August 2013 – May 2014
This project began after I read a Servo magazine article my freshman year of high school extensively detailing the steps required to scratch build a quadcopter. I was familiar with multirotors but this article was mind-blowing because I didn’t really consider someone could actually BUILD something as cool as a quadcopter from scratch. For the first time, I was really “taken” by an idea and super inspired, and decided I had to build a quadcopter! This coincided very nicely with me joining my high school’s robotics team and I had just recently learned how to do computer aided design (CAD) with Autodesk Inventor. Joining my school’s robotics team also made available a bunch of tools such as 3D printers, CNC machines, and laser cutters, all of which I found super empowering and I was open to the idea of using them in a project of some sort. This would be my first really large, ambitious project, and was my first real foray into the “maker” movement which was going strong at the time. I was hoping to build a fast quadcopter, powerful quadcopter that could get silky-smooth aerial footage, another big inspiration for this quadcopter once again being David Windestål’s aerial tricopter footage. Thanks to the Servo article I felt building my own quadcopter was within reach so I went for it!
Here are a couple videos I made of my quadcopter all those years ago; please excuse the Microsoft Movie Maker editor captions, poor renderings, and general cringe, we all have to start somewhere! The first video starts with a little montage of the build. If you’re interested in reading more details about my quadcopter project keep scrolling past the videos where I share more details.
Up until I built my quadcopter, I’d had relatively limited R.C. experience. As a kid, I’d had plenty of air hogs of both the quick charge and (2 channel? I think?) variety. From there I’d upgraded to a little (wingspan about 12″) parkzone flyer 3 channel, which I flew extensively for a while. However, after this I became cocky and bought a large model BF-109 with my weekly allowance. When it finally showed up in the mail, I could hardly wait! I assembled it and took it to the largest park my Dad and I knew of, and tried to take off. I discovered really fast that this thing was waaaay above my skill level. I only managed to get it a couple feet off the ground before I panicked, crashed it into the ground, after almost running out space in the park. I tried to take off again, with similar results, me crashing into the ground and damaging my brand-new plane. At this point I gave up and went home dejected, never to touch R.C. again until I built my quadcopter. It was a scary thought having never flown anything for years, having to fly this labor of love that I put a lot of money and time into.
I ended up settling on the electronic setup as follows;
- Turnigy Multistar 30 amp ESCs
- Turnigy SK3 Aerodrive 1240kv motors
- Turnigy 6x FHSS Transmitter and Receiver combo
- Hobbyking Power distribution board
- Zippy 2200 Mah 3-cell battery
- 3 cell low battery beeper
- APM 2.6 flight CPU
I designed the quad to have roughly the same dimensions as outlined in the Servo magazine article, about 500 millimeters diagonally across from motor mount tip to tip. Initially I designed it with metal c-channel arms in mind, and I designed the two arms to slot together in the middle. I planned to get the slots milled out on the school’s CNC machine. However, milling the arms out of metal on a CNC machine would have been a project in and of itself, and I had no CNC machine experience so that was intimidating to me. In retrospect, I could have easily manufactured these slots and drilled the holes in the C-channel to a “good enough” quality with a bandsaw or hacksaw and a drill, but I was going for perfection with the metal arms, which is not a good thing to do all the time. However, going back to my thought process then, these metals arms were the only thing holding me back from building my quadcopter. I already had all the electronics and wooden parts laser cut just sitting around. Pretty much on a whim, one weekend I just decided to use wooden arms and start building the quadcopter then and there, which I did! This actually was a largely good decision since not only is wood simpler to work with (no fancy metal cutting machines require), it absorbs vibrations from the motors and propellers a lot better than metal, which later on would already cause me enough headaches as is later on. After I made the impulse decision to build my quadcopter arms with wood and the tools I already had at home, on hand, the assembly process was quick and easy. I assembled the quadcopter together in a couple days and within the first weekend was ready for my first test flight in my backyard.
I eventually bought a GPS and began experimenting with the autonomous flight features offered by ArduPilot, the most important of which was Ardupilot’s return to launch feature. With a simple flick of a switch on my end, or more importantly, if my transmitter lost connection to the quadcopter, the computer would automatically take over and autonomously the quad back to its initial flying location and land. This not only was an amazing feature to see in action, but it me from completely losing my quadcopter on several occasions!
I also experimented for the first time with placing a brushless gimbal motors, and designing my own gimbal, but I didn’t take these efforts too far. I was trying to use the built-in gimbal connectors and gimbal software on the ardupilot, but I didn’t realize that they were only meant for controlling servo gimbals, not brushless motor gimbals. Needless to say, I never did take that too far, but I thought it was worth mentioning as this was the first time I tried to get a gimbal working. Later, I would get my own brushless gimbal working somewhat successfully on my tricopter, although even then I still had a lot of trouble with it.
Overall, I really enjoyed flying my quadcopter! It flew quite well, and it was fast and responsive.
However, the quad certainly wasn’t without its issues. First and foremost, vibrations were a huge and constant problem for me. I was using Turnigy motors which use low-quality bearings, which in turn wear out fast and almost always got ruined in crashes. I spent massive amounts of time balancing my motors and propellers trying to get them to run nicely without ruining bearings, and in the end I decided it just wasn’t worth the effort. To compound my issues I discovered only just recently (2016) that the propeller balancer I’ve been using up until now is completely worthless. The balancer I used didn’t run true and was completely unpredictable, causing me lots of agonizing guessing games in regards to the propellers and a lot of hair-pulling frustration. I actually suspected the prop balancer for a long time, but it was one of those situations where I was (again) too lazy to research and buy a nice one for waaaay to long. Don’t do that. If you want to build a multirotor, save yourself time and frustration and just buy yourself DuBros prop balancer, they’re surprisingly cheap and will work wonders. Another problem I was running in to was the lack of real estate and overall sketchiness of my quadcopter. One big, big oversight I made when designing this quad was my lack of attention regarding exactly where I was going to put some of my electronics. I was so excited to just get this thing flying I just ignored some of the little details which is a terrible idea. I went to route of “Well, this frame should have enough space for everything”. In my excitement and drive to get this thing built and flying, I knowingly overlooked a lot of little but important details in lieu of “I’ll figure something out when it’s built”. The method of figuring things out later worked, but it wasn’t pretty. For example, I secured my battery to the quad with duck tape and zip ties. I built a sketchy system to sling the GoPro beneath the quad so that the props didn’t show up in the shot (although they still did frequently) because I didn’t want to think about how to remove the propellers from the shot in the design process.
Finally, I wanted to try first person view (FPV) flying with my quadcopter, but I had no room for additional add-ons such as FPV transmitters and antennas. Add the high maintenance time require for balancing the motors, balancing the propellers (with a bad prop balance), combined with the overall sketchy aspects of the quadcopter, and after a while I was stuck with a frustrating toy that often ruined my aerial footage because of excessive vibrations despite my best efforts. Throughout my time with this quadcopter I came to realize what I was most interested in was not aerobatic or high-speed flying, but aerial cinematography. I wanted to get epic looking shots with my quadcopter, but I realized it just wasn’t up to the job of doing it. At this point, about a year after I had started on my quest to build a quadcopter, I realized I had reached the limits of my platform and began to look for something new. I wanted something that was more reliable, with more room for expansion, and that isolated the camera from vibrations much more effectively. Eventually, I would end up building my own tricopter.
That’s not to say I didn’t enjoy a lot of the time I spent with working with and flying my quadcopter! The quadcopter was overall A LOT of fun to fly, very fast, and very responsive. I enjoyed flying the quad in computer-stabilized modes as well as in an “acrobatic” mode where computer stabilization was turned off, partly because I didn’t yet understand how to tune PID algorithms effectively so I had trouble getting a super reliable stabilized mode. I dabbled for the first time with autonomous flight modes which was amazing to see, and I was out flying pretty much every moment I could be. While I did have problems with the quadcopter, overall I’m glad I took on this challenge it was a lot of fun!