After weeks of waiting, my QuadCopter arrived! Here are some hints and notes on the assembly process and photos of the result!
Firstly, Royal Mails Parcel Force was outdone by generic bog standard Hong Kong post. I was able to track the package all the way till ParcelForce obtained possession, at which point they claimed they could only track delivery once it had been signed for by myself.
So I unboxed the crazyflie, and marvelled at its size.
The motor mount assembly went along as directed, and I managed to get them attached with little difficulty.
I did have issues soldering, these are very small contacts and wires, and the go to people for soldering irons simply didn’t have one small enough, though I expected this might be the case. Here’s a filtered image with a 20 pence coin in the background for scale
After a trip to Southport, we soldered the contacts for each motor. Starting on Motor 4 and working are way towards the centre of the board we initially tried the suggested method of soldering on the top and inserting from the bottom. For Motors 3, 2, and 1 we pre-soldered the wires, inserted through the board, bent the wire with a screwdriver so it held in place, then soldered to the board.
This worked quite well, through we did create a short on Motor 2, but that was easily fixed. We tested by applying a multimeter to the soldered contacts, with readings of 2.1-2.6 Ohms. The assembly instructions were particularly helpful suggesting around 2 Ohms was an expected reading.
Finally, we attached the battery and the propellers, and setup the PC. The assembly instructions take great care to indicate clockwise versus anti clockwise propeller blades, but what they don’t mention, is that the anticlockwise propeller blades have the letter A inscribed on them. So there’s no need to squint at the curves to figure it out, just look for an A, if it’s there it’s anti-clockwise, if it isn’t, it’s clockwise.
For the controls, I bought an XBox 360 controller and wireless receiver specifically for this, testing on my home machine. I’ve had more experience with a PS3 controller in the past ( I worked at SCEE for a year ), and I’m mainly a PC gamer, so 360 pads are not so familiar.
Powering up the quadcopter, the initial tests ran and passed, and we successfully connected via the radio. Unplugging the quadcopter, we attempted a first test flight. The motors span as expected, but little thrust was generated.
It turns out we had placed the propellers upside down. After 30 minutes of scratching our heads, and another 15 minutes re-charging the battery on the theory it was too low on power to lift, we looked at the propeller blades.
It turns out that the central axis has a smooth and an etched side., the smooth side goes on the bottom closest to the motor, and the etched side goes on top. Swapping these fixed the problem.
There was also the issue that despite having a controller, there was no indication of what mapped on to what. We figured out one analogue stick influence thrust, and the other had an impact on direction.
Our initial tests proved tricky, To generate thrust I had to push forwards on the control stick, but as well as upwards motion, this also generated forward motion. It’s possibly a problem of weight distribution, but I’d like to set up a better test environment and practice. The instructions warn that while a 360 control works out of the box, it’s will be sensitive. Numerous people have said to apply thrust and tell it to move backwards at the same time, which seems pretty obvious, but the result was a fast veer backwards into a wall.
The device mappings in the control program also suggest a kill switch, but I’ve yet to determine which button on the 360 controller it’s mapped on to. It would be useful if the device mapping screen showed which buttons were mapped on to which function, alongside a light up key so that the controller can be pressed to see which physical button maps onto which virtual button/function
Flight Tests 2
Taking the quadcopter home I decided to reseat the battery and re-calibrate the 360 controller. This made a world of difference, the control mappings changed and the analog sticks made a lot more sense, though I haven’t figured out them all.
With a little practice I can make it lift off the ground without zooming off into a wall. Directing it in one direction is a little tricky as you have to figure out which way is forward and lifting too high will send it headed towards a wall. The wiki suggests adding coloured tape around Motor 1s arm to help indicate forward direction. Nonetheless, I can make it hover at ground level, move one way, then go back the way it came. Hovering on the spot is my goal but it’s proving difficult
One major difficulty I ran into was installing the crazy radio dongle driver on Windows 8. Unlike Windows 7, 8 saw that the driver was not digitally signed and point blank refused to install it. I had to reboot into troubleshooting mode and disable digital driver signature enforcement before I could continue.
Overall I’m pretty happy with what I’ve got. I could use some more practice, and it’d be great to have a bigger battery capacity, and one day I might look into camera control.