Building a Quadcopter

A rotary wing aircraft that I can fly outdoors has been something that I’ve been wanting to try for a little while, I even priced up the required items to build a quadcopter on hobbyking, but then procrastinated about buying it until I spoke with Scott who fancied building one too. So I ordered 2 full sets of kit so we could both build one each.

The shopping list I had for each copter was:

1 x Q450 V3 fibre glass frame

1 x Hobbyking multi-rotor controller V3.0

1 x QBrain 4 x 25amp brushless ESC

4 x Turnigy 1000kv D2830-11 brushless motors

1 pack of 4 10×4.5 slow fly props clockwise

1 pack of 4 10×4.5 slow fly props counter clockwise

 

It all arrived, so I built up the frame putting a drop of thread lock on each of the screws, and attached the 4 motors to the end of each arm, again with thread lock. I connected each motor to the speed controller, and just to check everything worked plugged it into the controller board and receiver and powered up… Nothing happened… So I thought I’d better read the instructions. First the speed controller required calibrating to the radio, setting the top and bottom of the throttle throw. So I did this. The instructions for the hobbyking controller board were ‘sparse’ to say the least, I eventually found by trial and error that setting the throttle to lowest and the rudder full left seemed to initialise the controller board, then every thing sprang to life. I now know this is ‘disarming’ the nav board… As opposed to arming which is done by moving the rudder to the right. Still none the wiser as to what that is if anyone knows?

Of course the first thing I had to sort now was the order of the motors, this was actually in the hobbyking instructions, but the firmware which shipped with the board was for a quad with a motor in front, behind, left and right. I wanted an ‘X’ configuration with a motor front-right, front-left, back-right and back-left. Fortunately I had a USBasp connector to update the firmware on the board, so when I eventually found the firmware update utility I update this to the X configuration from the hobbyking site, change the order of each motor to its connection on the controller board, powered up, put in left rudder to start the board and then a little throttle. This time all the motors appeared to start, and adding aileron and elevator appeared to speed up and slow down the motors as I would expect.  All is going well, next was to check the motors are all spinning in the correct direction and with a helpful little image in the instructions this wasn’t too much of an issue by placing a finger on the side of the motor case to see which way my finger was pushed, then reversing 2 wires if it was spinning the wrong way.

black_x

 

So next is attaching all the bits to the frame. Reading advice from other copter builders all seem to say that the controller board needs to be damped from vibration, my setup with the single ESC would seem to be advantageous in this situation since I could attach the controller board to the top of the ESC case using foam tape, then the ESC to the frame using foam tape the weight of the ESC should then help to damp the controller board from the vibrations in the frame. Although the heat management may be affected by attaching this to the ESC. We shall find out the effect soon anyway.

So after tidying up all the cables and tying them down this is what my copter looks like:

20141106_222311

 

At this point I got a little cocky and decided to spin it up with the props attached to see if it will shuffle around the floor in the correct directions, yaw the right way, etc. So I put a little bit of throttle on, not enough to take off, but enough to start taking the weight, then just touched the rudder slightly and the copter launched and smashed into the TV. After I extracted it, I removed all the props again (replaced the broken ones) and turned all the giros to minimum, and reduced the rates on my transmitter to 50% (with the dual rates switch set so I can increase this if needed) and put some expo in too. Strangely the pots on the board seem to need to be turned clockwise to reduce the rates. Turning them this way gave the lowest change in motor speeds for any given input, this way seemed a little counter intuitive to me. But I now felt confident again and being even more careful now I put the props back on and tried again, this time success, although there is a lot of effect from the walls and floor generally it seemed to shuffle in the way I expected.

All put together, this is now how it all looks:

20141107_072711

I still need to put a bit of tape on the receiver antennas, but next stop… a field test.

At the field we noticed that the single cable that I’d seen used elsewhere actually cross connected the Ailerons and Elevators, although we had set everything up so that the motors were working correctly of course the controller board didn’t have the motors set up correctly, so the first test flight was extremely unstable, the giros sending the copter more and more out of control. This was corrected, we then needed to add a significant amount of trim on various axis, but after this the copter took off and was controllable, but still a little wobbly. We turned up the gain on the giros on the controller board, this improved stability, this is the resulting flight by Mr Scott:

http://www.youtube.com/watch?v=s-FutairsgQ

I then flew a little higher and further, caught a bit of wind which started it swaying, because it was a bit further away it was difficult to tell which way it was swaying. I should have simply left it to the giros to sort itself out, but instead I lost my cool and knocked the power down which then made things worse and ultimately led to it crashing

20141109_130730

But I do take a lot of positives from the days testing. We managed to get some genuine flying time, the copter performed well and we did get a lot of information that we can take forwards and improve them and do better next time. And Scott managed to do all this without smashing his up which is even more of a plus so he can take it forwards without too much work.

Good fun… needs a bit of work  🙂

Update 30/11/2014

I mounted the FPV and OSD removed from the catalina when I updated to the Ardupilot, the following was flown Line Of Site, but recorded from the Ground Station

https://www.youtube.com/watch?v=kcGDjoquTRs

 

24 thoughts on “Building a Quadcopter”

  1. Thanks for all your help with the project Mr Chris, it’s great fun, gutted yours got broke, but at least like you stay we are still moving forward with it.
    Hope others will give it a try too 🙂

    1. It was great fun, thanks for joining in and pushing me into actually getting on with it.
      The updated firmware I was talking about, this is the forum thread:
      http://www.rcgroups.com/forums/showthread.php?t=1458663&page=5#post18609175
      if you move on a few posts there are some videos (http://www.rcgroups.com/forums/showthread.php?t=1458663&page=10), I’m sure they are better fliers than we are, but they look a huge amount more stable and flyable than ours were. So I’m really hopeful this will improve them again.

  2. Chris and Scott
    I have been trying to resist but it all looks too exciting.
    Would you still get the same things ? Do you thing it will be able to lift a go-pro ?
    Is there anything I get for you from hobbyking while I am ordering ?
    Robert

    1. Hello Mr Robert,
      They will easily lift a go-pro, I think it takes just about half power to lift off, if you are there this weekend I should certainly have the new firmware on mine and a new frame to give it a try
      The original frame that we ordered is now out of stock on Hobbyking, so I’ve order the 600mm aluminium one
      https://www.hobbyking.com/hobbyking/store/__42933__Hobbyking_X525_V3_Glass_Fiber_Quadcopter_Frame_600mm_UK_Warehouse_.html
      And you will need a USBasp cable to reflash the controller board which you have to do to change it to an X layout for the motors.
      https://www.hobbyking.com/hobbyking/store/__60379__USBasp_AVR_Programming_Device_for_ATMEL_proccessors_UK_Warehouse_.html?strSearch=USBasp

      I know Scott has been reading about using higher speed motors and smaller props for better responsiveness and stability, but we haven’t really managed to have a good play with it yet.
      And thank you for the offer, I think I have ordered all the replacement bits I need now ta.

    1. The flight controller board contains 3 giros, how they are used is down to the firmware. The firmware that we had on these was quite simple, where you can simply set the gain on each giro. I’ve found new firmware which uses the output of the giros more cleverly as per more advanced quadcopter stabilisation techniques. The so called PID stabilisation algorithm (see more here if you want to know more http://blog.oscarliang.net/quadcopter-pid-explained-tuning/)

  4. Good to see you having a go too Rob, it’s a good bit of fun, and very different.

    Thanks again for all the help Mr Chris, not dared fire it up with the new firmware yet, will wait till we get to the field I think, have changed the props and given it a crash helmet ready 🙂
    Quad

      1. Thanks very much chaps, and yes it’s currently held on by 4 bands at the mo Mr Chris so easy enough to get off 🙂
        Some interesting stuff about the throttle curve on there, did you have to reverse any of the pots?

        1. I did have to do the pot reversal, and the yaw giro reversal, that is all I’ve done and it hovers in the back garden, I’ve printed that page and will take it to the field tomorrow, where I will do the throttle calibration, and see how the instructions go for setting the pots.

          1. Just had a go myself, done the pot and yaw giro reversal, then calibrated the throttle as only two motors were powering up, all 4 power up ok now, can’t wait to try it tomorrow 🙂

  7. Chris and/or Scott
    Help.
    I am having trouble arming my controller board. Would one of you be able to plug same into one of your rxs for me to check it out ?
    Ta
    Robert

    1. I don’t think I’m going tonight, but Sunday we can try a receiver on it (weather permitting).
      Have you got your rates on full? and have you tried the rudder full left AND full right for 5 seconds each? Spektrum receivers needed the rudder reversing, so I first needed to push it full left to arm it before I reversed it.

      1. Changing the subject slightly from arming boards.
        I also seem to have a HK MultiWii pro board. It arms OK. When I look in (pardon me when I use the wrong words) the sketch files one of the lines seems to imply that, once armed, the propellors are always going round a bit. Is this right ?
        Playing with these files call to mind an old programming maxim we used to live by “if in doubt – rem it out”

        1. That’s good that you have it arming, but ours don’t spin the props when they are armed. Do you need to configure the throttle min and max positions? On ours you set the yaw pot to 0%, throttle the transmitter to full, power on the receiver and controller board, when the controller flashes a couple of times move the throttle to min. But of course I don’t know how to do it on your board

          1. Wasn’t allowed out this morning. Spent some time re-doing what I did to the no.1 controller board. Must have done something different but I don’t know what. Arms beautifully now !
            So now I have to decide which board to use in the quadcopter. Perhaps I could put them both in and switch between them in flight. The Taranis is very clever ! Perhaps not.

  8. Found this. It answers one of my questions.

    Starting the multicopter
    The engine launch is done by tilting the yaw stick right while having the throttle stick in minimum position.
    (An alternative method was also introduced recently: roll stick right while having the throttle stick in minimum position.)
    For security reasons, the throttle stick must be set to minimum.
    Now motors turn at an idle rate and the quadcopter is ready for flight.