Build Log


This page will describe some of the building process. For now, it is just blog format. More details will be added later.

Update: November




Update: August


We posted an aerial video of campus from the hexacopter!



Here's a picture of the hexacopter used for this shoot. It was a dual operator setup, the first operator (Christopher Vo) piloted the hexacopter, and second operator (Ted Markson) controlled the gimbal to compose the shot.

image

Update: April


We are still working on the image processing part of the project, but here's some PBS coverage of me:



I flew an autonomous mission at that event, here's the photo:

path

Update: December


We have put together a kit for the quad copter which costs $280 including 5000 mAh battery, gps, telemetry, frame, motors, and ESCs. This quad copter is 450mm wide, 800g without battery, 1200g with battery, and can handle an additional payload of at least 1200g for at least 25 minutes. For weight reference, a GoPro Hero 3 with waterproof case weighs 200g, and a Gumstix Overo with Summit Board and Caspa VL camera weighs less than 50g.

Here are some pictures:

quadcopter
quadcopter
quadcopter

Here is a video of it flying around indoors (using gyro/accelerometer only, no GPS or sonar). Altitude hold with sonar works well too, but most of the engineering building is covered in occupancy sensors based on ultrasound, which limits our ability to use sonar for position hold in the engineering building.



Update: 30 November 2012


We have uploaded some aerial footage from the hexacopter testing some altitude and GPS waypoint flights:



The recorded path is here:

path

Update: 15 November 2012


We have developed a quadcopter (4 rotors) for our indoor testing. Here are some pictures:

quad
quad

Update: 02 October 2012


We have performed several test flights with the fully assembled hexacopter and the platform is stable, including GPS location, loiter and RTL modes. See video:



We are working on testing a camera/micro-controller setup communicating to the APM 2 via SPI.

Update: 20 September 2012


Proper control of the hexacopter requires that the 6 motors are installed with the proper orientation.

Motor Orientation

I assembled and tested the spin direction for each of the motors, and then did a quick test on one of the motors with 8" prop and bench 11.1V power supply to just get an idea for the thrust the UAV is capable of producing. Of course at 11.1V it will not be exactly the same as the final version. The setup for this test is motors attached to 30A OPTO ESC. ESC attached to bench power supply and Spektrum AR8000 Receiver. AR8000 Receiver getting commands from the throttle stick of Spektrum DX7s controller.



Update: 17 September 2012


We received and tested the APM 2.

APM 2

The first step was to solder the header pins on. I chose vertical orientation for all pins.

For testing, we configured the CC BEC using the CastleLink programmer and software, for the default value (5.1V). Then I connected it to a 12V bench power supply and verified that 5.1V were output. It is connected to input 8 on the APM2 as shown in the photo.

I disconnected the BEC temporarily while we connected the APM2 via USB and used the Mission Planner Windows software to install the firmware for a 6-rotor hexacopter.

I also connected the 3DR radio and used the instructions here to configure it on the Mission Planner software.

I also connected the Spektrum AR8000 RC receiver (by connecting 2 power leads from the 5V rail on the APM 2 and 6 signal wires from the AR8000 to the appropriate input channels on the APM2). I selected the 5 default inputs (AILERON, ELEV, THR, RUDD, AUX1) and also selected an additional AUX2 channel to use. I propose to use AUX2 to either control camera tilt centering or something else in the future. I could also hook it up to the GEAR channel to add a remote switch.

Update: 10 September 2012


We received and assembled our X468 Camera Mount / Landing Gear early from RC Drones, so I put it together. The kit parts are a bit rough around the edges and requires some sanding and gluing, but it is pretty solid once constructed. I wrote a quick Arduino sketch (below) to just move the servos to test the tilt and roll of the mount with a heavy Point Grey Firefly 2. However, a more likely candidate for cameras would be the Point Grey Firefly MV which is lighter and has a USB interface.



The Arduino sketch to test the servos:
#include <Servo.h>;
Servo tilt, roll;
int pos=0;

void setup() {
    Serial.begin(9600);
    tilt.attach(3);
    roll.attach(5);
    pinMode(9, OUTPUT);
    pinMode(10, OUTPUT);
}

// the loop routine runs over and over again forever:
void loop() {
    analogWrite(9, 255);
    analogWrite(10, 255);
    for(pos = 30; pos < 150; pos += 1) {
        tilt.write(pos);
        Serial.println(pos);
        delay(15);
    }
    for(pos = 150; pos >=30; pos-=1) {
        tilt.write(pos);
        Serial.println(pos);
        delay(15);
    }
    tilt.write(30);
    for(pos = 45; pos < 75; pos += 1) {
        roll.write(pos);
        Serial.println(pos);
        delay(15);
    }
    for(pos = 75; pos >= 45; pos-=1) {
        roll.write(pos);
        Serial.println(pos);
        delay(15);
    }
    for(pos = 45; pos < 60; pos += 1) {
        roll.write(pos);
        Serial.println(pos);
        delay(15);
    }
    roll.write(60);
}


UAV Home Page
List of UAV Related Pages
Computer Science @ George Mason University