LED Stirling Engine


CAD, Electronics, Programming, Rapid Prototyping

During the 2015 Spring semester, I was fortunate enough to take this class on Machine Design and Manufacturing (MEAM 201) in which the semester long project is to design and manufacture a working Gamma type Stirling engine. Throughout the course, as we were designing our engines, we were introduced to and made very familiar with a multitude of concepts including tolerances, tolerance stacking, tap/clearance holes, engineering drawings, and precision machining. We were given the part drawings for the majority of the parts, but the base plate, bedplate, and connecting rods were up to us to design from scratch.

Seeing the creative engines that had been manufactured by students of previous semesters, I wanted to try something different and incorporate the basic electronics prototyping skills that I learned from mechatronics the previous semester. So after a number of brainstorm sketches and concept drawings later, I decided on creating an engine that had the capability to display basic persistence of vision images via an on-board printed circuit board.

There were a number of challenges that I ran into trying to realize the electronics portion of the project. Firstly, I had no idea how to design a PCB, so I spent a couple hours on YouTube following some of the walkthroughs on how to use EAGLE. After what was a manageable one week learning curve, I was able to model some simple practice boards. Designing the actual circuit, proved to be the most difficult. I wanted the board to have as high of a resolution as possible and opted for the ATtiny85 to power 12 Red LEDs via a multiplex circuit. In addition, I included a reset switch, a voltage regulator, room for two button cells, and a hall effect sensor to trigger an interrupt. And when it came time to actually CAD up the PCB, I was stuck with selecting from thousands of components. After searching far and wide for the cheapest available components, I was able to finally compile a BOM and start placing the component footprints and connecting the traces in EAGLE. After a several weeks of revisions, breadboard testing, consulting with professors, and making sure that I wasn’t going crazy, I sent out the PCB design and got it manufactured.

On the mechanical side, I had to individually machine every part, and as a result became very comfortable with the machine shop. I reckon that I spent around 10-15 hours on average per week in the machine shop for a total of roughly 100 hours. The mounting block, bedplate, baseplate, and piston flange, connecting rods, flywheel, balance weight, and web were machined on manual mills or the ProtoTRAK, while the heatsink, piston, displacer rod fork, crankshaft bearing tube, and air chamber were done on the manual lathes. The tolerances for these parts ranged from +/- 0.005″ to as small as +/- 0.001″

Integrating the electrical parts and mechanical system together presented a couple interesting problems. Since the hall effect sensor required a magnet in order to trigger an interrupt, I had to ensure that the magnet embedded underneath the envelope cleared the reset button that protruded from the surface of the board.

When everything came together, I was able to write a simple Arduino program that gave the POV display a pinwheel effect. I flashed this onto the ATtiny using ArduinoISP and was actually quite stunned when it worked the first time! I tried to then program the board to display words or symbols, and after spending what was probably 3 to 4 sleepless nights, I was only successful in getting the board to display things at very low speeds. I found the underlying issue was that the ATtiny was under powered and incapable of iterating through the code to display anything when the engine is running at 1100 rpm.

So in the end, I stuck with the simple pinwheel-effect code, and finished with an elegantly designed, electro-mechanically integrated, Stirling engine.

Materials Used

6061 Aluminum, Brass, Assorted screws and fasteners, resistors, capacitors, voltage regulator, hall effect sensor, magnets, PCB

Tools Used

SolidWorks, EAGLE, Manual Mill, Lathe, ProtoTRAK CNC mill, various taps, drills, and finishing sandpaper

Date

January - May 2015

Project Duration

5 Months