Quadcopter Frame Design

My first quadrotor frame design is simple, sturdy, reliable, and a bit ugly.  I have made no attempt to make it cute, flashy, or visually desirable in any way.  If and when I determine that the structural design performs well, I’ll use some better tools to make a more flashy design.  Until then, I’ll be flying on my raw cut aluminum frame.

My basic design is two 24 inch(610 mm) X 1/2 inch(13 mm) X 1/2 inch(13 mm) square aluminum arms.  I knotched both in the center so that they can cross each other.  There are two 5 1/4 inch(133 mm) alumimum plates holding everything strong and square.  The frame is very strong, rigid, and relatively low weight.

    • Scott
    • January 4th, 2012


    I’m thinking about building my first quadcopter. I was thinking of a similar design. How did this turn out for you? Was it stable? Is there anything you’ve learned while building that may help? Thanks for posting this!

    • Hey Scott,
      I haven’t actually got it flying yet. I’m finishing up a Master’s degree and haven’t found the time. The only thing left I need to do is finish up my code and attach the processor and sensor board to the frame.

      If I were to do it over again, I’d probably just buy a frame. Making one turned out to be a pain in the butt!

      I’ve learned a lot along the way! I previously built a stabilization system for a normal RC helicopter. I can’t think of any REALLY critical things to tell you, but I’m glad to answer any questions or be someone for you to bounce ideas off of.

      Don’t hesitate to email or post questions! Good luck


        • Scott
        • January 17th, 2012

        Thanks Nic!

        Is the working with the aluminum the part that made it suck? Hahah.. I have some tools including a drill press and CAD layout, so hopefully that will make things easier.

        When building your own quadcopter, how do you calculate what size motors and props will be needed? Thanks for answering my newbie question!

        I hope you finish yours soon!


      • Hey Scott,

        Yeah, the aluminum part wasn’t fun. A drill press would have made it really nice though.

        Unfortunately, I didn’t take the scientific route when deciding which motors and props to buy. I researched many different designs and made an educated guess. Obviously the better way would be to calculate the weight and moment of inertia on the frame and use physics equations to determine the proper setup. There is software out there that can estimate the amount of thrust that a given motor and prop combo will produce. Another issue with quadcopters is the amount of yaw you can expect from the motors and props. Just because the copter can lift doesn’t mean it can turn!

        Basically, I just played it safe and bought parts that many other people have bought. They are more on the cheap side so if I hate them, I’ll put more math behind my decisions. My combination supposedly can lift ~1Kg per motor/prop. That should be plenty for my initial design.

    • Jutland
    • July 8th, 2012

    Nice design, I like it. By the way which software is used to calculate thing like “the weight and moment of inertia on the frame”?

    • I don’t know of any software that does this. You’d have to make your frame, measure its qualities, and choose your motors and props accordingly.

  1. does it work by arduino ?/
    please answer me

    • The frame could be controlled by anything, however, I’ve chosen to implement the quadcopter software using an ARM Cortex-M3 processor in a LPC1768 chip. I has more processing power than typical Arduinos.

  2. what about using gyro ??

  3. what is the hardest thing facing my during making quadcopter ?

    • That depends on how much you want to do yourself. There are several ready-to-go systems out there, that require very little to no programming. Mine is a from-scratch approach. I bought a microcontroller and have implemented every library from the ground up. You could use an Arduino to cut down the interface design time. That would leave you with just implementing the control logic and building the physical aircraft.

  4. how much does it cost ???

    • The costs are constantly changing as parts are changing. At a minimum, a quadcopter will cost around $400. You can get really excited and spend $1000’s! If you like embedded system design, this is a really fun project to build.

  5. does it work ?

    • Mine is still incomplete. I have made an autonomous helicopter before using a T-Rex 600. Quadcopters in general are very successful: see here

  6. Hey,
    I bought a Triple Axis Accelerometer & Gyro- MPU-6050 to work with an Arduino Uno, Would it take much time and effort to work with this or it would be better to use for instance, a control board that requires no programming skills? I’ve just 7 weeks to be finished with my quad.

    • Of course it is faster to use someone else’s design, but you used the word “better”. It is “better” to implement it yourself! You should definitely keep the Accelerometer/Gyro combo chip and use it. I’m assuming you have this breakout: https://www.sparkfun.com/products/11028. If so, there is a listed library for the Arduino. I haven’t used it but I’d imagine that you could get it up and running quick. You may have to use a sensor fusion algorithm to combine the accelerometer and gyro data. You can do it in 7 weeks!

    • Biraj Acharya
    • June 1st, 2013

    Thanks nic for the post. I have measured the output of the receiver using oscilloscope. And I have found that the gap between two consecutive channels is just 1 microsecond(for Hobbyking Tx and Rx 2.4GHz). So I have been feeling difficulty to read the signals using microcontroller. Please could you specify the microcontroller you used for your project. And please could you send me your C codes at (acbiraj@gmail.com).

    With regards
    Biraj Acharya

    • That’s very interesting. 1 microsecond seems very hard to deal with. I’m using an ARM Cortex-M3 microcontroller in a LPC1768 package. It runs at 100MHz.

      I sent you my code.

      Given your short gap time, you may have to try another approach. Even if you make your interrupt routine VERY efficient, it will only give you a handful of clock cycles to get everything done. For example, 1 microsecond at 100Mhz is only 100 clock cycles. This is borderline impossible. You’ll probably have to read the signals in on separate timer modules and pins.

      If you do try to do it like I did, you need to strip everything out of the interrupt routine. Your interrupt routine should only pull a timer value off the hardware and place it into some data structure for processing in the main thread. This will minimize your interrupt routine time. I’m still not certain you can do this in 100 cycles.

      Another reader has used this receiver and had success. Read the comments on:

      Good luck. Let me know how it goes.


    • shubham
    • July 1st, 2013

    can you please tell me the dimension of this quadocopter???

    • The dimensions are listed in the text above. Are you looking for different dimensions?

    • Antti
    • July 7th, 2013


    Hope you get your quadro running! I have bit similar project ongoing, though i was going to use arduino mega. I have built the frame, also remote control (running arduino uno) . Mow ia am bit afraid if arduino is capable of all the math required, also my fear is the response time of my remote controller(using nRF24L01+) i.e. If the radio link is fast enough… Well time shows and if it does not work well enough, i will buy factory made rc controller. If its fast enough and working, i have possibility to transfer some data from quadro back to rc controlled and thus also for pc (these have 2 mbit/sec max transfer). I also have ordered motors (around 975 kv with 10×4,5 propellers), now just waiting for stuff to arrive.

    Really interesting project you have, lets see if you get it running before me 🙂

    • If it comes down to a race, you’ll definitely win! I haven’t had any time for over a year now!

    • tina
    • July 28th, 2013

    thanks for post this
    i’m working on my quadcopter, I bought the hole component, now to built the frame , how can i decide the dimensions and the weight of the frame?if i wanna use almu
    i used (4x A 2212 A2212 1000KV Brushless Outrunner Motor )
    10*4 , 5 r Propellers

    • I don’t have a good answer for you. I found an online calculator that estimated the amount of weight mine “should” carry (I can’t remember where this was). I overshot the weight by 3-4x to make sure it was OK. It would be nice if there was a better way to figure this out thought. Sorry I couldn’t be more help.

    • jhon
    • September 26th, 2013
    • Henry
    • December 17th, 2013

    Nice project! I am working on something similar and using the same LPC board. I was wondering if I could take a look at your code for reference?

      • nic
      • December 17th, 2013

      Sorry Henry, I don’t have a complete design, yet. This project has taken a back-burner due to other life events. However, if you have specific questions, please ask. I’ve covered a few topics already here, and hope to cover the rest. Once my project is complete, I’ll put my code on GitHub. Thanks for reading

        • Henry
        • December 17th, 2013

        Thanks for the quick response. I do have a question:

        For the PWM decoder/encoder drivers is there an actual decoder/encoder driver or did you have to code that yourself? ie. the pinout for the LCP board has 6 PWM outs but no ins.

        • nic
        • December 17th, 2013

        For PWM input, I developed my own mechanism that has gotten a lot of attention. See this:

        For PWM output, I used the PWM outputs from the LPC board. They work really well and interrupts aren’t even needed! I’ll probably do a post on this later. Let me know if you can’t figure it out before I post on it and I’ll send you some code snippets.

    • Jai
    • May 20th, 2014

    hey nic, is your quadcopter ready ????
    m also making the same project and will start very soon. First i have to make the frame. Could u please tell me that which transmitter shall i go for??? I don’t have any idea about transmitters.

    • Unfortunately my quadcopter project has been shelfed. It is all built, but I need to finish up the code. School + Work + Family = no time for programmable toys 😦

      I’m using the Spektrum DX5. I would recommend any Spektrum transmitter and receiver. They’ve worked very well for me in the past. I also have an airplane I fly with the DX5. It’s nice having a toy that works out of the box!

      I would buy as expensive of a Spektrum transmitter/receiver pair as you can afford. The DX5 is really cheap but not very configurable. The other extreme is the DX8 which holds custom configurations for numerous toys that can be recalled using the LCD menu. You really only need 4 channels for a quadcopter, but there’s always fun things to do with more channels like mode selection, lights, cameras, missiles, bombs, etc.

      Have fun! Keep me informed on your progress. Maybe your success will get me back working on mine 🙂

    • Talha pervez
    • October 10th, 2015

    Can u plz tell which software i can use to make all the measurements and proportions nicely. Btw ur project looks nice.

      • nic
      • November 30th, 2015

      If you were a real mechanical engineer you’d use autocad or solidworks, but I used a good old measuring tape and pencil!!

  1. No trackbacks yet.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: