OldDogBlog post 1.
I am an old dog. A gray beard. A codger. A (fill in the blank) I’m sure my wife and kids would have fun filling in that blank… I graduated from high school in 1980 and started college that fall studying what had become a passion of mine since at least 6th grade – architecture. I knew what I wanted to do with my life. I was on a track. I was set.
That lasted exactly one semester when it became clear to me that I had zero talent as an artist. I could do the drafting part of architecture, but the creativity part eluded me. I believed then, and now, that really good architects are also good artists in some way. That side of my brain was woefully inadequate so I transferred to a secondary interest of mine – engineering. I wasn’t much of a tinkerer or hobbyist so in that sense I was different from many in the engineering profession. I thought creating electronic gizmos was cool, but I hadn’t really done much of that as a hobby.
I’ve since learned that engineering is an art form and one that is well suited to how my brain is wired.
I was a semester behind my fellow engineering students so I decided to go to school year round instead of the normal two semesters. This got me through an engineering degree in three years instead of the normal (at that time) four years. It also had a small advantage in taking several classes in the short spring and summer semesters before the annual fall tuition increase.
In December, 1983 I graduated as a freshly minted EE. A few weeks later, just before the new year started, I moved from Michigan to Maryland and started my first real job with the Department of Defense.
I was fortunate to get hired into an organization that built things. It was one of the few places I interviewed that had soldering irons and work benches set up for each engineer in addition to having desks. The lab space and office space were co-located. It seemed an ideal environment to learn a lot so I took the job. I’ve never once regretted that decision in the 30+ years since.
Back then we were doing logic design using discrete 7400 parts on wire wrap boards. It was all hands-on and I loved it. I got to do every aspect of product design – mechanical enclosure design, controls/indicators/GUI design, circuit board design, microcontroller programming in assembly, test and debug, field deployments and, as an added bonus, write the operations and maintenance manuals. 😉
It was the perfect place for me. I knew nothing practical coming out of college. I built a great foundation here. I learned how to make.
Our designs were all drawn with pencil and paper using logic templates for symbols. If we needed a circuit board made, that was handed off to a layout specialist who would hand tape a design. Finished layouts were reviewed on light tables. Light would shine up through the layout plots making it easier to compare schematic to layout. I recall using red pencils to mark each wire on the schematic and the corresponding line on the artwork to ensure accuracy.
PCB layout tools were one of earliest CAD tools I remember. They were expensive and not hugely powerful by today’s standards, but they were huge time savers and layout folks were eager to adopt them.
Prior to FPGAs designers could try to use early precursors called PAL devices — Programmable Array Logic. These were primitive devices with small numbers of logic functions and interconnect available. They had limited uses and were not hugely popular. Designing circuits using a language called PALASM wasn’t simple or intuitive to me and on several occasions I recall looking into them and concluding “my brain doesn’t work that way.” Hats off to those of you who could use them.
When Xilinx first introduced their products I recall that a different design group from mine started to work with them. I talked to them and got some demonstrations and wasn’t overly impressed. The designers would enter a design, or make a change to a design, and then start to compile it using the DOS based tools on their IBM PC computers. If they were lucky they’d find out within 24 hours whether or not the tool could even compile the design.
I was way too impatient for that. Eventually I got a chance to see a demo of the new tools from an upstart company called Altera. They had some interesting parts and a tool called MaxPlus. That was one of the best DOS programs I have ever used. With MaxPlus you generally would know within 5 to 15 minutes whether or not your design was going to fit into the target part. With this tool and their parts it was suddenly a relatively simple process to do many design iterations in a single day. No 24 hour compile times. These parts were starting get useful and I was able to employ them in some of my designs.
By the early 1990s I had started doing custom ASICs for some designs where FPGAs couldn’t quite run as fast as I needed or lacked the density to meet my requirements. By 1992, however, I designed my last ASIC and have since been able to use FPGAs for the work I do. FPGAs went from design curiosity to design go-to. They became an essential part of my design toolbox and were especially useful for the many quick reaction projects I did at that time.
I never looked back.
Both Xilinx and Altera eventually transitioned to the fancy-dancy “Windows” OS when that became stable (your opinions may vary…) and widely adopted. In the intervening years MaxPlus morphed into Quartus and Xilinx Foundation morphed into ISE morphed into Vivado. I’ve used them all. At this point I’m reasonably agnostic regarding which vendor part I use. I still have a slight preference for Quartus but find Vivado to be very usable and powerful.
That brings me to the present. I’ll wrap it up by welcoming you to the OldDogBlog. This is part of a new company I’ve formed called MakerLogic. In upcoming posts I’ll describe why I started MakerLogic and also describe upcoming products and projects and the decision process behind them.