub-microsecond glitch due to improper PCB layout and resolved the design with a simple added jumper connection. The president of the Danish company was anxious to keep me in Denmark and offered me a fellowship position. Unfortunately, the tax situation in Denmark made it too costly for me to bring the family to live there.

What is on your bookshelf?

30+ years of textbooks, a Star Trek type II Phaser, and Dr Who’s sonic screwdriver.

Do you have any tricks up your sleeve?

My trick is not to rely on conventional methods. The best innovation comes from close evaluation of the constraints and assessing their value or lack thereof.

What has been your favorite project?

Helping a colleague from losing an important contract with Philips Medical, I shoe-horned in a 2 sq.in. add-on circuit to an 18KW RF pulse amplifier for MRI applications. It required 0.1dB linearization over a 73dB dynamic range operating air cooled 15C to 85C. I used an 8 pin PIC processor to implement a 7th order polynomial with dynamic correction in the gain control path and squeezed all analog processing with 32 bit floating point math into 2K of memory. That was fun.

Do you have any note-worthy engineering experiences?

In the early 90’s, I co-founded CSS Power, a telecommunication power company that provided power supplies and power distribution systems to major cellphone carriers. When I started, it was with a clean piece of paper.

Within 14 months, I created a catalog of products and setup manufacturing, automated testing, an inventory control system, audited all my vendors, and hit $1million in sales. By the next year, we were at $8 million in revenue. It didn’t last and was sold to Emerson, but all that experience was better than getting 10 PHDs.

What are you currently working on?

We have many active projects, but much of my personal activities have been recently devoted to new power semiconductor physics for a Japanese client. They seem to appreciate my ability to “see” the circuit paths in the “sandwich” and to let them know if the Swiss cheese has too many “holes”. Sorry for the pun.

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

My business is diverse engineering and our ability to take on larger projects is increasing. We have excellent mechanical and electrical capabilities as well as some expert software resources. We don’t advertise to avoid those “kicking the tires” and we find more “word of mouth” business from many industries. More military and heavy industrial projects seem to be visiting us and there are plenty of conventional power conversion projects to replace aging instrumentation equipment. A road map for catalog products is underway and more direct and OEM manufactured items will be introduced.

What is different about your business from other engineering firms?

My business is primarily about product development and being cost effective. Early in my career, I developed expertise in cost engineering and then full financial oversight. In the 1990’s, I developed software which I sold to the SBA (and Harvard Business) for New Product Development (NPD) that provided the most comprehensive analysis in determining viability of a project prior to implementation. Although, I don’t sell this software any longer, I provide the analysis as a service called “NPD Pro” and use it regularly to quote project schedules and costs with great accuracy.

What challenges do you foresee in our industry?

Most challenges these days are fiscal ones. Many vendors aren’t producing components and distributors are not maintaining much inventory which makes procurement difficult and slows progress of new projects. But bigger problems are looming in on the “Green” initiatives. Many markets, especially solar, are dependent on tax credits for the business models to work. I design very high efficiency converters for solar and fuel cell applications, but the cost can be difficult to justify except in extraordinary environments. Those that push for aid from government energy programs may actually be hurting the industry by focusing on cheap “band-aids” rather than letting the industry progress the technology to be competitive. Right now, the pay-back is awful without the aid so when the tax credits go away, there will be a nasty slump back to reality.

What advice do you give to young(er) engineers?

Simple,… if you don’t love what you’re doing, then do something else. Many of us, more successful engineers, didn’t start our careers thinking about making much money. We did it because we desired to learn and solve the puzzles in technology. Since the technologies are changing ever faster, to keep up you really need to have the interest or you’ll fall to the way-side and no self-respecting engineer wants to be “mediocre”. Like the passage in “Alice in Wonderland”; “…to stay in the same place, you must run as fast as you can. To get anywhere, you must run twice as fast.”