Quintcopter Plywood Design. Does It Fly?

Relativity simple multicopter, built using 3mm plywood and using the very common and cheap A2212 motor with 30 Amp brushless controller. Propellers are 1045 the flight controller is the simple KK2.1.5. Powered by a 3000mAh 3 cell lipo the whole model cost around £100.

Purely a model to address my curiosity of “why are there no Multicopters with 5 motors?”

(or put another way why you don't get quints!)

So let’s start this Instructables with the obvious question WHY? Why would you build a 5 motor drone? (or skip this section and accept it’s because “I CAN!”)

To answer this I am just going to explain how a Tricopter works and how a quad/hex/oct work. The flight controller (which is the heart of the drone) takes in the 4 channels from the receiver and works out what power to give to each motor. The transmitter is set up for a plane so the 4 channels are throttle, rudder, elevator and aileron. And the 4 channels do the following.

Throttle causes a change in altitude (up and down)

Rudder causes the model to spin on its axis (yaw)

Elevator causes the nose of the model to go up or down causing the model to fly forward or backwards.

And lastly Aileron cause the right or left side to go up or down making the model fly left or right.

SO now let’s look at the TRICOPTER. The throttle, elevator and aileron cause the 3 motors to change power and move the model. However the rudder (yaw) is connected to a servo which physically adjust the angle on one of the motors and causes the model to spin. The 3 motors can all rotate the same direction (either clockwise or anti) OR you can have two rotate one direction and the last the opposite. I prefer to do the later as I believe if you have all 3 turning in the same direction then the angle of the servo motor will be greater. Finally let’s have a look at the settings within the Flight controller. The 3 motors have different values for Pitch and Roll (or Elevator and Aileron whichever you prefer). The values are calculated by using the following equations. (Considering all motors are on the same radius)

Pitch (P) = cos(angle) * 100

Roll (R) = sin(angle) * 100

Where the angle is measured from the front going clockwise.

Take a look at one of the photos for more detail. With the flight controller it doesn’t matter which motor goes where as long as you program the correct P and R values for the position the motor is in, also each motor has a setting of 100 for the throttle. The rudder value for each motor is set to zero, that’s because the 3 motors have no effect on the yaw. Lastly we come to the servo, the servo is solely response for the rudder (yaw) and nothing else. In the settings the servo is given 100% rudder and an offset of 50. The yaw direction can be reversed by changing the sign (I.E -100%). Finally the servo can be placed on any arm as it only does the yaw so it really doesn’t matter if it’s at the front or back etc.

So now let’s look at the Quad/Hex/Oct. The first (and most obvious) thing to say is that the number of motors in each case is EVEN! And this is important but more important is that half of the motors will turn clockwise and the other half anticlockwise, and just as important is that the motors alternate around the frame. This is where the difference is between a tricopter where the yaw is controlled by the servo and a quad/hex/oct where the yaw is controlled by changing the speed of the motors. So what does that mean??? It’s very simple let’s consider a quadcopter, two motors will go clockwise and two will go anticlockwise. If you wish to make the model spin on its axis (Yaw) then you increase the speed of the clockwise motors and at the same time decrease the speed of the Anticlockwise motors, by doing this you have maintained the overall power of the motors and hence not changed altitude BUT crucially you now have a turning force because two motors are spinning quicker than the others.

So now let’s take a look at the flight controller. Its the same as the tricopter the Pitch and roll are going to be set the same depending on where each motor is, however for each motor the value of rudder is now set to 100 for the clockwise motors and -100 for the anticlockwise motors.

So where does that leave my Quint??

I have been making multicopters for a few years now, and initially i used 12mm carbon tubes and aluminium motor mounts. This made for a very light weight model, but the aluminum mounts bent and broke far to easily and the carbon would also snap on a heavy crash.

Then i tried a different approach , which was to use plywood. You can see one of my first model here which was a modular multicopter where you could add as many arms as you wanted. This turned out to be a nice strong model and i ran the model for some time just as a quad. Eventually the motors started to wear out and the wires failed! So i did a bit of a redesign of the arms and added a bit more strength.

Like most of my designs i try to make bits symmetrical and hence common. so the sides of each arm are obviously the same, but i have also made the top and bottom ply pieces the same size and symmetrical both axis. This really helps when cutting out and assembling the model as it doesn't matter which bit you pick up (top or bottom) as you know it will fit anyway around.

One of the main differences on this new design was to add an ends to the box section. at the motor end this piece is held in place by the motor mount, and when you are glueing the pieces together you push the arm onto the hub to hold all the bits together. The last bit to be glued into place is the motor mount reinforcement which i generally hold in place with a peg. The top and bottom bits can be retained whilst the glue drys with zip ties.

You should be able to see in the pictures that the motor wires and the power wires are fed through the holes in the ends of the arm, this is useful to keep things neat and held together.

My preference is to connect the motor wires directly to the ESC, i am not a fan of connectors so i get rid of them. There is one problem with this approach which is to change the direction of the motor you need to swop any two wires. But i have over come this problem by loading the ESC with SimonK then if i need to change the direction i can just change the firmware by connecting the esc to the computer using the Signal wire and ground.

The three motor wires which were connected to the esc are unsoldered and the motor wires are connected directly to the ESC, the wires which were removed are then used to lengthen the power wires and make a ring circuit where each motor connects to both its neighbors and then finally to the battery connector.

If you want to know how to load SimonK firmware into these HobbyPower 30A ESC then take a look at the video i did. LINK

This model uses the very easy KK2.1.5 board. Normally the board comes with a little LCD and 4 buttons but this version has a small programmer which you plug into the side. This just makes the controller a bit more robust!

In the pictures you can see the connections i have made to the receiver. The Receiver only needs to have one positive and one negative, so you may notice that all the colours are used (for signals) This is because the throttle is correct and has all 3 wires, the other 3 channels only use the signal wire so share one lead. In total there are 6 connections at the receiver and FC.

The esc connections are obvious, and if fact it doesn't matter which one goes where as long as the mixing table is correct for each motor. I keep it simple by starting with 1 at the front then counting around.

You may be able to tell from the pictures and the video that i really didn't think much about where the battery was going to go. So i temporarily taped it underneath the model.

And that's it. Now it's time to see if it works?

So like all of my new models i check each direction before taking off. You only need a little power and it obvious if its correct. if the movements match the transmitter then you know you have got the correct propellers on the correct motors, you have connected the correct wires both receiver and ESC's to the Flight controller, and lastly you have got the MIX correct.

Now i give it a bit more power to just get it of the ground. i only want the model to be a few inches off the ground so if something goes wrong it doesn't have far to fall!

Once i was happy i then try a little bit of movement in each way forwards, backwards, right and left. Lastly i check the yaw......

Almost as expected the yaw action has additional reactions! if you are smooth and slow with the yaw input then you cant notice the difference, however if you are aggressive with the yaw then a hard right results in a drop of the nose and a hard right result in a rise of the nose.

So i Now have 5 options...

This is the starting point. OPTION 1.

As standard i had all the rudder values set to either 100 or -100 depending on the direction of rotation. however if you think about it the clockwise motors add up to 300% and the anticlockwise add up to 200% so the first experiment was to change the values of the clockwise to 66% each so the 3 motors total 200% (the same as the anticlockwise)

This seemed so obvious when i thought of it! remove the rudder value of the front motor so it wont react to the rudder input meaning there will now only be two motors in each direction reacting to rudder.

Because i have made bicopters and a tricopter i had a couple of spare servo arms. And even more importantly i made the arms fit the same as the stationary arms. However to make this nice and easy i added a connector to the esc power wires so i could swop over the arm/motor quickly.

Now there is a disadvantage,as the servo arm is a lot heaver at 218 grams compared with 148 grams. This is partly my fault for using a high speed metal geared full sized servo, but experience has shown me that the small servos are very weak.

Also this isn't a plug and play swop as all the channels in the mixing table need to be changed and a 6th channel needs to be added for the servo. Having said that it doesn't take much more than 5 minutes to do and can be done in the field which is a real bonus of the KK2.1.5 board.

So as a test i decided to lower the throttle to 66% on the 3 motors.

SO it is possible to get a 5 motor multicopter working, and after a bit of adjustment you can get it to fly very well. However its not perfect and never will be. The simple answer is unless you are building a tricopter make sure you have an even number of motors.

As for this model... i will be making a hex hub and converting the quint to a hex!

And lastly as a celebration of the wonderful KK2.1.5 board i cut of two of the arms and made it into a tricopter! all of the changes were done at the field with no need for a laptop.