- Polo and the Pony Express: The T Acronyms
- Primary Purpose
- Secondary Purposes
- The Trill
- Tethered Rocking Orbiting Tracker (TROT)
- Tethered Haptics Using Morse Pulses (THUMP)
- Systems That Had to Be Created
Not everyone can afford to send spacecraft aboard rockets into space, and the space elevator is still science fiction due to materials that have yet to be invented. CubeSats are making progress, but they are tiny and still cost prohibitive for an individual. Their small surface area limits the amount of solar energy that they can gather unless sophisticated unfolding techniques are used.
High altitude, specifically Height Above Average Terrain (HAAT), is necessary for certain line-of-sight radio transmissions, and if we can get a craft high enough, it can be both an analog of a spacecraft and a teaching tool. Amateur radio operators know, for instance, that even a relatively small and inexpensive 5-watt, 2-meter HT (handie-talkie) can reach the International Space Station. In fact, they have such a radio onboard to communicate back down to us. One of my favorite authors, Arthur C. Clarke, was also an oft-forgotten radio specialist/physicist and helped to invent the communication satellite and inspire the space-elevator. He had a talent for predicting the future and wrote about interesting spaceships with onboard intelligence and docking space pods, all of which have intrigued me for years. While high-altitude platform stations, geostationary or moored balloons are intriguing, these types of aerial objects are heavily regulated in my country, especially where I live. Like lighthouses that warn ship captains of rocky coasts, it is understandable that our aircraft pilots need to know where aerial hazards are at all times.
I grew up and still reside near the main airport in St. Louis, the old Lambert Field where Charles Lindbergh was one stationed, and he once lived just down the street from where I am today. My grandfather loved airplanes and frequently recounted the regret he felt when he had to burn a captured Japanese Zero in WWII, and I went to the airport with him often to watch the passenger planes land and meet my father, as he traveled a lot. While it was convenient for traveling, the thought probably never occurred to them that proximity to an airport placed severe restrictions on what we could actually fly in the air. This was an unfortunate irony as a kid, since I loved aeronautics; my neighborhood was of large percentage McDonnell Douglas aerospace engineers (now Boeing, near the site of where the Mercury space capsules were constructed). One of my classmates even became a NASA astronaut, and I watched in awe as he installed the Dextre robotic arm aboard the ISS. In 2020, he was on the first human flight of a commercial SpaceX Falcon 9, their Crew Dragon Endeavour capsule parachuting into the ocean, the first time NASA astronauts did this since 1975, when I was 5 years old. My parents and grandparents, however, were not engineers and had little math and science background. Three of them were elementary school teachers, masters of the English language and the arts, inspiring limitless possibilities within me if I put my mind to it. My father was not a teacher, and English was not his strength. Being Persian, he was the only one of my elders that had learned algebra as a child, but had forgotten it by the time I encountered it in school. But he was an explorer himself, an artist/designer at heart, and through his creations, I saw him bring to life some of those possibilities. One day, out the blue, he built a kite out of gift-wrapping paper, a style that I had never seen before, and it flew beautifully.
But, while they couldn't help me with my algebra, they provided me with the encouragement, freedom, and tools at an early age so that I could teach myself through language and communication, so that I could gain the knowledge to claw my way up that alien mountainside to see what those mathematicians and scientists saw as they reached the summit ahead of me. Some people stand on the shoulders of giants, but I stared at the backs of their boots. I read books, built my communication systems, wrote about my adventures and imagined fantastical stories, and when the Internet arrived, like Kevin Flynn in Tron: Legacy said, "I got in".
But long before that, I remember a special day in the 1970's, when I was around 6 or 7, before I had even seen my first computer, when my grandfather walked me up to the dim attic of the private school where he taught, and as as the light cast a beam into the dusty, wooden room, the particulates swirling in the air, the first things that emerged were the shadowed shapes of round spheres of different sizes on rods; a painted, polystyrene solar system, probably left over from a student's project in the 1960's. I cannot convey how profound that moment was, a model of some of our mysterious outer universe, here, static and timeless, high up inside a dusty, school attic. There were old science books there too, intended for the higher grades, and one of them showed children connecting light bulbs to dry cell batteries and communicating using paper and string, a note-passing gondola. The idea fascinated me and I built a mammoth version when I got a few years older.
The Sciences are a different world than the Arts, but I found common ground, saw the connections. I bridged the two schools through philosophy and reason and found Universal similarities that I've spent a great deal of my life trying to understand, which is the primary reason I created TrillSat--it is a union of these elements to help me, personally, understand the world, to find its boundaries. Science is fairly good at finding out where our reality's outer boundaries are, but it has no idea what comprises its inner structure, like a hidden, fractal maze. This is the realm of the Arts.
As a product of Spielberg's films like Close Encounters of the Third Kind and E.T. the Extra-Terrestrial, I'm still radiating my childhood wonder and ingenuity. I watched people use patterns of sound and light to communicate with unknown intelligent life forms, watched an extraterrestrial build a radio beacon out of an old record player, circular saw blade, harvesting power from a windy tree connected to a tether. It is fascinating realizing that the arrangement of matter around us can simply be "re-arranged" to tap into new communication channels that transcend our lonely box. Consumer products were the 80's kids' raw materials, and imagination and wonder were all we needed to uncover the magic inside them, the magical box within the box.
We didn't have 3D printers yet--we had plastic, lots of it, but few ways to build from it.
Today, we can "sing into existence" those plastic objects we dreamt about as the stepper motors vibrate while playing their G-code notes, creating those tiny boxes and hexagons to build the infrastructure needed to bring Clarke's communication satellite to life using inexpensive parts and consumer electronics. Just like E.T., no rockets nor carbon-nanotubes needed. We have nylon paracord, inexpensive microcontrollers that sip nanoamps and have enough speed to drive motors under PWM control. We have cordless screwdrivers, rare earth magnets, tiny BLDC motors and MEMS sensors. We have cheap, futuristic materials in our very hands. We live in a wonderful time of technological enlightenment, perhaps even a Golden Age, but may not realize this for many years. And like the wonder that I had climbing trees and flying kites as a child in the green fields under the vast, blue skies of the American Midwest, we can re-experience the same wonder again using Trees, Tethers, and tiny Spacecraft.
I've been working on my TrillSat project since 2016 and decided it was worth the effort to publish it early and enter it into the the Hackaday 2018 challenges. I have been documenting the project in greater detail, along with my philosophical commentary, on my website at http://greatfractal.com/TrillSat.html and the information on this page is more condensed for the purposes of that challenge. The craft, in addition to the open hardware design challenge, also has subsystems that qualified for 3 of the other challenges (robotics module, power harvesting, and human computer interface) so I clarified those specific subsystems and sub-challenges/solutions in more detail.
Polo and the Pony Express: The T Acronyms
This project uses tethers (ropes), and ropes date back to pre-historic times, so there are a lot of ancient analogies, both nautical and equestrian, that can be used to describe it, and there are a lot of T acronyms that match the satellite-like "T-shape" of the craft itself. The letter T is integral throughout, for the "Tether" hangs from a "Tree", mimics a "Transcontinental Telegraph Transmission" to send "Texts"; even a telegraph pole or tree looks pictographically like a T. T is so common in English usage, in fact, that its Morse code representation is simply a single dash (-), or "dah" (something I took advantage of in my NEAT directional UI for a Morse-code based roguelike game in 2015).
TrillSat or TRILLSAT (Tethered Radio-Interfaced Laterally-Leveled Spiral Axis Tracker) is the name of the entire project, but TRILLSAT-1 is my first prototype that was entered into the Hackaday 2018 open hardware design challenge. The name is an intentional variation of "TSAT", a particular type of single-axis solar tracker that it improves upon (Tilted axis), as opposed to a VSAT (Vertical axis) or HSAT (Horizontal axis). It relies on the TROT subsystem and allows testing the THUMP subsystem (although THUMP is not required for TRILLSAT-1).
TROT (Tethered Rocking Orbiting Tracker) is the name that I gave to the capstan-driven, robotic solar tracker, the distinct platform that both elevates the craft and tilts the panels to harvest energy for the other systems. I entered this system into the 2018 Hackaday robotics module challenge and also entered it into the power harvesting challenge. I called it TROT to represent the two-beat gait of a horse as it "trots" along the tether, like a sideways rocking horse, tilting back and forth with a planetary mass on rod that orbits around, kind of like an ancient polo player swinging a mallet as the horse trots. Note that the T-shape of a polo mallet, or stick, is actually the ideal design for the planetary mass, placing the mass at the far end of the rod to increase panel tilting torque, something I noticed when designing its interior chamber.
It also seems to "swim", like an Olympic freestyle swimmer tilting back and forth in the water, or "buck" like a rider on a mechanical bull with an outstretched arm, but the equestrian analogy had more overall similarities. TROT doesn't include the contents of the hanging Radio Box, and one could put whatever they want into that box, really, but I used it to house the CPU and radio systems for TRILLSAT-1.
THUMP (#Tethered Haptics Using Morse Pulses) is an experimental backup haptic Morse code system that is intended to communicate with the TROT platform when all other higher systems are offline. This is a separate project which I entered into the 2018 human computer interface challenge. Using the same polo rider analogy, you could think of it as the rider using the reigns (tethers) to send signals to the horse by pulling on the bit (mouthpiece). In the case of THUMP, you are thumping on the tether (reigns) with your hands to send signals to the TROT mechanism (horse) via the accelerometer (bit).
This is analogous to "cable thumpers", also called surge wave generators, that have been the names for industrial testing devices for decades. They send a high-voltage electrical wave pulse down the cable to generate an underground thumping sound at the location of a fault (for fault detection), but this system does the reverse--it sends a thumping sound down the cable (a mechanical wave pulse) which then generates an electrical pulse at the location of the craft (for Morse code communication).
So TROT with THUMP forms a skeleton platform like a horse with reigns, with a central saddle with saddlehorn for stabilization (the Gyrofan), and two hanging pannier saddlebags (the Circuit Boxes) that allows basic autonomy and manual control of a single-tether, solar powered robot, but it's not the full system. I then added the rider (the higher-level algorithms), loaded up the saddlebags with mail, so to speak, and turned it into a pack horse.
Like the short-lived Pony Express that once traveled between my state of Missouri and California before the first Morse Code transcontinental telegraph, it then becomes a TRILLSAT, a useful XMPP-to-AX.25 radio communication relay.
Restoring Communication When Our Infrastructure is Down
For the open hardware design challenge, it was an obvious one: this project is literally uplifting (via a capstan-driven hoist), but it specifically addresses the challenge of restoring communication in times of emergency when our entire infrastructure is down for long periods (cell networks, power grid, Internet), and all we have are the Sun and trees. The challenge is also to do this with the lowest cost possible, with easy-to-obtain materials, make it as as small, portable, and lightweight as possible, minimize the number of moving parts, and create a free and open system that can be used and improved upon by anyone (although an amateur radio license would be required if the ham radio transmitter is incorporated).
Hurricane Matthew devastated Haiti while I was programming this project and taking classes to get my second ham license which reminded me what can happen when we lose such infrastructure. I first got my amateur radio license in 2013 and am a tinkerer-type (advancing the art) who bonded more with 80's microcomputer cyberpunk than with the post-war, analog days of radio. So I rarely participate in emergency communication activities, but I greatly value those that volunteer their time to keep up these skills and networks, promoting this important facet of the amateur radio service.
But TrillSat doesn't have to be a ham thing. New consumer radios like the inexpensive ESP32 LoRa boards could be used in place of the 2-meter radio, with some range and bandwidth limitations, but one of the great things about amateur radio is the openness. Under law, we can't encrypt or hide the communications, so the signals teach by their very nature, and it therefore relies on civility and goodwill to keep it in place. TrillSat is currently experimental, not robust nor dependable, and is only a proof-of-concept prototype that I built for myself, with the hope that the knowledge gained may help others in some way.
As long as one has the Sun, trees, and a smartphone (in this case Android AOSP running Xabber with Qi wireless charging and WiFi), in theory, one would be able to use TrillSat to "text" other people with other TrillSats (or packet stations) miles away, leave store-and-forward "e-mail", send real-time messages, or read the status of the distant craft. And one could do this indefinitely, day and night, summer and winter, rain and snow, as long as the hardware holds out. By erecting a TrillSat, it also briefly changes frequency periodically to announce itself (beaconing) on an APRS network even while AX.25 calls are still in progress on a different frequency, allowing other people to know it's out there, what frequency on which it is operating, and other short bits of information. Because Linux has its own AX.25 stack and tools, it can even act as a packet digipeater to route messages to nodes over great distances, utilizing network backbones if needed. There is currently an emergency packet network in my state within reach of me. If you're not familiar with AX.25, it's an old protocol based on the even older X.25, a protocol that fell mostly into disuse once TCP/IP emerged.
If you are within WiFi-range of the craft and indicate that you are available for chat (no WiFi router is needed if one is not available), and this range is fairly significant on an ESP8266, remote radio messages will route to the smartphone in real-time and send an alert (similar to receiving an SMS text). If you're not in range or you don't want to be bothered, the remote caller is notified that you are unavailable and can leave you mail.
But additionally, one can also fully control the craft by texting it commands and tell it to lower and wirelessly "dock" (using straps to hold the phone in place) and re-charge the smartphone when its battery is low. If there isn't enough power available, the craft has various stages of shutdown, including one that draws very low power, which incorporates haptic Morse Code reception by thumping the tether.
The Ancient Library
While it has a primary purpose, that is not the primary reason I built it which is that it is a teaching tool for myself and others.
Before we had our telescopes, the starry, night sky was mostly black and white, a monochrome duality, like the old B&W science shows I watched when I was young. There is a mystery when you de-immerse yourself and see form instead of color, words instead of pictures. It is one reason I study the dichotomy of Interactive Fiction and Roguelike games, describing a world through narration versus depicting a world spatially. I could have built a traditional radio station with a base, feedline, and antenna mast, but I decided to isolate and raise the entire station in the air on a single, thin tether.
The craft is a "bot", an actor, so to speak, an atomic entity, not dependent on a larger system, something I explored when programming other bots, like my home automation robot or the rats or lizards in my solar, microcontroller game. But TrillSat can also parse words, give replies, just like those interactive fiction games--whatever you decide to program it to say...
The intelligence doesn't have to be that sophisticated, since once a bot does something, interacts with the halls and walls, for example, it becomes more than a bot; it has a personality (crude AI) shaped by those walls (context). This duality is profound, and by building something (anything, really) not just reading about it, you share an experience with that bot, you both go on the adventure and become something greater than yourselves. As you solve problems, you encounter solutions of adventurers before you, following in their footsteps, seeing what they saw, and sometimes staring at the backs of their boots. It's a very engaging and memorable way to learn, and sometimes you come up with unique solutions that nobody really considered.
I went one step further, as I am drawn to fractal designs like my father, and aligned the planetary mass and Qi Dock such that, in the future, I can add a Qi receiver coil, ESP32 LoRa board and Lithium-ion battery pack inside the planetary mass itself, instead of using it as a "dumb" mass. This would create a second, independent orbiting bot. The main craft could then be programmed to communicate with it over the same WiFi network, allowing the pod to assume independent control over its orientation where it can then request docking when it gets low on power. I call this future experiment the "Planetary Space Pod", something one could imagine Clarke's Dave Bowman piloting around the Discovery One, and it is in a special position on the craft, being free to rotate outside of the tether, yet a direct electrical connection with the main craft is not possible due to rotating capstan. Instead of using a slip ring, I left this as an exercise to further explore the pod logistics, and it's a perfect way to experiment with LoRa, to see if I can communicate with that tiny pod miles away.
Arthur C. Clarke often wrote about objects or machines with intelligence, such as the mysterious cylindrical Rama, the 2001 Monolith and HAL 9000, which presented themselves as sources of knowledge. This project for me is a source of knowledge, revealing to me the inner structure of the Universe, and I even created a wooden ark storage box symbolic of the one in Spielberg's Raiders of the Lost Ark (or even the wooden crate that encased it at the end of the film), as its contents were also a source of knowledge, a message, and of course TrillSat is a device designed to relay messages. Clarke didn't shy away from combining such ideas into his science-fiction. He bridged the worlds of magic and science, and in the foreward to 2001, he famously stated:
" . . . But please remember: this is only a work of fiction. The truth, as always, will be far stranger."
Notice something about the craft, though: it has many characteristics of a ship at sea. Compare it to an aircraft carrier: a long flat top, a single axis of forward motion, which changes its pitch in waves, which rolls port and starboard. Note that the gondola mechanism under the craft acts like a weighted keel, the planetary mass acts like a sailboat boom or even an anchor when it nears the ground. It uses a rope and capstan winch, common on sea vessels, contains an inertial, anti-rolling, stabilizing gyro, just like ships. It also has the characteristics of an "airship" like the old Zeppelins, which also had underhanging "gondolas", hanging Zepp radio antennas, the use of Morse Code.
When you see the ship analogies instead of the more easily-imagined satellite or spaceship analogies, the tether can actually be seen as a medium, a "tether sea" with its own characteristics, and it brings to mind a much larger and more mysterious world where radio traveled great distances over the sea and was necessary for survival, like the early Morse Code operator on the RMS Titanic that sent for help. Technologies were tested in real-world use with lives on the line. I live in the land-locked center of the US but am lucky to be near the confluence of two giant rivers, with its plentiful source of water and rich farmland, and I can at least ride towboat ferries or watch the river barges pass. Without that tiny bit of experience, fishing in a lake with my grandfather, and taking SCUBA classes in a swimming pool, I'd have no nautical knowledge whatsoever, and nautical concepts kept coming up time after time. Regardless of buoyancy or ship design, even seemingly simple things like rigging and knots are profound. Simple machines are computers too; they have their own algorithms, but are programmed differently. Radio is also a computer, in this case analog, and the "program" is the tuning, discrimination of a signal in the noise, analogous to a digital database search query. Our interaction with static things transforms a denotation into a connotation (like the diegesis/mimesis of the ancient Greeks). As "actors", we bring things alive through our interactions with them.
Then there are also bridge and amusement park analogies: when the panel is suspended mainly at the ends of the tether, it acts like a long bridge with a heavy weight (the gondola) in the center. Note the box girder-like frame on the sides similar to a bridge. And when the planetary mass orbits, it takes on an irregular path as the panel tilts under weight, just like an exciting carnival ride.
So when you build something new, it isn't really new, people have been engineering such things for hundreds or even thousands of years, but you toil away, week after week, in a fog until natural constraints and your mind itself begins to shape the machine into something recognizable, often fractal. Through the process of recognizing your own creation, you begin to reconnect with that lost knowledge, reconnect with the ancient people and civilizations before us.
Creating and building will access a library of information just as much as a perusing a literal library. And the craft itself, with its PBBS is a literal library, serving out information from its menu of choices, just like those old phone-line BBS's. I learned many, many things during this journey, and one of the most profound ones was a greater appreciation of the mysterious relationship between pi and e, those two transcendental, irrational numbers that affected the parametric path of the craft, using trigonometry based on both the unit circle and the hyperbola. Like traversing the halls of a roguelike game, I would never have paid much attention to this if I hadn't chosen to walk inside this ancient library, so to speak.
A Clarkeian Satellite/Spacecraft Analog
And finally, there is the easily-imagined analog, the spaceship or satellite which I chose to entertain, as spaceships are fascinating objects, purely functional machines that often take on geometric forms and pure colors, since the environment is harsh, yet less complex outside of a biosphere, closer to pure physical law. I have a fascination with rotation, and rotation upon rotation. Many science fiction film special-effects teams have built miniature spaceship models out of white plastic, and the unique shapes always make me wonder what strange use they have in that futuristic society. I used inexpensive white PETG primarily for solar reflection (to keep the craft cool), but it brings to mind the craft in 2001: A Space Odyssey. And just like the elongated Discovery One, the elongated TrillSat has its own power source, propulsion, attitude control/telemetry/orbital mechanics, radio systems, onboard ship computers, docking port, and the potential for a space pod. And it really does serve the functions of Clarke's geostationary "communication relay" and brings to mind the benefits of his space elevator.
- It is a geosynchronous communication satellite, but just at "low" orbit.
- It is a self-contained spacecraft that orbits the Sun, but just within earth's atmosphere, and it has a tiny orbiting, self-contained space pod that can "dock" with it.
- It is a planetary probe that has to withstand the elements, full of sensors, but the planet it is probing is earth.
- It is a space-elevator, but the tether is attached to a tree, not extending out of our atmosphere.
Of course, a real space-elevator would crush the capstan in an instant due to the immense tension of the tether, so it would have to have a climbing mechanism instead, but it provides a fascinating model for experimenting with spaceship logistics, systems concepts, kinematics and orbital mechanics. Again, no expensive CubeSats needed.
Clarke, near the end of his life, studied fractal geometry, which has been an interest of mine for several decades, and like many machines, the craft is indeed a collection of numerous self-similar elements (too many to list here), but the most notable two are:
- The tether that supports the craft to elevate it, at a smaller scale, supports the gondola to hold it against the solar panel frame, at a smaller scale, supports the pulley frame and end cap against the motor housing, and at the smallest scale, supports the capstan against the motor axis. And if the test frame is used, a wire rope (steel cable) supports the entire setup. Tether sections supporting tether sections.
- Planet earth orbits the sun, changing its orientation under gravity, which provides wireless power (solar radiation) to the craft, and the planetary mass orbits the craft, changing its orientation under gravity, which provides wireless power (induction) to the planetary space pod. And inside the motor housing that drives the planetary assembly is the screwdriver's own planetary gearbox. Orbits within orbits, wireless power transfer within wireless power transfer.
Notwithstanding the acronym in the title, it's called TrillSat because it was partially inspired by Red-Winged Blackbirds. I used to run along a long path between an natural, oxbow lake next to the wide Missouri River, and as I ran toward the river bridge, I'd pass by a series of those birds perched high on the telephone wires, almost equidistantly spaced apart, sounding their characteristic trill sound as I ran below; a fast alternation between two frequencies, which some have described as "oak-a-leeeeeeee" or "conc-a-reeeeeee". In those distance-running traces I could not help notice that it sounded like AFSK (Audio Frequency Shift Keying) packet radio data transmissions (like an old computer modem), and I begin wondering if they were actual data transmissions, with the first part sounding like a kind of initialization or ID ending with a longer trill which gave me the idea that I could perch the entire packet radio station (power source, computer, radio, antenna) up on a tether, just like those birds on a wire, and communicate in peer-to-peer fashion with other "birds" like itself.
But the trill is more than just audible/electronic, it is designed into the mechanics of the craft and expresses itself in the dualistic, back-and-forth oscillation, the rocking, the cycloidal/sinusoidal mechanism of the craft as it rocks itself using its pendular, planetary mass. A trill is also commonly referred to as a roll, like a roll of the tongue, and the craft rolls port and starboard. Like my distance running, the TROT mechanism "trots" up and down the tether, but when combined with the complete radio system at high elevation, the entire craft is no longer a horse, but a bird, and begins to "trill", becoming a TRILLSAT.
Tethered Rocking Orbiting Tracker (TROT)
TROT is the capstan drive locomotion, planetary assembly, and spiral axis tracking system that forms the basic platform for the TrillSat craft. I entered it into Hackaday 2018 robotics module challenge and also entered it into the power harvesting challenge as it qualified for both, the functions are inseparable in this case. Other than high-above ground (average terrain) the purpose of robotic mechanism is to maximize solar energy using a single drive motor and single tether. The craft oscillates in an alternating binary fashion as it traverses or "trots" up and down a tether, like a zig-zagging rocking horse.
The mechanism is already functioning on the TRILLSAT-1 prototype and basically consists of the solar panel frame aligned longitudinally along a catenary tether with a hanging gondola and planetary assembly.
The CHALLENGE was one that I already encountered when I set out to build TrillSat: How do you elevate an efficient solar tracker for effective line-of-sight radio communication height, using a single tether, 3D-printing most of the parts, while minimizing the cost, weight, power, and the number of moving parts?
I used trees as a source of elevation, with a tether attached high in a tree and then pulled due south and staked into the ground at an angle close to the latitude (in my case 38.6 degrees).
This orientation means that the panel will never be in the shade (in the northern hemisphere) and the default angle is optimal for a Tilted Single Axis Tracker (TSAT). It uses a capstan winch/hoist mechanism instead of spools, takeup-reels, or climbing mechanisms. A capstan is a cylindrical machine that uses friction (calculated through the capstan equation) to hold a flexible line or rope around the cylinder. The honeycomb infill PETG plastic can withstand quite a lot of compression, so it acts almost like a rope climber (where you don't have to worry about tightening or locking take-up spools to keep the tether taut) except that the mechanism is simpler and the rope is interwound around the shaft, making it more difficult to thread and setup. You don't have to worry about a changing spool diameter, weight, torque, or RPM as the craft moves since the number of winds around the capstan remains fixed and it never accumulates excess.
But the capstan has to have enough friction for the amount of winds/diameter. It can't bind or tangle on itself, and it has to allow high motor torque since it takes substantial force to lift a heavy craft and planetary mass, so it requires careful design.
One of my first mistakes in building it was to make the erroneous assumption that capstans are simply a rope around a shaft--there was more to it than that. The shaft has to be symmetrical since the amount of rope entering and exiting the shaft has to remain constant. The shape is relatively hard to 3D-print in one piece due to the overhanging tapering. Flares and guides have to be positioned so the rope doesn't overlap and tangle, and it has to enter and exit different ends of the shaft.
This adds an asymmetrical lever effect to the capstan which increases with tension, but luckily the steel screwdriver mechanism can withstand these forces. Another issue is that, because capstans rely on friction, the friction created by the capstan cannot be less than the friction created in the rope guides for the rest of the unit, or the capstan will slip or require a larger size that may not work in your situation. You can decrease the diameter of the capstan to increase torque and increase its length (allowing more winds) to increase friction, but the stresses on the frame can be large. You can also pull the rope tighter, but again, the stresses on the frame are large and this just amplifies the friction of the other guides, nullifying some of the effect. Since my design uses an off-axis capstan, there were 4 sharp bends in the rope which increased in friction as the rope tightened, so I had to add 4 grooved pulley ball bearings to keep their friction (via a rolling effect instead of a sliding effect) below the capstan friction as the tension increased. Simple screw-eye type bushings did not work. I used the inexpensive 624vv, but the grooves are so shallow in relation to the rope thickness that careful angles and guides were required to keep the rope in place, which barely fit within the 200mm dimensions of the 3D printer print bed. And larger pulleys would be difficult to fit into the ends of the battery frame, since I was already hitting the max print width of the printer bed.
Obviously the ideal capstan would not require 4 changes in direction (and 4 pulleys) but would remain in-line with the rope, but this design does not allow the motor to easily perform its secondary function of rotating a planetary mass in 360 degrees, as the mass and rod has to be free of any obstructions such as the tether itself.
The best design I could come up with was to orient the capstan at the end of the mounting apparatus, removing all obstructions from the planetary orbit. The geared, cordless screwdriver was ideal for this, as it is strong enough to rely on a single-ended shaft and doesn't require a bearing at both ends of the shaft (which would create an obstruction), while also being inexpensive, light, and using relatively low current. While this orientation allows the motor to create unwanted rotation when used, causing the craft to roll around the tether, slow start algorithms reduce this effect, and the orientation has the advantage of allowing the motor itself to perform active sway cancellation caused by wind detected by the accelerometer with respect to gravity.
So as the capstan mechanism traverses the tether, the planetary assembly rolls the entire craft and changes the tilt angle of the panel.
To drive this amount of weight up an angled tether, which turned out to be around 1 lb (458 grams) plus the angled weight of the craft in some cases, the voltage of the screwdriver had to be increased, which increases its current under Ohm's law. But the massive jerking incurred by the high speed, high torque mass will throw the craft around violently unless slow start PWM routines are added. Some of the stress is relieved by a passive ball damper in the planetary mass. I built a fairly powerful H-Bridge from logic-level MOSFETs directly into the handle of the screwdriver and then used the 16-bit hardware timer on the ATtiny 1634 (for phase and frequency correct PWM) to ramp-up the voltage gradually using different PWM routines which include both rheostatic and active braking. These methods were necessary to properly drive the "orbital mechanics" of the mass, as the characteristics change on its orbit like a car with low horsepower driving up a series of hills trying to maintain constant velocity.
The pulley frame is designed so that, when correctly assembled, the tether can only enter in one orientation, and this orientation allows particular "compass" set points via two unipolar Hall-effect switches and two earbud headphone magnets in south polarization, inset into a 3D printed magnet holder attached to the screwdriver shaft. Minimal parts were used, and the sensors and magnets are were arranged such that the sensors are only triggered at the same time at solar noon, providing absolute confirmation of motor position. Solar noon is when the mass is at its high point, 0 degrees, North, as this takes the stress off of the planetary rod. Sunrise, 90 degrees, and sunset, 270 degrees, can be detected relative to this, and no position sensing is done during the night phase (when the mass is passing by the lower side). There are 4 distinct quadrants in 2 directions (8 unique phases) that need to be considered due to the position of the mass in combination with the direction of the craft, but I combine the two night quadrants into a single quadrant, and call these 3 quadrants "tridants", since they are no longer quadrants, and I'm not concerned with monitoring the orbit throughout the night, which would also be difficult since I lack Hall-effect sensing at that position.
Since the daytime is when the panel tilting will occur, the motion of the mass from sunrise in the East toward noon is upward--therefore the craft is oriented to drive downward, allowing the downward weight of the craft to assist with the upward lifting of the mass using the more powerful and power-efficient sign-magnitude PWM routine, but once the mass swings downward to approach sunset, the mass moves more slowly, under the controlled-braking locked-antiphase, to resist the free-fall forces. At night, the planetary mass is reset back to sunrise position to lift the craft back up again and reset the mass at sunrise, waiting for the morning sun. The screwdriver planetary gear, in combination with the the tight tether, has enough resistance to maintain its position in this arrangement. It may slip slightly at sunset, but this is not critical, since night is approaching. It is more important that it holds its position for sunrise, after the depletion of the batteries during the night, which is possible due to the balanced arrangement of the craft and mass.
Once this rough alignment is achieved via the Hall-effect sensors, a more precise alignment can then be achieved by waiting for the craft to stabilize and reading its tilt angle with the accelerometer to within .1 degree, making small adjustments using slow start/stop routines and open-loop timing, and then reading the tilt angle again, analogous to the "finite" and "impulse" burns in a real orbital maneuver.
Since the accelerometer can also detect pitch, not just roll (with respect to gravity), it is also used to find the point on the tether curve most appropriate for that time of year. The panel is aligned longitudinally along the rope to provide stability, and this means that it has to tilt (roll) laterally. Like a Zeppelin, this provides a nice shaded area underneath to hang a gondola. When the tether is at an angle based on your latitude, the craft rolls like a Tilted Single Axis Tracker with 97+ percent efficiency. It cannot tilt fully horizontal (to keep weight and rod length to a minimum) but due to Lambert's cosine law and indirect light, the benefit is non-linear and a full tilt is not as necessary. But since the accelerometer can also read pitch, it can travel up and down the tether during different times of year, parking at different points on the catenary, exceeding the efficiency of a TSAT (in theory, not necessarily in practice), not quite a dual-axis tracker, but a spiral one.
I've tried to visualize the parametric path of the various parts of the mechanism since it was difficult to construct the exact parametric equations or create a robotic kinematic simulation.
It's also difficult to get out a ruler and carefully measure and plot the points to see exactly what is occurring. Things that might seem intuitive often do not match reality upon scrutiny. Aristotle's intuition was fooled by the properties of falling bodies (which Galileo disproved), and the "brachistrochrone curve", part of a cycloid, took a bit of effort for brilliant 17th-century mathematicians to isolate (although Isaac Newton is said to have done it in a day).
It wasn't until after the 2018 challenge that I spent months in algebra, trigonometry, inverse elliptic integrals, equilibrium equations, and vector decomposition to extract the equations of motion of TrillSat into around 30 equations which I then applied to a "Delta-v Motor Drive System" to predict the torque curve, which I had to pre-generate on the more powerful Pi ARM processor and then stream to the ATtiny motor driver microcontroller in real-time. It's analogous to a feed-forward Delta-v thruster burn in orbital mechanics, and I had to maximize all aspects of the craft precision to minimize cumulative error in-between sensors, when it is mostly under dead-reckoning. I used the 6 edges of the large Hall-effect switch tolerance zones for precise phase-lock motor correction, a drawback that suddenly became a benefit.
The solar panel should tilt once per side about 1078 times to go all the way up and then back down (a complete period) a 50-foot (15.24 meter) tether (although in practice this will be shorter), and the capstan rotates one revolution about 539 times over that same distance.
From above, looking down on the craft from the sky, excluding the catenary and tilt of the panel, it is clear that the Planetary Mass takes on the path of a trochoid known as an extended, or prolate cycloid, since the capstan could be through of as a rolling circle, rolling along the tether, and the Planetary Rod extends the cycloid.
But during this dance, driving up a conic path, the force of gravity begins to affect the craft differently, for that force is not hit head-on, a single vector, but a partial component that begins to hit the sides of the mechanism as it tilts and orients differently in space, altering the absolute geometry and the subsequent mathematical patterns that describe it.
The axis upon which the panel rotates, due to the kinematic constraints of this mechanism with one degree of freedom, is not a single Cartesian axis, and translation is also occurring, so it could be thought of as its own coordinate system, with a single axis curvilinearly "bent" into a spiral. Imagine if the tether was a long, arcing threaded rod with helical threads. If you were to spin a nut around the "axis", in addition to its obvious rotation, it would also slowly move forward or back, slowly move up the arc and back down. This allows a more accurate tilt angle using one degree of freedom than a pure TSAT, since it can also change the pitch. This is well-suited to the slow, seasonal changes in the earth's tilt angle as it orbits the sun, since since the pitch also move slowly, and the roll moves faster, just like the fast changes in diurnal solar elevation from Sunrise to Sunset.
The TRILLSAT-1 prototype employs the TROT mechanism, but it is a packhorse, not a racehorse, as it was difficult to achieve the proper balance needed to keep the weight to a minimum. The tether needs to be at the center of gravity between the solar panel frame and the gondola assembly to allow the planetary mass to be small to tilt the craft to appreciable angles, but it was difficult to estimate the final weights ahead of time, during 3D printing. The craft and mass were heavier than anticipated at 8.28 lb (3.76 kg) causing it to hang lower than necessary (the weight suspended between 2-ropes problem) and the tether was tighter than expected, adding unnecessary stress to the drive motor and frame.
But the force of the tether that lifts the craft, has several beneficial effects: In addition to providing more capstan friction to assist the drive system, it also holds the gondola and electrical terminals in place against the panel frame, it holds the pulley frame and endcap against the motor housing, and it holds the entire planetary assembly, the capstan and heavy mass, inside the screwdriver shaft.
It is a fascinating design with several dynamic elements that needs to be further explored. In TRILLSAT-1, I also added a gyroscope to the gondola as an experiment to see if it could act as a reaction wheel along the same axis and provide internal stabilization. Due to the balance issues I mentioned earlier, the gyro speed and weight need to be exceedingly high to become useful and the gyro mass itself causes a nasty feedback effect which upsets the balance of the craft if not accounted for earlier. Once all elements are in balance on the axis, however, the craft should operate well in theory.
In TRILLSAT-1, I added several other elements to the TROT platform besides the gyro: Qi docking with an option space pod, circuit boxes for radio and power regulation, etc, but in a pure TROT platform the radio box can carry other sorts of computers, electronics, or radio systems, it doesn't have to carry the Pi Zero W, ESP8266 and packet radio system.
Animation of the TROT Mechanism
An animation of the TROT mechanism in HTML5 WebM VP9 format is shown below (The motion is not to scale and is exaggerated for clarity):
But if you can't see the animation above, a traditional animated GIF can be downloaded by clicking here, although it is 88% larger and contains 718 frames, at 9.33 MB.
Tethered Haptics Using Morse Pulses (THUMP)
In early 2017, I published an algorithm called SIMTHEO (Simple International Morse THEOry) to efficiently decode hand-sent International Morse Code (discrete letters only) on an ATtiny 1634 microcontroller. I created it to both improve my roguelike game recognition but also so that I could adapt it to the TrillSat tethered robotic craft, which uses some of the same microcontrollers.
My original idea was to add the SIMTHEO algorithm to the ATtiny called Sawyer (which is used for motor control) on the TRILLSAT-1 prototype which maintains a connection to two CdS LDR sensors. This algorithm works with "normal" Morse Code when a continuous signal is varied in length and could therefore be applied to light communication (like a flashlight being pulsed on those LDR sensors) with the craft hanging high in the air.
Later, I realized that the other ATtiny 1634 (Huckleberry, the power regulation CPU) that was connected to the accelerometer interrupt line, mainly for freefall detection, could potentially sense haptic taps through the taut tether in the daytime, not just at night, requiring only human touch at the tether near the ground--no flashlight needed. But Huckleberry is confined to power regulation functions and cannot control the motors to send a vibrating pulse back down the tether for two-way communication. Sawyer, however, does have PWM control over the motor and could generate such a pulse, but these two I2C slave devices cannot talk to each other directly through I2C, only through the I2C master CPUs which would be unavailable in an extreme low-power situation.
So I created a separate Hackaday project called #Tethered Haptics Using Morse Pulses where I created the ability to daisy-chain the decoder, allowing it to run simultaneously on both ATtinys to allow haptic two-way communication over the tether. I then entered this project in the 2018 Human Computer Interface Challenge.
I had to modify the SIMTHEO algorithm, inverting it to time the off-times, not the on-times, since I don't have a continuous signal anymore. The LIS3DH accelerometer senses acceleration, not velocity, and so a continuous acceleration for the lengths of time needed to signal a normal "dah" would require massive thrashing of the tether, unless the pauses were used.
Systems That Had to Be Created
I enjoy building interrelating systems and coming up with creative acronyms, and while the basic functions of the craft are easy to understand, it was really quite difficult to achieve this unique combination of systems at low levels of cost, and to do it within my own time, energy, and knowledge limits. Most of it was in fragments until the very end when I could assemble all of its interacting pieces (electronic, mechanical, software, physical and logical systems), a very uneasy feeling for two long years, that made me wonder if I would ever get it built. It is the most difficult project that I have ever constructed.
- To minimize motor complexity while achieving high solar panel efficiency and reduce the craft's pendular motion, I designed the TROT mechanism. This required that I build a powerful MOSFET H-Bridge into a screwdriver (taking advantage of its strong planetary gearbox) and then used custom PWM algorithms, hall-effect sensors and an accelerometer to perform slow starts, braking and "orbital mechanics" of the planetary mass within the mechanical and electrical constraints. Note that the same algorithms that create an "unstable equilibrium" for inverse pendulum robots can be employed in a simpler fashion to create a "stable equilibrium" for normal pendulum robots, but this incorporates both normal and inverted pendulums at some points. But the pendulum can also act as an active roll damper.
- To drive the motor "smoothly and accurately" for long-term locomotion, elevation control, and solar alignment, I had to create the Delta-v Motor Drive System, a high-level system that uses a mathematical model of the torque curve for mostly feed-forward control, using a complex system of auto-calibration, increased precision, pre-generation with real-time streaming of motor torque values (electrical current) using hardware flow control, and phase-lock motor-correction heuristics to keep the motor on its idealized "orbit".
- To further stabilize the craft's pendular motion, I built an experiment Gyro (for inertial damping and reaction wheel) around a lightweight, but powerful DVD-ROM 3-phase BLDC motor, pushing the limits of an ATtiny microcontroller to control two different types of motors and PWM, again using custom PWM algorithms and hardware timers/interrupts. Gyros are really actuators, not sensors, and I wanted to explore this application. I like the idea of "invisible forces", and there is also a passive ball damper inside the planetary mass. Note that the planetary mass cannot be machined using a subtractive process and is an object that I created to demonstrate the unique benefit of additive 3D printers.
- I had to write my own rudimentary XMPP server in Lua on an ESP8266 (since I couldn't find one as most people use them for MQTT instead). I have only tested it with a particular version of Android Xabber, my favorite open-source XMPP (Jabber) client.
- In early 2017, I published a microcontroller-based, hand-sent Morse decoding algorithm called SIMTHEO that I had written for my roguelike game but also thought could be adapted to this project, since it uses the same microcontrollers. As described, above, in conjunction with the accelerometer, it was modified into a new algorithm that compares the pauses and not the tonal lengths (called SIMTHEO SHH) to allow haptic communication with the craft via the tether (via human tapping) in times when either the power levels are too low or the radios or clients are not available. This system is a separate project called #Tethered Haptics Using Morse Pulses .
- I couldn't use the wonderful Chirp software since I had to take more granular control over a UV-5RA flash memory than what it provided. I needed to quickly change frequencies on-the-fly and take over its LCD, a use-case that it wasn't really designed for, and I had to write my own code to control the radio interface that I built to replace a TNC, which uses the Direwolf software modem, a virtual UART and a virtual PTT line. This allowed me to be able to quickly change the radio's frequency to the APRS frequency while AX.25 sessions were in progress on a different frequency, using an elaborate "stupor" method to pause the session in progress, and then change back, which allowed a single, inexpensive transceiver, a single VFO, to perform double-duties.
- The craft had to be printed within the dimensions of my Prusa i3 printbed, while being designed to withstand extreme weather in all four seasons in my temperate zone (PETG, aluminum, brass, stainless steel parts, air and rain channels). For temperature control of the Lithium-ion batteries, a gasketed battery box was created using black, high emmissivity paint for radiated solar heat which can be temperature-controlled via panel tilt and a ducted Gyrofan and aluminum heatsinks, monitored with a series of temperature sensors of different types.
- I built a multi-user PBBS from scratch using the Unix interface instead of the C library, something very uncommon and underappreciated in my opinion, which had its own set of difficulties due to the delimiter incompatibilities and the parallel processing of multiple text streams in combination with AX.25/APRS-to-KISS-to-UART-to-TCP-to-802.11n-to-XMPP routing. A lot of people had written packet BBSes in the 1990's, but I haven't seen many people do this recently, and very few in Python, so I wanted to give it a shot.
- There are 4 CPUs (Raspberry Pi Zero W, ESP8266, two ATtiny1634s) that can be re-programmed in-circuit and use 4 different programming languages between them (Python 3, BASH, Lua, C) that have to sync data, time, and control their states using a minimal number of wires, which was logistically complex (AX.25/APRS/KISS, TCP/IP/XMPP, 802.11n, UART/UV-5RA Serial, I2C, 1-Wire), incorporating Python multiprocessing, C interrupts and job mechanisms, true parallel operation of two interdependent microcontrollers, hardware timers, atomic operations, various global state tracking, and IPC communication methods that cross abstraction layers.
- Unlike a typical wheeled robot that doesn't transmit over amateur radio frequencies, it is inherently difficult to test, calibrate, or demonstrate the unit indoors. On a table, the dynamic torques are not available and the motor behaves completely differently, and it hangs from a long, tight tether that requires a lot of sunlight, so I had to build a wooden Ark and Test Frame that was inexpensive, foldable, and could support the tension of the tether and also function as a storage, transportation, and demonstration apparatus. The wooden Ark is an integral part of a larger machine, as I was fascinated by the 1947 Uncrating and Assembly of the P-47 Thunderbolt Airplane documentary where the wooden crate was essential to the assembly of the plane. I had to familiarize myself with pine, steel cable (wire rope), turnbuckles, and clevises, a vast departure from 3D-printed plastic. I also had to create a software simulation of the AX.25 packet communication and build dummy loads around a second radio and packet interface called TRILLSAT-G (for groundstation) since I can't just spew garbage all over the ham frequencies.
- It seems to have enough battery storage (two battery banks) to maintain communication through the night, but I haven't performed a full day/night cycle in actual testing. It uses DC-DC boost converters, but for high-power operation it switches both battery banks in and out of series (doubling the voltage and not subject to the boost converter current limitation), using a mix of logic-level MOSFETs and relays. Surprisingly, the electromechanical relay is not obsolete; its low contact resistance and powerless states (latching) are ideal for voltage and current-sensitive circuits.