Lately the news has been buzzing with stories of drone technology. Be it for delivery, recreation, or military purposes, unmanned flight is the future of aviation. Beginning in December 2013, I began working on a drone with several of my friends. We have created a drone that is 100% our own design. Using Autodesk Inventor we designed, drafted, and printed each part of our copter's frame. We converted our files from .IPT to .X3G file, allowing us to print our frame on a Makerbot 2x Replicator. All of the external wiring was done and reviewed by each person working on the copter. The three other people who created this design have the Instructables accounts BobBurghart, CoreyOrlovsky, and RichardP4.

Step 1: Acquiring Components

Before any of the frame could be designed, we needed to consider the size, shape, and placement of all components. Our design is centered around an ArduPilot 2.6. In addition to this, we have one power distribution board, four speed controllers, four motors, and an FM receiver that must stay safely on the copter. All of these parts, with the exception of the GPS and FM controller for signal reasons, are kept safely inside of our frame, and will not be vulnerable to crashes.

Below is a list of all purchased components on our device:

APM 2.6:http://store.3drobotics.com/products/apm-2-6-kit-1

Radio Set (Android Compatable):https://store.3drobotics.com/products/3dr-radio

Compass Kit:https://store.3drobotics.com/products/3dr-gps-ublo...

PDB: https://store.3drobotics.com/products/quadcopter-p...


Speed Controllers: https://store.3drobotics.com/products/esc-20-amp-S...



Power Module:https://store.3drobotics.com/products/apm-power-mo...

RC Reciever:http://www.ebay.com/itm/like/171474770621?lpid=82


All other parts of our device were made in AutoDesk Inventor, and fabricated by us using a Makerbot 2x Replicator.

Step 2: Designing the Center

Our center was designed to fit all components and allow no movement in the event of a crash. The bottom part of our frame is designed to house both the battery and power distribution board. The top of our frame is designed with compartments for our ArduPilot 2.6, FM Receiver, and GPS. Both were designed on Autodesk Inventor and printed on a Makerbot 2x Replicator under high quality settings with the bed heated to 100 degrees Celsius and the extrusion head at 240 degrees Celsius. We used PLA plastic instead of ABS because the PLA printed from our printer was found to be more durable. In step 8 are the files for the center of our quad copter.

Step 3: Designing the Arms

Our frame and arms are designed to fit together using three 4" long 1/8" thick screws per arm. Our arms are 21 centimeters long and are designed with trusses to create the strongest structure while also keeping the weight low. Our motor mounts are designed to house our motors if the housings fly off. The motors are attached with screws underneath them. These mounts were designed using Autodesk Inventor and printed using the same settings as our previous prints. The arm file in step 8 is a .stl file of our arm.

Step 4: Printing

Export all files as .STL files from Autodesk Inventor's file type .IPT. From here we open Makerware software and scale all files to 100%. Set to the settings described in Steps 1 and 2, then save the printable file onto a secure digital (SD) card. Put this into your printer, and wait for the prints to finish.


Step 5: Wiring

Plug the Battery into the PDB and the PDB into all speed controlled motors. Make sure that the polarity is swapped on the "push motors ( motors on opposing corners with pull propellers) The wiring is quite straightforward:

1) Plug battery to APM Power Management cable, and APM power cable connected to the APM power input, the Deans connector should be out from Power Management cable to the Quad copter Power Distribution Board
2) ESC are to be hooked into the Quad copter Power Distribution board

3) The PDB should lead to all speed controllers and motors

4) The receiver should be plugged to the APM, Set all control components and plug them into their corresponding ports.

Many copters have been made using this setup.

Step 6: Propeller Setup

All propellers rotating clockwise should be push propellers (motors 3 and 4 in the diagram), and all props rotating counter clockwise should be normal propellers ( motors 1 and 2 in the diagram). Ensure that you have proper motors for your copter and proper propellers for your copter. Our propellers are 10.5" props that have effectively lifted our machine.

Step 7: Flight

Now that you've created your quad copter, it is time for flight! Our machine is flyable by either GPS or FM transmitter. Have fun!

Step 8: Compatable files

There were some complaints of non-Inventors designers being unable to view or edit my files. I reformatted our parts as .stl files, and hope that they are usable by all.

If further problems occur, please post a comment and I'll be sure to respond as fast as I can.

<p>is that US because im from Australia</p>
<p>Hi..i hav a 3d printer.so wat will b the minimum cost for this pjct?</p>
<p>I just bought the parts from a combination of 3d robotics, amazon and ebay for ~$450</p>
good I need this project code
<p>How much did this cost you?</p><p>Im interested in making this but need to save money</p>
<p>i dont own an android but if theres a cheaper way to control it, i would be interested. thank you for any help.</p>
<p>I'm confused about the radio interface. Is this to be controlled by an Android app (do you have a link?) or by the RC controller, or by either?</p><p>Also, I have a Printrbot Simple, and I'm not sure it can manage the full arm length. The build area is 150 by 150, but it will need room for the raft. Any suggestions on how to modify the arm to fit in this area? I don't think it will be as strong if I cut it in half and print both halves.</p>
Whenever I had that problem I would cut the arms in half and print the two and then (if using ABS) glue together with acetone. Cuttimg also allws you to define a flat surface from which to print up. I also have a 3Doodler which I use to merge parts together and gives me joints of the material fused together. Granted handling complex joints is a pain but I use acetone for the original weld and then the 3Doodler to clean up seams and very minor warping. Finishing with a small file recovers the part and we are in business -- sometimes I dont feel like reprinting and waiting if I can repair and move on.
<p>i have spent countless hours running experiments with a Makerfarm 3D printer and i found that printing in smaller parts and welding using acetone is a good idea. It reduces the risk of the prints failing and having to start over. And if it does fail you don't waist as much time. </p>
<p>Hey mgoodfel,</p><p>With our radio setup and with our FM reciever, the copter is flyable from android and controller! Either control setup can be prioritized. We decided to prioritize the RC aspect of our drone in case we must manually land. As for the arms, you'd have to print them in two parts. It will not be nearly as strong, but there is no other way to ensure that it will fit together with the rest of our build. </p><p>Kyle</p>
<p>Well, the RC controller is $140, so if you could do without it, that would be nice. The RC receiver link seems to be bad now. Could you update it?</p>
<p>You need some sort of RC controller to arm the machine. I'll update the link no problem.</p>
<p>Over all, how much would this cost, and is there a way I could cheapen it? </p><p>I am too young to be able to fund a 1,000$ drone. <br>And also, this looks awesome. </p>
Hey. What is the total weight of the frame, I am building a quad and 3d printing it but I need an estimated weight
<p>please add a video<br></p>
<p>where are the videos of the flights?<br><br>does it fly well?<br>speed? lenght of flight?</p>
<p>Hey Patrice,</p><p>I'm sorry we weren't able to fly our copter for you on Friday. Every Tech teacher needed to prepare for a Google seminar being given at our school. Due to this, we had no supervision to fly, and by our school's rules, we could not fly. Perhaps on Tuesday (Monday is Columbus day, so we have off) we can get you your videos and statistics. </p><p>Thank you for your patience,</p><p>Kyle</p>
<p>yes please add a video!?</p>
<p>I plan to post videos of flight, more professional pictures, and detailed analysis of calibration and flight tomorrow. It has flown decently well, but landing seems to be a problem haha. I'll be sure to record all of your desired statistics upon our flight tomorrow.</p>
<p>Have you done a stress analysis on the truss based on prop thrust and motor torque?</p><p>Your truss is fine for handling the motor torque and the constant section modulus is correct for the constant bending moment over the length of the beam produced by the motor torque.</p><p>On the other hand, the prop thrust is in the vertical plane and creates a bending moment that is maximum where the truss mates to the body and basically zero at the rotor. This varying bending moment requires that the section modulus of the beam be max at the body and minimum at the rotor. Since, by inspection, the beam appears to have a constant section modulus there might be an overuse of material resulting in a craft that is over weight and a payload that is under weight.</p><p>I am aware that material density can be varied with 3D printers and that you may have varied the density. that is just fine for that results in a varying section modulus thus compensating for the constant cross section.</p><p>I would think that an oval cross section would be optimum for the beam with the major axis in the vertical plane. The beam would have to be printed with its length in the vertical plane. I doubt that any webbing would be required for the cross section of a tube is very &quot;nice&quot; (efficient) design. Again the cross section should be varied over the length of the beam. Also the hollow tube is excellent for housing the component (wire) required to supply the motors with electrical power.</p><p>Please take my treatise of your creation kindly for that is how I wanted my thoughts be conveyed.</p><p>Again, very nice. Keep creating what I cannot, due to time and finances. I will enjoy from afar!</p>
<p>Our ideas were seeded in creating a body that was able to withstand impact and only break in certain designated areas to ensure the safety of our components(like a crumple zone) The only time our arm broke was upon a failed landing. This was exactly as it was designed to break. The upward force of our motor is negligible from the pressure sustainable in the PLA arms, so we aren't really concerned on that point. We designed our arms with triangular honeycombing because the large quantity of faces ensured structure. If you wish to modify our design, we would be honored and you could post your own design that I'd be sure to look at!</p>
<p>Thanks for the kind reply.</p><p>I'd like to clear up some misconceptions.</p><p>The thrust is not insignificant. It is very important. It is the lifting force. The thrust per rotor, with four rotors, is greater than (craft weight + any payload(electronics))/4 + any (design acceleration)/4.</p><p>Adjusting density is very nice. Because of the constant cross section over the length of the arms, I hope you varied the density over the length of the arm?</p><p>Yes, crash adds interesting loading and is proportional to the rate of deceleration at the crash. Flexing of the copter frame helps absorb crash loads. Flexure is not only aided by material properties, but is aided by section modulus as well. And no matter the impact loading the bending moment in the arm is greatest where it mounts to the body.</p><p>Again keep up the innovative designs!!</p>
<p>The upward force of the motors on the arm is what I was calling insignificant. The thrust upward is nowhere near insignificant, I simply meant toward your concern of the breaking of our arm. Our printer (or perhaps our inexperience with it) would not let us vary the density inside of our arm, and we therefore needed to design it with the supports as they were.</p>
<p>Finally a nice step by step to build your own quad. Thank you for posting this. </p><p>Two stupid hardware questions. Do you need the 7 channel receiver you linked to when the transmitter you linked to comes with one? It looks like the APM comes with the Power Module, do you need a second one?</p><p>Thanks again.</p>
<p>Are there any videos of it in flight?</p>
<p>Hey Jake,</p><p>We do have one video of it in flight and all members are searching their computers for where it was stored. In its current state, all components are taken out of it for pictures we plan to post on this instructable. I will be sure to post a flight video once it is rebuilt.</p><p>Kyle</p>
<p>Hey Wroger,</p><p>We used a 25% fill, and our frame is quite light. Balsa warps when scratched, so we feel that the balsa frame would require such a great rebuild time that modifications would get lost in recreating what we already have. It is important to consider factors past aerodynamics.</p><p>Kyle</p>
<p>Hello,</p><p>I pulled the arms for this quad into TinkerCAD to look at them. I noticed that you have inches defined as the dimensional units (for all you guys trying to pull the STL in and slice it), but I also noticed that when I import it, the arms are 500mm long! Now if you printed these on a MakerBot 2x, that can't be right. So obviously something is messed up with the import. What is the length of the arm end to end and the maximum width (which will be the outer diameter of the motor mount? I can scale it down from there. Thanks!</p>
<p>Hey Samern, <br>Something is definitely wrong with either the import or export, I'll be sure to check that this issue isn't my foul up. The length of our arms were 21 cm (210 mm by your scale) and the scale should work on the constrained object, adjusting the width properly as you adjust the length. the motors are 28 mm by 30 mm, and our motor mounts were designed slightly larger than the motors (without any 3D drafting software at my home computer I do not have exact dimensions) so as long as your 21 cm arms are slightly larger than the motor size you'll be good to go. Ill attempt a re-export on tuesday.</p><p>My apologies, </p><p>Kyle</p>
Thanks Kyle. I don't think it's you. It's more likely the export software. I would suggest you embed/define the units as mm. When I reduce the part down, the magic number seems to be 40%, which gets me to 211.67mm, which is pretty close to what you have. The diameter of the motor mount comes to just about 40mm at its widest point. That seems about right because a 30mm cylinder fits perfectly inside the opening for the motor. I can upload the STL I have but have no way to do so without creating an Instructable of my own. BTW, Instructables won't render the STLs which suggests something is wrong with the file. Normally, I would see the STL rendered right in Step 8. In any event, I am hoping to mount an FPV onboard this. Do you have a final weight to your design? I'd like to see if a motor upgrade is required before I try to load a camera on it.
<p>The magic number is actually 39.37%, which *surprise* is the ratio of cm vs. inch.</p><p>Everybody: This is engineering. We use metric units in engineering.</p>
Well isn't that interesting. Oddly TinkerCAD blew chunks when I tried to scale down below 40% to get closer to 210mm (I got 206mm). Those files definitely need an update. Thanks!!!
<p>It might help to change your default units to metric before importing.</p>
<p>Hey Samern,</p><p>We exported in metric, and designed in metric, so there shouldn't really be any problem with standard units on our side. The best thing I can suggest is opening in a file previously set to metric units. </p><p>Sorry, <br>Kyle</p>
Confused as to how you arrived to PLA being more durable?
<p>We preformed our own stress tests, and results were pretty conclusive with PLA coming out on top.</p>
<p>Hi, I noticed that you have said that PLA is more durable than ABS?? Form a structural point of view, ABS is superior to PLA! stronger, more flexable better in compression and tension! please check material data sheets and compare and you will see. A lot of 3D Printing people will tell you PLA is &quot;Better&quot; this is only because it is better to print, meaning it does not warp as much or shrink like ABS will. However a properly tuned printer will print ABS just fine.</p><p>I have probably stated a massive debate about PLA vs ABS..lol, but material data sheets on mechanical properties do not lie!</p>
<p>We did not use data sheets, but instead conducted our own stress tests of compression and snapping. Our data concluded stronger PLA (possibly because our ABS was warped), but out of our two plastics, on our machine, the PLA was found to be superior. Perhaps our printer was tuned incorrectly, but for the purpose at hand, PLA was a much better option.</p>
<p>One more thing to note on PLA, It can be composted! and is environmental friendly due to the ease it can be broken down.</p>
<p>This is excellent. I'm building one myself that can hover steadily and carry two cameras. One for good quality HD video and one as the pilot. I think it's called first person viewing. I've got the disc and dome spun from aluminum that will house everything with 4 holes for the propellers. You guess it! Flying saucer. </p><p>It's 30&quot; round and spun in 3 pieces. I have abandoned the idea of welding them (24 gauge is too thin), and found some adhesive used to build aircraft and buses. So far I've reached the stage where I fitted the guts from a foam framed quad I bought but it does not have the power to do much more than bounce across the grass. Your information has solved some issues for me. Hopefully the next flight will see phone calls about UFOs to the news stations. I'll let you know when is ready to fly . I've documented it all for an instructable.</p>
<p>Very Nice. I would like to make my own as well. Thanks!!</p>
<p>would it be possible to modify a rc heli to be a drone?</p>
<p>STL files, please. </p>
<p>Hey, </p><p>STLs are posted in step 8</p><p>Kyle</p>
<p>Nice work. </p><p>I did not see all of your references, but I assume you designed the frame and adopted the controllers. I would like to see how your PID tuning proceeds. I congratulate you on building and flying your Quad!! They are hard to make. A good site for help is <a href="http://diydrones.com/." rel="nofollow">http://diydrones.com/. </a> </p>
<p>Hi,</p><p>I'm a high school instructor and my students want to make this: can you create a list of <strong>all the electronic components with brand names</strong>, including the battery, Transmitter, joystick reccomended. We have been wanting to do this and want a good basis for success before we dive into one. Your project is Awesome and I looks like it took some time and thought. We have all the items(tools) you have inventor / Rep 2X/ seems like a great match....Thanks for sharing. ..</p><p>Originally We really wanted to create one using bluetooth and the ipad as a controll, (every kid has an IPAD) but I fear that with all the OS changes the software will be out of date a month after we purchase the app to controll. </p><p>Thanks,</p><p>Todd </p>
<p>Hey Todd, </p><p>I think that this build is great for high school students as long as you have several smart and dedicated kids! It took us about a year to design and assemble it. Though we have a working model at the time being we will continue to upgrade and innovate our design until college. I linked to all non-printable components we use, and I wish you luck in your build! Ill be sure you keep you posted with all updates and changes.</p><p>Regards, </p><p>Kyle </p>

About This Instructable




More by Nolan5454:Homemade Fishing Lures for Cheap Spooky Ghost with Arduino 3D Printed Quadcopter with Arduino 
Add instructable to: