Introduction: Bicopter A2212 1400Kv Motors Multiwii Fc

Picture of Bicopter A2212 1400Kv Motors Multiwii Fc

Experimental Bicopter. based on a multiwii fight control board and two motors/propeller's each able to swivel and controlled by servo's.

For this build I am using one of my favourite construction method.

1) I draw out the parts on a Raspberry PI using Qcad.

2) I print out the drawings at a scale of 1:1.

3) I stick the paper onto 3mm plywood using "prittstick". and double up as necessary (front and back, two sides)

4) Cut out the bits on my fret saw.

5) Stick all the bits together using Gorilla glue.

the last picture shows the model assembled but not glued, with this type of construction you should be able to make the model lock and hold itself together, and before I glue the model I always fit all the pieces together to make sure they are going to fit.

I just added a couple of screen shots from the raspberry PI qcad program (LibreCAD).

To add qcad open terminal and type

sudo apt-get install qcad

And if you are going to print then you will also need to install cups

open terminal and type

sudo apt-get install cups

Step 1: Putting the Bits Together.

Picture of Putting the Bits Together.

So after cutting all the bits out and checking they fit I then in one go glued all the bits together using gorilla glue which I find very good. It only takes about 2 hours maximum to fully harden and foams up to fill small gaps.

With the main structure glued I then added pieces of 3 mm ply to the back sections where the servos will be mounted.

At the same time I glued threaded inserts into the ends of 12mm carbon tube. this design allows the tube to turn by twisting on bolts screwed into the threaded inserts in the end of the tube (the picture shows it better than I can explain it) you can also see in the picture the motor mounts pushed onto the end of the tube. at the outer end of the tube I have used a piece of fibre glass sheet to support the end.

The servos are also screwed into place. the servo used have metal gears which is a must otherwise as soon as you have a crash landing the gears will strip.

Step 2: Mounting the Hardwear.

Picture of Mounting the Hardwear.

Now its time to get all the bits into place.

To allow the motors to move they need a connecting arm, which is also made out of fibreglass and screwed to the motor mount using two of the motor screws. at the end of the arm is a pushrod fitting which holds the push rod into place using a grub screw.

The flight control board is mounted at the top, I choose this place as its near to the roll centre of the propellers. the holes to mount the board were drilled before the ply was glued into place.

With the motors and flight control board in place you can now put all the wires in and tidy up the spare cable by wrapping it around the frame.

The receiver is fitted into place using tie wraps and the connections made to the flight control board.

At this stage I decided to put Velcro strips down the side to fix the top section to the bottom where the battery is held.

Step 3: Completed Model.

Picture of Completed Model.

So the last bits are the propellers and links from the servos to the motor mounts.

the propellers are 9" *4.7 and obviously counter rotating.

Its very important to get the neutral values correct so make sure you hold the model upright and securely and then connect the horns to the servos at as near to horizontal as you can. then adjust the link to make the motors mounts exactly horizontal and exactly the same as each other. You can see in one of the pictures that there is a connector in the connecting rod to allow for adjustment, this connector is just a terminal block as removed from a terminal strip. (we call it choc block in UK but I am not sure if that will translate!)

I put the model on the scales and it weighs in at 923g (including the battery but not the propellers)

Step 4: Flying the Model

This model is very hard to fly! it needed a lot of adjustments just to allow it to fly (PID's) And it will only fly in the calmest of conditions, as soon as the wind pushes against the model it runs out of servo travel and stops yaw control.

still it is fun and its quite an achievement getting something you made to fly!

Comments

AryaVenugopalan (author)2016-09-10

It has a lot of vibration, but if you think about it it has many practical applications especially because for one quadcopter, you get 2 of these

mr_fid (author)AryaVenugopalan2016-09-10

If i had progressed this i would have spent a lot of time sorting out the PID settings to correct the vibrations. i may have another go with this next year!

AndreMendes (author)mr_fid2016-10-05

Sounds good for me, I wanna see this with out vibration in fly, I like so much, mainly design.

mr_fid (author)2016-01-16

Hello One of the picture above has the numbers written on the back.

so if you are using a multiwii with 328p then...

on one side you have 11 = Servo and 9 = ESC

and on other side you have 3 = servo and 10 = ESC.

In the picture above the model is on the floor face down (so you can see the back) and the arrow on the flight board will be facing into the floor. The other channels just connect to your receiver .

Regards

ulisses (author)2016-01-15

can put wiring diagram on the board Crius?

osterac (author)2015-10-15

I love this! Some suggestions: the shape of the craft looks like it would catch a lot of wind. If it could be constructed more like a traditional multirotor in a horizontal configuration, only with the battery hanging down to stabilize it, perhaps it would catch less wind. I'm not familiar with the multiwii flight controller, my personal preference is the KK2. I have one in my tricopter and I do remember seeing a dual rotor setting in the menu.

osterac (author)osterac2015-10-15

You also might want to experiment with how much tilt you put on the the rotors if you haven't already- you can always use more lift.

osterac (author)osterac2015-10-15

Come to think of it, why are they tilted? Seems like it would create turbulence.

mr_fid (author)osterac2015-10-15

Lots of comments! like this model it needs lots of trial and experiment's!

Why are they tilted???? well the original idea was for this model to be a backpack for an action man toy, so the shape was chosen to match the action mans shoulders! it turned out that the action man had eaten too many pies and was far to heavy to lift!!!

derekiswise (author)mr_fid2015-12-15

You might be able to vacuum mold a copy of the figure if you have access to a vacuum-form machine. Then you would have a very light copy of the figure that them could be carried by the bi-copter.

If you don't have a vacuum-form machine there are some very good designs online to make one using an adapter that fits onto a normal vacuums hose, and uses some relatively cheap ni-chrome wire to heat the plastic.

Best of luck, Thanks for the intructables.

mr_fid (author)derekiswise2015-12-15

Good idea. If I was to continue this model, I would have made the man out of balsa wood with carbon fibre arms and legs. only problem is the kids... if they see me attacking action man with a hacksaw they wont be impressed. I just finished a Tricopter for the action man and that works well, with a few improvements it should fly very nicely with a full weight action man. Check out my "Action Man Tricopter"

derekiswise (author)mr_fid2015-12-15

That's awesome! I will check it out!

osterac (author)mr_fid2015-10-18

Haha, that's a fun idea. I can't imagine a bicopter carrying much cargo though- it's like a table with two legs and it's pretty impressive you got it flying in the first place. A tricopter will give you much more lift and stability but may look strange as a backpack.

BeeblebroxTheHalfth (author)2015-10-25

Awesome job! Could the wobbling be due to the servos not being strong enough or the tilt joints not being stiff enough? I had the same problem with a tricopter I built. In my case, I used a tiny 9g servo to control the tilt system and even though the control gains were set correctly, it used to wobble and go crazy cause the servo couldn't keep up. But I may be wrong cause your servos look beefy enough and the twitching is due to it being hard to control.

Also the angled motors might work better if the were angled the other way, as in the thrust being directed outward rather than inward as the airflow would be cleaner and less turbulent. Same reason why a quadcopter descends unsteadily when descending straight down, as it's entering the 'dirty' air.

molvikpm (author)2015-10-17

If you set the D and I gains to zero do your oscillations eventually damp out at zero stick input, or does your model become unstable with increasing oscillation?

mr_fid (author)molvikpm2015-10-19

Its hard to say, but I did spend a lot of time getting to this stage. And now its flying I can try a small adjustment each time and see if it improves, however its fun to fly it and ignore the oscillations!

molvikpm (author)mr_fid2015-10-19

I'm guessing (an educated one) that the answer to my question is that it would become unstable. If you haven't already, you can try implementing a moving average window to both moderate and slightly delay the system response to the error input. The window size will be a trade off between crisp input response and stability. You might also try to program in a small dead band around zero error to minimize hunting. You will get more drift with a dead band, but it should eliminate the hunting if you get it right. Dead banding will also help reduce the effects of mechanical hysteresis (slop) in your bearing/linkage system.

wordpool (author)2015-10-17

HI,

I wonder if removing the PID would actually make things easier? Unless you need to run on autopilot it may be that your eyes and brain can make a better job of stabilising flight than the software. Apart from anything else it would confirm that the chatter is/is not from untuned PID rather than any other source.
My very limited experience in other fields taught me that two sets of inteligence (you and the PID) can fight forever. It looks like the servos are powerful compared to the masses they are trying to move, so overshoot should be minimal and correcting for long term drift is what your RC is for.

mr_fid (author)wordpool2015-10-19

This model really was very hard to get flying, and was very nearly binned as I continually crashed it. What you have said is not a bad suggestion. And in the video you will see me test launching the model and adjusting the trims so I can feel its level, so there is a lot to be said you human control.

Tim P (author)2015-10-15

awesome, one of the blades were just a little wobbly, but besides that,
AMAZING. I'm into remote controlled helicopters (bought online), so
seeing this is awesome, since you made and engineered this yourself!

one question though, how much did it cost for you to make this?

mr_fid (author)Tim P2015-10-16

The two motors, speed controllers and propellers would be £18 the controller £12 and the wood I had, so not a lot. For reference the amount of wood is 2 pages of A4 paper (because that's what I print out on!)

Tim P (author)mr_fid2015-10-18

ok cool, 50$ in supplies isn't that bad, even if i didn't have a good helicopter controller already

Edgar (author)2015-10-17

Neat! Won't some counterweights to the front and back, 90 º to the main structural plate, help balance it?

DejayRezme (author)2015-10-17

This is pretty cool. I guess if the rotors could swivel 90 degrees you could actually do some kind of vertical liftoff / aerodynamic flight plane?

Georgeh3 (author)2015-10-15

Hahaha, pretty impressed, not much like many of the other Bi-copters I have seen. So why did you decide to tilt the motors and what are your PID settings?

mr_fid made it! (author)Georgeh32015-10-15

The original idea was for this model to be a backpack for an action man toy, so the shape was chosen to match the action mans shoulders! it turned out that the action man had eaten too many pies and was far to heavy to lift!!!

Please see below for GUI picture of settings

landimatteo (author)mr_fid2015-10-16

I'm almost sure that replacing that D_yaw=0 with something higher can help.
I
don't know your bicopter inertias, then I cannot provide you with the
results of a decent simulation, but I just tried a very rough and
simplified simulation (I guessed something about 1.4e-4 [kg*m^2]). with
D_yaw =0 the step response oscillates almost forever, with just 0.001
the settling time is about 1s, with 0.01 it settles in 0.15s, with 0.15
settles in less than 0.1s, with 0.16 settles in less than 0.06s

Georgeh3 (author)mr_fid2015-10-15

Hahahaha, just my luck xD What about a quadcopter pack that folds up on two servos so that this action guy thing hangs down? Also you could probably get some airplane motors that would be plenty powerful for most things, though they mat be more expensive haha. Thanks for the info.

mr_fid made it! (author)2015-10-16

Dear All please find attached a PDF drawing of the model. It should be made from 3mm ply and doubled up. Also to aid printing there is a 1cm scale along the side, if when you print this it srinks then you will need to play with settings.

james.dougal.549 (author)2015-10-15

nice build I noticed your getting a lot of servo chatter you need to get heavy duty servos to stop the chatter will make it smoother :)

mr_fid (author)james.dougal.5492015-10-16

At first I used slower servos and they were better!

GaryK18 (author)2015-10-15

Hi

Have you balanced/aligned the blades?

mr_fid (author)GaryK182015-10-16

I once spent a long time balancing all my propellers, then promptly ploughed in into the ground. so generally as long as they don't shake to much I don't bother. in the case of this model the servos are more of the problem

RichardBronosky (author)2015-10-16

Great work! ? It's funny to hear all these people talk about weight shift and pendulum as a design change. Do they not realize that this thing already IS a pendulum and when you rotate the servos to the same side it IS weight shift. By using the 2 servos in a configuration that allows rotating to opposite sides you gain yaw control. Otherwise you'd have to yaw via speed differential which in turn would cause thrust differential, effecting balance, requiring complex compensation. Good luck with that, Armchair Engineers.

mr_fid (author)RichardBronosky2015-10-16

Yes at first I thought it would be a nice idea to have the battery moved by servos to control the direction, then I realised the Flight control board would need a didn't firmware to work. At present up/down = ESC power, left/right = ESC power (imbalance) forward/backwards = both servos move together, yaw = both servos move opposite. For a battery moving craft it would be, up/down = ESC power, left/right = servo, forward/backwards = servo, yaw = esc imbalance. so you would have to write your own "MIX TABLE" if using Multiwii.

TehseenHasan56 (author)2015-10-16

great

MattTuck (author)2015-10-15

Have you considered using a gyro or eccentric rotating mass for forward/backward motion?

MattTuck (author)MattTuck2015-10-15

You could then change to sturdier mounts for your props and get rid of at least SOME of your vibration.

mr_fid (author)MattTuck2015-10-15

I think adding more mass isn't really an option as I think its heavy for the amount of lift. And I think the vibration problem is mainly the PID settings causing the servos to overreact. It needs a lot of work to "dial in" and the nights are drawing in!

Georgeh3 (author)mr_fid2015-10-15

Looks like P gain, may counter act it with the derivative though.

MattTuck (author)Georgeh32015-10-15

P being momentum?

Georgeh3 (author)MattTuck2015-10-15

P being porportinal, it causes higher speed ocilations like in the video when it is too high. Cause the servo to overshoot, and then backtrack to make up for overshooting it, keeps repeating.

MattTuck (author)mr_fid2015-10-15

I'm talking about removing the servos and adding a similar mass...

So it moves fwd/bcw via a pendulum effect caused by rotating a small mass at the bottom.

So long as there is sufficient torque it should cause the bicopter to pivot thus changing the thrust vector providing fwd/bcw motion.

Mass+new motor~2*servo

Georgeh3 (author)MattTuck2015-10-15

Or use some CF props, the plastic creates a lot of vibrations that can confuse the FC due to the vibrations.

mohamedb9 (author)2015-10-15

good project but it misses the technical names off used parts such as the flying control board,etc and the source code which is paramount

mr_fid (author)mohamedb92015-10-15

The flight control board is a Crius MWC Multiwii SE V2.5

It is loaded with Multiwii version 2.4

The Flight control board has the following.

ATMega 328P Microcontroller
ITG3205 3-axis MEMS gyro
MPU6050 6 axis gyro/accel with Motion Processing Unit
HMC5883L 3-axis digital magnetometer
BMP085 digital pressure sensor

I hope that helps.

mohamedb9 (author)mr_fid2015-10-15

Thank you for you reply,

It seems that you use a preloaded program.

As beginner, I prefer open source solutions to better understand the behaviour of each component.

Cordially

mr_fid (author)mohamedb92015-10-15

Multiwii is open source. you use the Arduino programing environment to configure the machine and load the firmware.

sbkenn (author)2015-10-15

My personal preference would be for concentric contra-props/rotors. Control would be quite similar, I think.

About This Instructable

80,983views

501favorites

License:

Bio: I love making things. I have for as long as I can remember liked to make stuff. Now days I have two kids (Thomas and ... More »
More by mr_fid:Radio Controlled Electric Hovercraft  Release Hook for Quadcopter (The Flying Beaver Called Pig!)Simple Quadcopter, Basic Tools Only.
Add instructable to: