Back in the early 1970's I was studying electronics at Cal Poly Pomona and working in the environmental test lab for a Southern California based instrumentation company. The company produced pressure, position, and acceleration transducers for both aerospace and industrial applications. A lot of our products went into the space program, and some of them even ended up on the moon.
The challenge facing the test lab was to devise test methodology to prove to our customer's satisfaction that the transducers actually would perform to their specifications. A lot of the testing was fairly run of the mill, but sometimes we would run into major headaches. The customer engineers that wrote the specifications - often NASA scientists - found it easy to crank out the technical requirements without giving much thought to how the end products might actually be tested.
For example, for one space project we designed an angular position transducer that measured the angle of a missile fin. The product specs called for reliable operation under a number of tough environmental conditions - shock, vibration, salt spray, high and low temperature, temperature cycling, vacuum, and a wide range of humidity. Developing the test plan was pretty straight forward until we tackled the low humidity requirement. The spec called for testing all the way down to 2% relative humidity, and there was no commercially available equipment that could cycle to that level at the time. For a while it looked like we would have to go back to the customer and ask for permission to modify that particular parameter, but our sales group was dead set against it. They knew that any request for spec modification might throw a monkey wrench into the works, and give our competitors a chance to try and overturn the contract placement.
It took a lot of research (this was pre-internet by a couple of decades), calculation, prototyping, and groundwork with the customer, but in the end we managed to successfully pull off the testing without modifying the spec. Our approach was extremely simple. Since the humidity section in the spec didn't specify the operating temperature (there were other, separate sections that did), we devised a test method that would drive the relative humidity up and down by varying the temperature in a sealed volume. Of course the pressure inside the volume also went up and down, but we relieved that using a balloon type bladder. The design passed the test the first time around and everyone was extremely happy.
Designing reliable test jigs and cycling equipment also required a lot of creativity. At times it also required a lot of persuasion and sometimes heated argument with the older engineers that were set in their ways. For timing type jigs they loved to use synchronous motors equipped with step down gear boxes. They would fashion cams out of aluminum, then jury rig switches with cam followers to toggle power circuits on and off. Some of their contraptions were real works of art. I really wish that I had taken photos to document their creations.
But, by the early 70's a lot of the transducers we designed were starting to include electronics. At first it was just some simple circuitry - a few transistors and components. Very quickly the component complexity started to grow, and before long ICs were starting to appear in the newer designs. The angular position transducer I mentioned before included a 709 operational amplifier chip. And as the device complexity ramped the design engineers started to wonder why the test lab wasn't on the program. They felt that they were designing state of the art equipment (which they actually were), but the test lab was using thirty year old methodology to qualify their designs. Frankly, they didn't seem to have a problem with it as long as their designs passed the testing. It was only when they failed that the finger pointing started. And, like all engineers, they had a lot of personal pride and their egos wrapped up in their designs. Their 'babies' couldn't possibly fail, so the fault must be with the test equipment.
In order to preempt their claims, we decided to design a new test fixture based on semiconductor technology instead of motors and cams. It took several months, and we were beset by the normal purchasing approvals, component deliveries, and false start delays. But in fairly short order, and with a minimum of hiccups, we rolled out the new timing test fixture and used it to run a full product qualification test. The fixture centered around a number of 555 timer IC chips. For a 1970's design they were surprisingly robust and tolerant of miss-treatment. It was actually fun to design with them. I ended up using them in a number of other designs, both at the office and at home for hobby type projects. Like most engineers, once I latched on to a solution and found that it worked well, I immediately tried to apply it to every other problem that came up. Sometimes it worked extremely well, and at other times it, or more correctly my designs, met with abject failure - sometimes humorously. When they failed I'd have to scramble around and quickly come up with some other non-555 approach. Over time I moved on to other designs, other ICs, and eventually out of the circuit design business altogether.
That was 30 years ago. Then over the weekend I was doing some research on the Robosapien robot and happened across some articles detailing Robosapien hacks - or modifications. Amazingly enough, there, right in the middle of a couple of the designs, was my old friend the 555 timer chip. It's changed a little over the years. Now it's CMOS, and the price is just a fraction of the price when it was originally introduced back in the 70's. But the basic circuit and chip characteristics are exactly the same as they were 30 years ago. I have no idea who designed the original 555 chip, but whoever it was, my hat is certainly off to them. Brilliant, absolutely brilliant.
"The 555 timer IC was first introduced around 1971 by the Signetics Corporation as the SE555/NE555 and was called "The IC Time Machine" and was also the very first and only commercial timer ic available. It provided circuit designers and hobby tinkerers with a relatively cheap, stable, and user-friendly integrated circuit for both monostable and astable applications. Since this device was first made commercially available, a myrad of novel and unique circuits have been developed and presented in several trade, professional, and hobby publications. The past ten years some manufacturers stopped making these timers because of competition or other reasons. Yet other companies, like NTE (a subdivision of Philips) picked up where some left off."