You can control the motor speeds with the remote control, which means that you can change the height of the drone (pitch functionality),
but you cannot move in different directions (roll & yaw functionality). Flight controller PCBs have integrated all remote controllers
channels connections for programming, but the programming logic is not implemented.
QUESTION 2.2. : WHAT IS NOT WORKING OR NEEDS TO BE IMPROVED?
(does not affect functionality, only design improvement) In power supply PCB I had planned a screw terminal
with 5 mm pitch size wires coming out of the battery holder on the power board, but when I got
the battery holder I found that the wires were narrower than I thought and would actually have been enough
with a 3.5 mm pitch screw screw terminal.
Landing gears attaching the drone to the plywood frame was a challenge, it required the use of force and
a little tweezers to separate the parts. Despite the fact that the compatible dimensions were modeled 1: 1,
it should be taken into account that using a laser cutter creates a curf, as a result of which the model does not
coincide with reality. To improve this project, it would be necessary to (a) find out how much the laser cutter
result decreases from the model parameters or (b) reduce the size of the landing gears model.
It is better to use the same type of materials for the connection line fasteners, currently I have a small rectangle,
it would be better to print them.
For security purposes, ESCs are fixed with plastic lock straps, it would be better to develop other solutions such
as 3D printed ones such as clothes pegs.
For safety reasons, batteries should be provided with a safety box to prevent them from moving. At the moment,
for safety reasons, I have the batteries secured with heavy duty strong tape.
Since I only realized at the last minute that I don't have a safety switch for the batteries (when they are inserted
the power must be turned off), so I added a lamp switch to the short solution. It would be better to integrate a
electronic component switch into the PCB.
Protective covers for propellers in the event of a fall would be useful.
QUESTION 3 : WHAT QUESTIONS NEED TO BE RESOLVED?
Why sometimes when the drone is turned on for the first time, all ESCs are initialized, but sometimes only three
of the four engines rotate, sometimes one of the four, after starting the remote control. After turning it off
and on again, it works, which leads to the conclusion that it could not be associated with a bad solder or
connection to the PCB.
Whether is it possible to further adjust the PWM values to make it more sensitive?
QUESTION 4 : WHAT WILL HAPPEN WHEN?
For the implementation of this project it is important to start step by step and only gradually complicate. How I did it:
1. I created a PCB to which a motor can be connected. I created a program where the speed values change in the loop and the motor rotates accordingly.
2. I created a PCB to which a motor can be connected and at the same time a remote controller channel can be connected. I created a program so that the motor rotates depending on the state of the remote controller.
3. I created a PCB to which four motors can be connected, connected all to the static power supply in the laboratory. I created a program for four motors depending on the state of the remote controller.
4. I created a power supply PCB, connected it to the previously created flight controller board and four motors.
5. I developed a design for system integration.
6. Performed test flights and improvements found in test trials.
QUESTION 5 : WHAT HAVE YOU LEARNED?
Results from a laser cutter are not suitable for mass production - several fasteners have different kerf.
It is not possible to implement large-scale functional PCBs in the form of one side - double side PCBs is the best choise.
It is not advisable to use a lot of vias and rivets, but if necessary, take a lot of time and patience to incorporate it into the PCB and solder it.
Asymmetry will always make you suffer (even if you don't think so at first). The PCB was not symmetrical for me, later making the frame design,
it was difficult because the board had to be fastened inside the frame.
When working with batteries, special care must be taken and attention must be paid to current values.
When working with different materials and integrating them into one system, one must not forget about natural losses, for example, in laser cutting - kerf,
or in the process of PLA 3D printing, the first layer of the object is always of slightly lower quality. For these reasons, I manually corrected these
errors with sandpaper. It would be best, of course, to predict this already in the design.
In the end, I wanted to transfer everything to the frame of the new carbon fiber material, but when I tried to transfer all the screws from plywood, I concluded
that the previously used screws could not be screwed in the appropriate places in the new material. As a result, I was unable to transfer the entire drone to
everyday drone material. The design of both the laser cutter and the waterjet machines was the same, but I concluded that the laser cutter has the bigger deviations
from the design what is related to kerf and during the final project I chose the screws to the laser cutter solution, but the waterjet machine is a much more accurate
and this ir reason why previous chosen screws are incompatible. To transfer everything to the new material, it would be necessary to replace all the screws.