Engineering Lesson: The Corollary to Murphy's Law

By Bruce McNair

Murphy's Law Everyone knows about Murphy's Law – anything that can go wrong, will. In the early 1970s, I learned about its corollary – Not only will things go wrong in a demonstration as per Murphy's Law, but the degree to which disaster will occur is directly proportional to the management level of the observer.

I was working as a junior engineer for a large defense contractor on my first digital design project. We had been working a few months on a quick reaction project for the White House to build a transportable satellite terminal that would follow the President and allow him to talk over a high quality secure link to other identical units at various strategic places he might need communications.

I had the job of designing two of the dozen plus boards in the device, filling three desktop 19" rack cabinets (transportable by a squad of Marine guards!) With all the TTL logic, we had a 90 amp 5 volt supply operating the digital logic nest. A senior engineer wisely suggested we put a very low resistance 2 watt resistors in series with each board's Vcc lead to act as a fuse, JUST IN CASE ANYTHING WENT WRONG. He obviously had met Mr. Murphy many years earlier. Our debugging prototypes were wire-wrap boards and it turned out, so were our six "production" units built from our tested and working wire-wrap net lists.

Things were right on schedule and going very well, getting from architecture drawings to a fully working prototype in the three months after I started. One of the mechanical engineers working on the project expressed concern that the pins on our wire-wrap boards were subject to damage since they extended 3/4 of an inch behind the board. We had the room between boards, so he decided to add a sheet of plastic to serve as a protector above the pins for the production units.

Lesson One: If you want to jump to a production unit from a prototype, make sure EVERYTHING about the prototype models what the production unit will be.

One Saturday in the middle of the summer, we were ready to do the final demonstration of the working system to our division vice present, the highest ranking person at our facility. All of our test equipment was hooked up and were ready to "fire up" the system. I was sure my 60-DIP signal acquisition and tracking hardware would work perfectly, it would be the first digital hardware to allow everything to start up. I was also sure my similar complexity clock recovery plus MUX/DEMUX hardware would work as well.

We all held our breath, as the VP watched smoke start billowing out of the baseband analog and digital hardware nests. Strangely, only the RF nest was spared but, then, it had no wire-wrap boards in it. As we quickly shut things down, the VP sadly shook his head, turned and walked away.

Lesson Two: If you can, startup the complete system at a lower voltage or at least a current limited supply to avoid any surprises when 90 amps starts flowing.

As it happened, the Augat boards we used had a Vcc plane on one side and a ground plane on the other side of the board. The mechanical engineer had drilled through them in the four corners of all the boards and used steel screws and spacers to hold the plastic shields in place. Connecting the power supply through the screws and across the protective resistor had clearly demonstrated the meaning of I^2*R to all of the EEs on the project in ways we hadn't imagined.

Lesson Three: When a 2 watt resistor is called upon to dissipate 20-30 times its rated power, it can generate beautiful clouds of smoke before it fails. I used this knowledge a few years later as a fellow amateur radio operator debugging a homemade 3500V, 1 amp power supply during lunch time – 1/10th watt resistors are very easy to hide in a stack of 100 watt bleeder resistors!

Fortunately, the power distribution wiring in the nests were unharmed and the only damage was to the protection resistor and our egos. Our mechanical engineer discovered the value of plastic screws. We were able to quickly repair the boards, change the screws and re-run a successful demo to the VP later that afternoon.

If you have an electronics story you'd like to share, please send it to [email protected].