The first step I took was a search for designs that I found visually appealing. A common thread was that I was drawn to the intricate mechanical designs more than the hidden smooth examples. I decided that stepper motors and some type of micro-controller would be the best way to control the planets with precise individual control. Many would say that the choice of an Arduino was less than ideal considering the mathematics required to calculate Heliocentric Planetary locations, and they would be right. The use of a Raspberry Pi is almost ideal for the application. Powerful, small, and a complete pain for me to use. I have tried and failed to come to terms with programming them. The C++ variant of an Arduino just makes sense to me, and I have used dozens of them over the years for projects. The Due fit the bill. Fairly fast machine, a TON of outputs and I had one in my shop… But I am getting ahead of myself a bit, the very first thing I did was sit down and pickup various pieces of round stock and bearings to get my mind around what ways I could drive 8 planets from stepper motors.
What do I use for planets? 3D print them? Hand paint various sized ball bearings? I found some marbles my kids don’t play with anymore and we sorted them out into suitable planet’s at first. We found some suitable candidates but A quick google search of marble size planets came up with some really good options where one set in particular stood out. There was an extremely detailed earth in a set that was above and beyond what I was expecting to find. I think I ordered the set within 10 minutes of finding them. Once the central stack was complete, I had to start thinking of how to display the finished unit. I wanted to ensure that the orbiting planets would not hit objects and a flat plate would fit the bill. Problem was the size and bare surface was too imposing. A star field with constellations was my wifes brilliant idea while lamenting over the problem at lunch.
DESIGNED AND BUILD BY BRYAN LORD
March 2018 - January 2019.
I came up with the idea of driving everything off jack shafts through a central hub. Originally, I was going to make the entire drive train out of gears driven directly off of the stepper motors, but it gradually evolved into a belt drive system. It took me quite a few months to come up with a system that I liked with the visual items that seemed to be right in my minds eye. The central system came together relatively quickly once I committed to the central stacked bearing design. For the first time that I can recall for a project, I did not have a clear plan on what the Orrery was going to look like at the very end. I typically design something almost to completion using Solidworks before I even make my first few chips on the mill or lathe, my mind just visualizes things differently. I started to build the central stack without much care for the final size. It was a combination of CAD design and on the fly machining. I wanted to keep the overall orrery small, but by the time I finished all 8 planets It was evident that the size was going to become fairly large. Coupled with the concentric circle size of Neptune, I had to walk into the house with my tail between my legs and tell my wife that I thought the finished item would be over 2’ wide! I decided that evening that I did not care how big it was going to be, if it had to live in the basement, so be it. I wanted to complete this project regardless of the outcome.
I also wanted a method of entering dates for the user. I decided to continue with the mechanical feel and build my own date system. The feel turned out to be a great tactile experience of spinning a mechanical device to cue up the date of choice, along with linear selector to select a desired month. Its sort of like a multi-tactile device you would give to a small child with sliders and knobs and buttons.
Up to this point in the build I had gone from concept to working system in only about 4mths but I hit a wall in the summer. I was busy in the summer and fall with family, racing and a once in a lifetime trip to Japan. But I was also stumped on how to finish the model. Would it become a table top device? It would have to be a big table! Or do I find an end table I like and mount it onto it? Seemed like I was copping out. If I was going to building this contraption, I wanted to build it all. I toyed around with a few bases that never really gave me the finished product I desired. Then the thought of rings supporting the model came to mind. It would provide support and keep the focus on the Orrery. I feared a solid base would take away the overall presence of the Orrery itself. The problem was the length of flat bar, and the thickness required to support the weight. It would become a very heavy piece not to mention the nightmare of bending round rings of stout metal flat bar. Wine barrel rings suddenly came to mind. They are pre-made , cheap, “should” be easy to find. Checks most of the boxes I was looking for. Turns out, finding a barrel is not that easy, or cheap. Most of what I could find was pretty expensive especially when I was intending on tearing it apart for some metal strapping. My father found a place near me that was making furniture out of old barrels and was discarding the rings. I found 4 rings roughly the same size. Did you know they are all various sizes and custom for each barrel? I did not. They turned out to be allot thinner than I had originally thought they would be. In the process of fitting them together to make a base it occurred to me that they could be orientated into an orbital sphere. I think the result turned out great. With a center shaft to hide the power for the system and add some much-needed support, the base came together over a weekend.
With the hardware sorted, all that was left was painting and the still unresolved programming. I hate painting. That’s all I am going to say about that. What’s more intriguing is 1’s and 0’s! Its amazing what a $30 Arduino can do in the hands of someone with no formal training outside of high school, and a handful of parts from China,. Apparently the limit is the meat behind the keyboard. Pretty astonishing really. But I digress. Throughout the project -almost from day one- I knew that the program to orientate the planets would be tough. Allot of barriers stood in the way for me. Understanding the orbital mathematics is about as far outside of my wheelhouse as working on the JPL lander. I tried to bash together a formula based on orbital periods of planets from an orientation. Utter failure is the best way to describe this attempt. Next up, lookup tables! Making tables for each planet consisting of 5,000 data points was as appealing as dental surgery… After a few nights searching I stumbled upon some papers written by a Swede Along with Kepler’s formulas. Now trigonometry is not my strong suit, so tackling the orbital math was quite daunting. Dozens of hours were spent trying to get the formulas to work in excel before converting into C++. Thankfully the blind luck of selecting the Due made this viable. The Due is unique in the Arduino family for having exceptional extended math capabilities. Thankfully programming fell into place shortly afterwards. 10 months of on and off fabrication and design has turned into an educational and visually appealing piece that I hope sparks many years of enjoyment.
The idea of an Orrery has always intrigued me. They are typically a mechanically geared device that simulates the orbital periods of our solar system. There are some fine examples that you can find photos of online that are magnificent in their workmanship and materials. But I wanted to make one that could show me what the orientation of planets were on any given day, with the ability to present a static display or show the planets orbiting at the correct rates. Of course, an Orrery is anything but accurate in terms of, well anything really. Planets don’t orbit the Sun in concentric circles, or maintain a constant speed (Relative to the earth as they are elliptical in orbit around the sun). About the only thing that is similar is that they are all orbiting on the same plane. Probably the greatest difference is the spacing, if you were to accurately scale up the orbital periods from the marble sized planets I am using, Neptune would be 12km away!
Why is there no Pluto you may ask.... Pluto is inconvenient, so I ignored it.