Building the Big Dog

By Gary Striker

It was the 1970s and I was a lowly engineer at a weapons research lab where we designed military airborne and ground-based defense systems – missile, gun and avionics control systems. Many now-famous individuals, such as computer guru Peter Norton, came to my workplace in Massachusetts with one objective: give engineers micro-circuits and development applications so they could take them home and experiment with them. It was like giving a kid the key to the candy store.

In 1977, the newly released 8080A processor was $350 per copy without the support devices. I concentrated my efforts on a specific processor that I was very familiar with. In the early 70s while still an undergraduate student, I participated in the architectural development of various large-scale gate-array devices (early integrated circuits) that set minimum design requirements for the Intel 8080 series processors. This processor spawned out of the 8008 and the 4004 microprocessor, which was a collaboration of bit-sliced resistor-transistor and diode-transistor circuit design logic used in various military applications; very fast, but not totally applicable to microprocessor clocking circuit requirements. The keen interest in military uses fueled the development of these and subsequent devices.

I never fully got the 70s technology out of my system because in 2009 I designed and built a microcomputer-based system for myself based on that same 8080A chip and named it the "Big Dog." This started as a large-scale design objective...five years later, Phase I (the microcomputer system) is complete with Phase II (the radio communication system) in progress. The project continues to evolve and challenges have appeared where I had to invent creative solutions to keep the Big Dog growing.

For example, I stripped a 19" roll-about music rack to mount and wheel around the assembly in process. It became a permanent part of the system. The wheel-around part was extra handy when the Big Dog was up on a workbench during wiring and assembly. In addition, it was a huge help during troubleshooting.

Big Dog's permanent systemOriginally a quick-fix, the black music rack is now part of the Big Dog's permanent system.

Designing a system from nothing might seem like an impossible task, but this is clearly not the case. A structured approach will keep you within acceptable boundaries. Here are my suggestions of things to keep in mind when building your own system.

Design Advice

Before designing a system, set reasonable goals for what it will be able to do. Regardless of how simple the design goals happen to be, they require a concentrated thought process to get it done.

Record everything and don't throw anything away. A designer should be able to think seamlessly between the digital function to be achieved, and the hardware and code that will accomplish the objective. It's important that a multitude of options become apparent at the same time and that they're all saved, at least in your mind. Remember, there's no sense throwing away good criteria at the expense of bad design.

SchematicSchematic of the Big Dog (click to enlarge)

Get it to do what you want. My objective for the Big Dog was to build a flexible controller for home brew receiver/transmitter applications. It needed to control phase locked loops, digital/analog frequency converters and a host of functions to work in the radio frequency (RF) communications environment.

For the Big Dog, the 8080A processor family and support devices were (and still are) popular and available at Jameco, so that's where I began my design process. I decided to keep it simple and stay within memory confines and focused on my major objectives: direct control of the processor, control of every memory location's bit state and logical control of every instruction cycle.

SchematicThe Big Dog's Input/Output Prototyping and Display Circuit Boards

Hardware Advice

Find the intersection of availability and budget. What hardware do you have and need? Do you have the necessary tooling and support hardware? Where are you going to get the hardware you need? Remember, a friend in need is a pain in the butt, so take advantage of all resources, including the junk yard.

The first question I had to answer was how to connect the multitude of wires necessary to make the Big Dog function. Wire wrap was the obvious solution – it supports a very flexible layout process, is very reliable and changes can easily be made. Then there were high frequency bandwidth considerations. The Big Dog can process about 667,000 instructions per second, which is terribly slow by today's standards but OK for a viable controller. This slow operational speed tamed the pulse switching bandwidth issues considerably.

The Big Dog's hardwareThe Big Dog's hardware needs a lot of power. Its power supply (above) can light up the whole Eastern Seaboard,
which does an excellent job of taming unwanted parasitics.

Development Advice

It may get out of hand. I'll be honest, my development process got slightly out of hand. Interfacing to an outboard device was necessary to provide some sort of display feedback in the verification phase. A half-dozen small programs were written and loaded to a 32 kb memory device via an external device programmer. The computer system allocates half of the available 64 kb memory to EEPROM and the rest to RAM. The LED address, data and status indicators provided direct feedback for each instruction.

The Big Dog in the Lab, Phase IIThe Big Dog in the Lab, Phase II

The Big Dog will continue to evolve. The input/output assembly is currently designed to write to and read from 32 separate devices each. To date, an output display board is provided and an ASCII parallel keyboard is successfully interfaced. The entire system is modular including the display and control board, the central processing board, the input/output board, the power supply assembly and the I/O devices.

Final Words of Wisdom

Start small. Record ideas, concepts and block diagrams that lead to prototype circuits. These are bread-boarded temporary test circuits that can be easily set up on one of Jameco's various proto board devices. Remember that you are not under contract to do anything, so have fun dreaming and formulating ideas. Once those small ideas take shape, then let loose!

Need help or suggestions building your own system? Ask the Jameco community for ideas on the community forum. If you want to build your own version of the Big Dog, email [email protected] and ask for my schematics, test code, Jameco part numbers, photos, etc. I'll be glad to provide any part of the circuit and software design data to get you up and running in short order.

If you feel discouraged, stay cool. Research every step of the way, get up-to-speed on technology and just go for it. You will surprise yourself!

What's your big design project? Jameco loves to feature projects in our newsletter. Send us your engineering dream, your failure or any story about that big project you are working on. Send your submission to [email protected].
Gary Striker lives near LaBelle, FL and has been involved in electronics and computer design since the 1970s by way of academia, research institutions, government and private industry. His interests include aviation, music, ham radio and scuba diving. Gary has been a loyal Jameco customer since 1975.