DIY Quadcopter For Around $200

We think [FlorianH] did a bang-up job of prototyping his Minima Quadcopter on the cheap. The total bill comes in right around $200 and we’re very happy with the quality of parts as well as the results.

Here you can see the top of the double-sided board which he etched to host all of the components. At each corner there is a power MOSFET which drives the motor. At first glance we thought that the Xbee module was acting as the radio control and processer as well. But on the underside you’ll find an ATmega32 which is responsible for reading the Gyroscope sensor and Accelerometer, processing these signals and driving each MOSFET via PWM lines to provide stability.

You can see some flight tests after the break. [FlorianH] mentions that there is some oscillation in the feedback loop when both the gyro and accelerometer are used. But cut the accelerometer out of the equation and the platform is rock-solid.

This build uses carbon tubes to mount the motors, which we think will be a little more robust than the all-PCB designs are.

[youtube=http://www.youtube.com/watch?v=s5ALTDcdjEM&w=470]

28 thoughts on “DIY Quadcopter For Around $200

  1. I must say that I am always amazed at what a person can come up with. I have a wish that I could be as able as most of the people that post their designs on this site. This is a great example of hard work and it is a great design. Living here in Las Vegas it doesn’t appear that there are any people or groups of people interested in anything along this line or any thing like this. If there is, I would like to know where they are because I could sure use some schooling to do some of the things I want to do but haven’t the knowledge. Great project.

    1. You have one of the greatest stores ( in my opinion ) in Vegas, Fry’s… I would be surprised a store with as large of a hobby electronics section as they have, wouldn’t want to help organize or host a meetup.

      In my area ( Orlando ) we have a small store called Skycraft, though I would really like to have more options than just that one store. I have heard there is a hackerspace in my area, but I have yet to get out there and join ( must kick myself in the boot to get on that )

      As for learning, you just need to pick a project, then break it into small chunks and chip away at it. You learn one thing, then move on to the next. Over timer you will acquire a better understanding of electronics and the processes that help you to complete your project. It does take time, but you will be so happy that you spent the time doing it.

      1. @M4CGYV3R – yes :) That place IS awesome, but I do like variety! It is about five minutes from my house so I spend a ton of time ( and money ) there.

        By the way, I believe those are actually jet fighter fuel tanks, not rockets… though, they do have rockets hanging from the ceiling inside!

      2. @mess_maker Skycraft is the only reason my friends and I made it through school. No matter what project we came up with they had the parts.

        @kevin there is some interest in electronics and robotics out here.
        /usr/lib is a library/hackerspace on Fremont street.
        Check out Pololu robotics has a robotics/Ruby hacknight, Thursdays.

  2. Great job! Lots of work there.

    In my experience, soldering those pluggable jumpers you can buy from Makershed and Adafruit is a bad idea. The wire is already loosely soldered to the pin under the overmolding, and when soldering the pin to something else 50% of the time the wire comes loose (open) underneath the overmolding… and you are left scratching your head for a little bit while your circuit is borked ;-)

  3. If only the propeller arms would fold in so I could carry a swarm around in my pocket.

    Nothing like a swarm of quad-rotors in a muggers face to deter him from crime and make him decide to go into engineering instead.

  4. I’m going to try to make something similar under 100€.
    2€ atmega8a for ESC
    06N03 from scrap
    28€ Zigbit (without the ZigBee crap layer)
    25€Wii Motion Plus
    4×4€ SPEED400 motors
    ….

    I just hope the SPEED400s will be sufficient to lift it off.

  5. This is a pretty sweet project! Even though the HaD summary missed that it uses brushed motors instead of brushless.

    “Mainboard features:


    Power MOSFET brushed motor drivers”

    Adding brushless motors would be a good next step.

  6. Ok so I just got a teensy in the mail today….. Looks like I am going to be making a quad copter teensy shield now lol. Very nice job! It’s a simpler version of the one posted up the other day. I have a bma180 3 axis accelerometer just need an xbee or bt module motor drivers and gyro.

  7. Hey Guys,

    thanks for all the kind remarks.

    @andrew: In gyro-only mode there is no gyro drift. The gyro only corrects any detected involuntary rotation directy. The error isn’t accumulated. It’s like flying a helicopter. The pilot has to level it out.

    Yes, the quad uses brushed motors. Not brushless. That would have required a lot more parts. That’s another build :)

    P.S.: a set of brushless motors is already in the mail :)

    1. Nice build. I really liked the idea of using just brushed motors despite the fact that it’s a bad idea.

      I want to ask some thinks. Are there really just two perpendicular PD regulators? So you take (wanted angular rate – real angular rate) and angular acceleration in one axis, multiply them with P and D constants and calculate needed PWM difference between the two motors on that axis?
      Just that is needed to make it fly? (If so, I’m amazed :) I mean really..)

      Are you experiencing any latency problems with Xbee since you are using it for real manual flight (like heli you said)? I’ve been always discouraged from using Xbee/Zigbee and pretty much everything with higher OSI layers so I’m experimenting with bare 802.15.4 PHY in Atmel’s AT86RF230/212. You are the first of anyone who I can think of who managed using Xbee for real manual control.

      1. Hey Grawp,
        yes and no. There is a PD controller at the end of the control loop. And it’s doing most of the work. The trick is and it’s the center of most quad-copter discussions how to find the correct angles since both sensors have their problems. My current auto-leveling loop doesn’t do any of those clever things (Kalman,DCM), yet. It just accumulates the gyro angles and weakly blends in the accelerometer for P and uses the gyro for D directly. You can browse through my code on github to see what I mean. The gyro-only loop is even simpler it takes the desired angle (joysick position), sticks it directly into P and the gyro directly into D.

        Regarding your second question, the latency is not a problem for me. There obviously is some (not only is the communication bundled up into network packages but the XBee also waits for a specified number of character times before it sends it). But I don’t notice it when I fly the quad normally and I haven’t done any aerobatics yet. Having said that, I have had a few strange problems with the XBees which I attributed to buffer overruns. But I can’t be sure. If you are handling the protocol yourself at least you know exacly what going wrong.

    1. @uphiearl: Sorry, I only ever made that one board. No spares. But the eagle files are on the webpage.
      Don’t you already have a pretty nice controller board in your mQx ? Why would you want to replace that?

  8. My first thought was my inner cheapskate screaming that the gyro and radio components stand out as a bit pricey. After a bit of googling around, it seems that MEMS gyros are indeed expensive, but you could save tens of bucks by doing what that picopter dude did and just use an ATmega with built-in RF functionality. The ATmega128RFA1 goes for about eight bucks, which saves you around $20 for each the controller and the copter.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.