Merge branch 'master' into stm

This commit is contained in:
Oleg Kalachev
2026-08-11 23:17:04 +03:00
parent c3b818c2ae
commit b8c687f3ed
117 changed files with 2264 additions and 810 deletions
+26
View File
@@ -27,3 +27,29 @@ Soldered components ([schematics variant](https://miro.com/app/board/uXjVN-dTjoo
<br>Assembled drone:
<img src="img/assembly/7.jpg" width=600>
See an alternative assembly process photos here: https://drive.google.com/drive/folders/1FG5BH9RCzdf1XmJcC70PymiRMXcz6Fx7?usp=sharing.
## Motor directions
> [!WARNING]
> The drone above is an early build, and it has **inversed** motor directions scheme. The photos only illustrate the assembly process in general.
Use standard motor directions scheme:
<img src="img/motors.svg" width=200>
Motors connection table:
|Motor|Position|Direction|Prop type|Motor wires|GPIO|
|-|-|-|-|-|-|
|Motor 0|Rear left|Counter-clockwise|B|Black & White|GPIO12 *(TDI)*|
|Motor 1|Rear right|Clockwise|A|Blue & Red|GPIO13 *(TCK)*|
|Motor 2|Front right|Counter-clockwise|B|Black & White|GPIO14 *(TMS)*|
|Motor 3|Front left|Clockwise|A|Blue & Red|GPIO15 *(TD0)*|
## Motors tightening
Motors should be installed very tightly — any vibration may lead to bad attitude estimation and unstable flight. If motors are loose, use tiny tape pieces to fix them tightly as shown below:
<img src="img/motor-tape.jpg" width=600>
+3 -3
View File
@@ -12,8 +12,8 @@
* `acc` *(Vector)* — данные с акселерометра, *м/с<sup>2</sup>*.
* `rates` *(Vector)* — отфильтрованные угловые скорости, *рад/с*.
* `attitude` *(Quaternion)* — оценка ориентации (положения) дрона.
* `controlRoll`, `controlPitch`, ... *(float[])* — команды управления от пилота, в диапазоне [-1, 1].
* `motors` *(float[])* — выходные сигналы на моторы, в диапазоне [0, 1].
* `controlRoll`, `controlPitch`, `controlYaw`, `controlThrottle`, `controlMode` *(float)* — команды управления от пилота, в диапазоне [-1, 1].
* `motors` *(float[4])* — выходные сигналы на моторы, в диапазоне [0, 1].
## Исходные файлы
@@ -35,7 +35,7 @@
### Подсистема управления
Состояние органов управления обрабатывается в функции `interpretControls()` и преобразуется в *команду управления*, которая включает следующее:
Состояние органов управления обрабатывается в функции `interpretControls()` и преобразуется в **команду управления**, которая включает следующее:
* `attitudeTarget` *(Quaternion)* — целевая ориентация дрона.
* `ratesTarget` *(Vector)* — целевые угловые скорости, *рад/с*.
+9 -9
View File
@@ -110,7 +110,7 @@ float angle = Vector::angleBetween(a, b); // 1.57 (90 градусов)
#### Скалярное произведение
Скалярное произведение векторов (*dot product*) — это произведение длин двух векторов на косинус угла между ними. В математике оно обозначается знаком `·` или слитным написанием векторов. Интуитивно, результат скалярного произведения показывает, насколько два вектора *сонаправлены*.
Скалярное произведение векторов *(dot product)* — это произведение длин двух векторов на косинус угла между ними. В математике оно обозначается знаком `·` или слитным написанием векторов. Интуитивно, результат скалярного произведения показывает, насколько два вектора *сонаправлены*.
В Flix используется статический метод `Vector::dot()`:
@@ -124,7 +124,7 @@ float dotProduct = Vector::dot(a, b); // 32
#### Векторное произведение
Векторное произведение (*cross product*) позволяет найти вектор, перпендикулярный двум другим векторам. В математике оно обозначается знаком `×`, а в прошивке используется статический метод `Vector::cross()`:
Векторное произведение *(cross product)* позволяет найти вектор, перпендикулярный двум другим векторам. В математике оно обозначается знаком `×`, а в прошивке используется статический метод `Vector::cross()`:
```cpp
Vector a(1, 2, 3);
@@ -144,9 +144,9 @@ Vector crossProduct = Vector::cross(a, b); // -3, 6, -3
В прошивке углы Эйлера сохраняются в обычный объект `Vector` (хоть и, строго говоря, не являются вектором):
* Угол по крену (*roll*) — `vector.x`.
* Угол по тангажу (*pitch*) — `vector.y`.
* Угол по рысканию (*yaw*) — `vector.z`.
* Угол по крену *(roll)* — `vector.x`.
* Угол по тангажу *(pitch)* — `vector.y`.
* Угол по рысканию *(yaw)* — `vector.z`.
Особенности углов Эйлера:
@@ -162,8 +162,8 @@ Vector crossProduct = Vector::cross(a, b); // -3, 6, -3
Помимо углов Эйлера, любую ориентацию в трехмерном пространстве можно представить в виде вращения вокруг некоторой оси на некоторый угол. В геометрии это доказывается, как **теорема вращения Эйлера**. В таком представлении ориентация задается двумя величинами:
* **Ось вращения** (*axis*) — единичный вектор, определяющий ось вращения.
* **Угол поворота** (*angle* или *θ*) — угол, на который нужно повернуть объект вокруг этой оси.
* **Ось вращения** *(axis)* — единичный вектор, определяющий ось вращения.
* **Угол поворота** *(angle* или *θ)* — угол, на который нужно повернуть объект вокруг этой оси.
В Flix ось вращения задается объектом `Vector`, а угол поворота — числом типа `float` в радианах:
@@ -177,7 +177,7 @@ float angle = radians(45);
### Вектор вращения
Если умножить вектор *axis* на угол поворота *θ*, то получится **вектор вращения** (*rotation vector*). Этот вектор играет важную роль в алгоритмах управления ориентацией летательного аппарата.
Если умножить вектор *axis* на угол поворота *θ*, то получится **вектор вращения** *(rotation vector)*. Этот вектор играет важную роль в алгоритмах управления ориентацией летательного аппарата.
Вектор вращения обладает замечательным свойством: если угловые скорости объекта (в собственной системе координат) в каждый момент времени совпадают с компонентами этого вектора, то за единичное время объект придет к заданной этим вектором ориентации. Это свойство позволяет использовать вектор вращения для управления ориентацией объекта посредством управления угловыми скоростями.
@@ -198,7 +198,7 @@ Vector rotation = radians(45) * Vector(1, 2, 3);
<a href="https://github.com/okalachev/flix/blob/master/flix/quaternion.h"><code>quaternion.h</code></a>.<br>
</div>
Вектор вращения удобен, но еще удобнее использовать **кватернион**. В Flix кватернионы задаются объектами `Quaternion` из библиотеки `quaternion.h`. Кватернион состоит из четырех значений: *w*, *x*, *y*, *z* и рассчитывается из вектора оси вращения (*axis*) и угла поворота (*θ*) по формуле:
Вектор вращения удобен, но еще удобнее использовать **кватернион**. В Flix кватернионы задаются объектами `Quaternion` из библиотеки `quaternion.h`. Кватернион состоит из четырех значений: *w*, *x*, *y*, *z* и рассчитывается из вектора оси вращения *(axis)* и угла поворота *(θ)* по формуле:
\\[ q = \left( \begin{array}{c} w \\\\ x \\\\ y \\\\ z \end{array} \right) = \left( \begin{array}{c} \cos\left(\frac{\theta}{2}\right) \\\\ axis\_x \cdot \sin\left(\frac{\theta}{2}\right) \\\\ axis\_y \cdot \sin\left(\frac{\theta}{2}\right) \\\\ axis\_z \cdot \sin\left(\frac{\theta}{2}\right) \end{array} \right) \\]
+27 -27
View File
@@ -87,13 +87,13 @@ Flix поддерживает следующие модели IMU:
#include <FlixPeriph.h>
#include <SPI.h>
MPU9250 IMU(SPI);
MPU9250 imu(SPI);
void setup() {
Serial.begin(115200);
bool success = IMU.begin();
bool success = imu.begin();
if (!success) {
Serial.println("Failed to initialize IMU");
Serial.println("Failed to initialize the IMU");
}
}
```
@@ -108,21 +108,21 @@ void setup() {
#include <FlixPeriph.h>
#include <SPI.h>
MPU9250 IMU(SPI);
MPU9250 imu(SPI);
void setup() {
Serial.begin(115200);
bool success = IMU.begin();
bool success = imu.begin();
if (!success) {
Serial.println("Failed to initialize IMU");
Serial.println("Failed to initialize the IMU");
}
}
void loop() {
IMU.waitForData();
imu.waitForData();
float gx, gy, gz;
IMU.getGyro(gx, gy, gz);
imu.getGyro(gx, gy, gz);
Serial.printf("gx:%f gy:%f gz:%f\n", gx, gy, gz);
delay(50); // замедление вывода
@@ -135,36 +135,36 @@ void loop() {
## Конфигурация гироскопа
В коде Flix настройка IMU происходит в функции `configureIMU`. В этой функции настраиваются три основных параметра гироскопа: диапазон измерений, частота сэмплов и частота LPF-фильтра.
В коде Flix настройка IMU происходит в функции `configureIMU`. В этой функции настраиваются три основных параметра гироскопа: диапазон измерений, частота сэмплирования и частота LPF-фильтра.
### Частота сэмплов
### Частота сэмплирования
Большинство IMU могут обновлять данные с разной частотой. В полетных контроллерах обычно используется частота обновления от 500 Гц до 8 кГц. Чем выше частота сэмплов, тем выше точность управления полетом, но и больше нагрузка на микроконтроллер.
Большинство IMU могут обновлять данные с разной частотой. В полетных контроллерах обычно используется частота обновления от 500 Гц до 8 кГц. Чем выше частота, тем выше точность управления полетом, но и тем больше нагрузка на микроконтроллер.
Частота сэмплов устанавливается методом `setSampleRate()`. В Flix используется частота 1 кГц:
Частота сэмплирования устанавливается методом `setSampleRate()`. В Flix используется частота 1 кГц:
```cpp
IMU.setRate(IMU.RATE_1KHZ_APPROX);
```
Поскольку не все поддерживаемые IMU могут работать строго на частоте 1 кГц, в библиотеке FlixPeriph существует возможность приближенной настройки частоты сэмплов. Например, у IMU ICM-20948 при такой настройке реальная частота сэмплирования будет равна 1125 Гц.
Поскольку не все поддерживаемые IMU могут работать строго на частоте 1 кГц, в библиотеке FlixPeriph существует возможность приближенной настройки частоты сэмплирования. Например, у IMU ICM-20948 при такой настройке реальная частота сэмплирования будет равна 1125 Гц.
Другие доступные для установки в библиотеке FlixPeriph частоты сэмплирования:
* `RATE_MIN` — минимальная частота сэмплов для конкретного IMU.
* `RATE_MIN` — минимальная частота для конкретного IMU.
* `RATE_50HZ_APPROX` — значение, близкое к 50 Гц.
* `RATE_1KHZ_APPROX` — значение, близкое к 1 кГц.
* `RATE_8KHZ_APPROX` — значение, близкое к 8 кГц.
* `RATE_MAX` — максимальная частота сэмплов для конкретного IMU.
* `RATE_MAX` — максимальная частота для конкретного IMU.
#### Диапазон измерений
Большинство MEMS-гироскопов поддерживают несколько диапазонов измерений угловой скорости. Главное преимущество выбора меньшего диапазона — бо́льшая чувствительность. В полетных контроллерах обычно выбирается максимальный диапазон измерений от –2000 до 2000 градусов в секунду, чтобы обеспечить возможность динамичных маневров.
Большинство MEMS-гироскопов поддерживают несколько диапазонов измерений угловой скорости. Главное преимущество выбора меньшего диапазона — бо́льшая чувствительность. В полетных контроллерах обычно выбирается максимальный диапазон измерений от –2000 до 2000 градусов в секунду, чтобы обеспечить возможность быстрых маневров.
В библиотеке FlixPeriph диапазон измерений гироскопа устанавливается методом `setGyroRange()`:
```cpp
IMU.setGyroRange(IMU.GYRO_RANGE_2000DPS);
imu.setGyroRange(imu.GYRO_RANGE_2000DPS);
```
### LPF-фильтр
@@ -172,16 +172,16 @@ IMU.setGyroRange(IMU.GYRO_RANGE_2000DPS);
IMU InvenSense могут фильтровать измерения на аппаратном уровне при помощи фильтра нижних частот (LPF). Flix реализует собственный фильтр для гироскопа, чтобы иметь больше гибкости при поддержке разных IMU. Поэтому для встроенного LPF устанавливается максимальная частота среза:
```cpp
IMU.setDLPF(IMU.DLPF_MAX);
imu.setDLPF(imu.DLPF_MAX);
```
## Калибровка гироскопа
Как и любое измерительное устройство, гироскоп вносит искажения в измерения. Наиболее простая модель этих искажений делит их на статические смещения (*bias*) и случайный шум (*noise*):
Как и любое измерительное устройство, гироскоп вносит искажения в измерения. Наиболее простая модель этих искажений делит их на статические смещения *(bias)* и случайный шум *(noise)*:
\\[ gyro_{xyz}=rates_{xyz}+bias_{xyz}+noise \\]
Для качественной работы подсистемы оценки ориентации и управления дроном необходимо оценить *bias* гироскопа и учесть его в вычислениях. Для этого при запуске программы производится калибровка гироскопа, которая реализована в функции `calibrateGyro()`. Эта функция считывает данные с гироскопа в состоянии покоя 1000 раз и усредняет их. Полученные значения считаются *bias* гироскопа и в дальнейшем вычитаются из измерений.
Для точной работы подсистемы оценки ориентации и управления дроном необходимо оценить *bias* гироскопа и учесть его в вычислениях. Для этого при запуске программы производится калибровка гироскопа, которая реализована в функции `calibrateGyro()`. Эта функция считывает данные с гироскопа в состоянии покоя 1000 раз и усредняет их. Полученные значения считаются *bias* гироскопа и в дальнейшем вычитаются из измерений.
Программа для вывода данных с гироскопа с калибровкой:
@@ -189,23 +189,23 @@ IMU.setDLPF(IMU.DLPF_MAX);
#include <FlixPeriph.h>
#include <SPI.h>
MPU9250 IMU(SPI);
MPU9250 imu(SPI);
float gyroBiasX, gyroBiasY, gyroBiasZ; // bias гироскопа
void setup() {
Serial.begin(115200);
bool success = IMU.begin();
bool success = imu.begin();
if (!success) {
Serial.println("Failed to initialize IMU");
Serial.println("Failed to initialize the IMU");
}
calibrateGyro();
}
void loop() {
float gx, gy, gz;
IMU.waitForData();
IMU.getGyro(gx, gy, gz);
imu.waitForData();
imu.getGyro(gx, gy, gz);
// Устранение bias гироскопа
gx -= gyroBiasX;
@@ -226,9 +226,9 @@ void calibrateGyro() {
// Получение 1000 измерений гироскопа
for (int i = 0; i < samples; i++) {
IMU.waitForData();
imu.waitForData();
float gx, gy, gz;
IMU.getGyro(gx, gy, gz);
imu.getGyro(gx, gy, gz);
gyroBiasX += gx;
gyroBiasY += gy;
gyroBiasZ += gz;
-1
View File
@@ -1 +0,0 @@
usage.md
+2
View File
@@ -0,0 +1,2 @@
<!-- markdownlint-disable MD041 -->
Build instructions are moved to [usage article](usage.md).
+57 -9
View File
@@ -6,7 +6,7 @@ The firmware is a regular Arduino sketch, and it follows the classic Arduino one
<img src="img/dataflow.svg" width=600 alt="Firmware dataflow diagram">
The main loop is running at 1000 Hz. All the dataflow goes through global variables (for simplicity):
The main loop is running at 1000 Hz. The dataflow goes through global variables, including:
* `t` *(float)* — current step time, *s*.
* `dt` *(float)* — time delta between the current and previous steps, *s*.
@@ -14,12 +14,12 @@ The main loop is running at 1000 Hz. All the dataflow goes through global variab
* `acc` *(Vector)* — acceleration data from the accelerometer, *m/s<sup>2</sup>*.
* `rates` *(Vector)* — filtered angular rates, *rad/s*.
* `attitude` *(Quaternion)* — estimated attitude (orientation) of drone.
* `controlRoll`, `controlPitch`, ... *(float[])* — pilot control inputs, range [-1, 1].
* `motors` *(float[])* — motor outputs, range [0, 1].
* `controlRoll`, `controlPitch`, `controlYaw`, `controlThrottle`, `controlMode` *(float)* — pilot control inputs, range [-1, 1].
* `motors` *(float[4])* — motor outputs, range [0, 1].
## Source files
Firmware source files are located in `flix` directory. The core files are:
Firmware source files are located in `flix` directory.
* [`flix.ino`](../flix/flix.ino) — Arduino sketch main file, entry point.Includes some global variable definitions and the main loop.
* [`imu.ino`](../flix/imu.ino) — reading data from the IMU sensor (gyroscope and accelerometer), IMU calibration.
@@ -28,6 +28,7 @@ Firmware source files are located in `flix` directory. The core files are:
* [`control.ino`](../flix/control.ino) — control subsystem, three-dimensional two-level cascade PID controller.
* [`motors.ino`](../flix/motors.ino) — PWM motor output control.
* [`mavlink.ino`](../flix/mavlink.ino) — interaction with QGroundControl or [pyflix](../tools/pyflix) via MAVLink protocol.
* [`cli.ino`](../flix/cli.ino) — serial and MAVLink console.
Utility files:
@@ -37,20 +38,67 @@ Utility files:
### Control subsystem
Pilot inputs are interpreted in `interpretControls()`, and then converted to the *control command*, which consists of the following:
Pilot inputs are interpreted in `interpretControls()`, and then converted to the **control command**, which consists of the following:
* `attitudeTarget` *(Quaternion)* — target attitude of the drone.
* `ratesTarget` *(Vector)* — target angular rates, *rad/s*.
* `ratesExtra` *(Vector)* — additional (feed-forward) angular rates , used for yaw rate control in STAB mode, *rad/s*.
* `ratesExtra` *(Vector)* — additional (feed-forward) angular rates, used for yaw rate control in STAB mode, *rad/s*.
* `torqueTarget` *(Vector)* — target torque, range [-1, 1].
* `thrustTarget` *(float)* — collective thrust target, range [0, 1].
* `thrustTarget` *(float)* — collective motor thrust target, range [0, 1].
Control command is processed in `controlAttitude()`, `controlRates()`, `controlTorque()` functions. Each function may be skipped if the corresponding target is set to `NAN`.
Control command is handled in `controlAttitude()`, `controlRates()`, `controlTorque()` functions. Each function may be skipped if the corresponding control target is set to `NAN`.
<img src="img/control.svg" width=300 alt="Control subsystem diagram">
Armed state is stored in `armed` variable, and current mode is stored in `mode` variable.
## Building
### Console
To write into the console, `print()` function is used. This function sends data both to the Serial console and to the MAVLink console (which can be accessed wirelessly in QGroundControl). The function supports formatting:
```cpp
print("Test value: %.2f\n", testValue);
```
In order to add a console command, modify the `doCommand()` function in `cli.ino` file.
> [!IMPORTANT]
> Avoid using delays in in-flight commands, it will **crash** the drone! (The design is one-threaded.)
>
> For on-the-ground commands, use `pause()` function, instead of `delay()`. This function allows to pause in a way that MAVLink connection will continue working.
### Parameter subsystem
Parameters subsystem (`parameters.ino`) uses standard [Preferences.h](https://docs.espressif.com/projects/arduino-esp32/en/latest/tutorials/preferences.html) ESP32 library to store parameters in non-volatile memory. Each parameter is a regular global variable, which is registered in the `parameters` array.
To add a new parameter:
1. Define a global variable for the parameter, two types are supported: `float` and `int`.
2. Add an entry to the `parameters` array, with the parameter name, a pointer to the variable, and optionally a callback function to call when the parameter is changed.
3. Everything else will be handled automatically.
See examples of adding new parameters in commits: [c434107](https://github.com/okalachev/flix/commit/c434107), [a687303](https://github.com/okalachev/flix/commit/a687303).
> [!NOTE]
> Since all the parameters are internally stored and passed as floats, the safe range for `int` parameters is -16777216 to 16777215.
## Adding a subsystem
To add a new subsystem:
1. Create a new `*.ino` file for your subsystem.
2. Define setup and loop functions for the subsystem, for example `setupMySubsystem()` and `loopMySubsystem()`.
3. Use `Rate` class if you need to limit the loop frequency, for example:
```cpp
Rate mySubsystemRate(100); // 100 Hz
void loopMySubsystem() {
if (!mySubsystemRate) return;
// Do something...
}
4. Add setup and loop calls in to `setup()` and `loop()` functions in `flix.ino`.
## Building the firmware
See build instructions in [usage.md](usage.md).
Binary file not shown.

After

Width:  |  Height:  |  Size: 70 KiB

+1 -2
View File
File diff suppressed because one or more lines are too long

Before

Width:  |  Height:  |  Size: 140 KiB

After

Width:  |  Height:  |  Size: 15 KiB

File diff suppressed because one or more lines are too long

Before

Width:  |  Height:  |  Size: 101 KiB

After

Width:  |  Height:  |  Size: 13 KiB

+136
View File
@@ -0,0 +1,136 @@
<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 813.79 508.65">
<defs>
<style>
.a, .d, .f, .j, .p {
fill: none;
}
.a {
stroke: #d5d5d5;
stroke-width: 31px;
}
.a, .p {
stroke-miterlimit: 10;
}
.b {
fill: #ff9400;
}
.c {
opacity: 0.8;
}
.d {
stroke: #d80100;
}
.d, .f, .j {
stroke-linejoin: bevel;
stroke-width: 13px;
}
.e {
fill: #d80100;
}
.f {
stroke: #57ed00;
}
.g {
fill: #57ed00;
}
.h {
fill: #c1c1c1;
}
.i {
opacity: 0.12;
}
.j {
stroke: #0076ba;
}
.k {
fill: #0076ba;
}
.l {
font-size: 50px;
font-family: Tahoma;
}
.m {
letter-spacing: -0.01em;
}
.n {
letter-spacing: -0.01em;
}
.o {
letter-spacing: 0em;
}
.p {
stroke: #000;
stroke-width: 10px;
}
</style>
</defs>
<g>
<g>
<line class="a" x1="259.14" y1="432.84" x2="532.46" y2="340.23"/>
<line class="a" x1="311.69" y1="313.16" x2="481.62" y2="461.39"/>
<ellipse class="b" cx="311.35" cy="312.8" rx="88.68" ry="47.94"/>
<ellipse class="b" cx="532.53" cy="340.42" rx="88.68" ry="47.94"/>
<ellipse class="b" cx="479.72" cy="460.7" rx="88.68" ry="47.94"/>
<ellipse class="b" cx="259.21" cy="433.03" rx="88.68" ry="47.94"/>
</g>
<g class="c">
<g>
<line class="d" x1="396.65" y1="386.83" x2="564.66" y2="35.72"/>
<polygon class="e" points="582.76 51.95 579.15 5.42 540.66 31.8 582.76 51.95"/>
</g>
</g>
<g class="c">
<g>
<line class="f" x1="396.77" y1="387.06" x2="69.41" y2="341.09"/>
<polygon class="g" points="79.42 318.93 36.15 336.42 72.93 365.15 79.42 318.93"/>
</g>
</g>
<ellipse class="h" cx="396.36" cy="386.61" rx="47.21" ry="25.52"/>
<path class="i" d="M398,375.67l-14.42,12.95a4.32,4.32,0,0,0,2.35,7.5l16.81,2.11a4.33,4.33,0,0,0,4.8-5l-2.39-15.06A4.32,4.32,0,0,0,398,375.67Z"/>
<g class="c">
<g>
<line class="j" x1="396.77" y1="385.56" x2="396.77" y2="92.64"/>
<polygon class="k" points="420.1 99.47 396.77 59.06 373.43 99.47 420.1 99.47"/>
</g>
</g>
<text class="l" transform="translate(0 292.27)">y/left</text>
<text class="l" transform="translate(268.4 81.58)">z/up</text>
<text class="l" transform="translate(600.99 43.18)">x/<tspan class="m" x="43.87" y="0">f</tspan><tspan x="59.3" y="0">or</tspan><tspan class="n" x="104.47" y="0">w</tspan><tspan x="141.14" y="0">a</tspan><tspan class="o" x="167.38" y="0">r</tspan><tspan x="185.16" y="0">d</tspan></text>
<g class="c">
<g>
<path class="p" d="M470.35,114a52.71,52.71,0,0,1,103.57-2.31,51.67,51.67,0,0,1-1.33,25.92"/>
<polygon points="562.15 128.29 563.93 154.16 585.44 139.69 562.15 128.29"/>
</g>
</g>
<g class="c">
<g>
<path class="p" d="M344.16,164.77a52.66,52.66,0,1,0,103.78,16.31"/>
<polygon points="460.61 184.04 446.18 162.5 434.74 185.76 460.61 184.04"/>
</g>
</g>
<g class="c">
<g>
<path class="p" d="M138.4,411.11a52.71,52.71,0,0,1,18.43-101.94A51.68,51.68,0,0,1,182,315.65"/>
<polygon points="170.73 324.01 196.43 327.44 186.55 303.47 170.73 324.01"/>
</g>
</g>
</g>
</svg>

After

Width:  |  Height:  |  Size: 3.5 KiB

+110
View File
@@ -0,0 +1,110 @@
<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 664.49 478.47">
<defs>
<style>
.a, .d, .f, .j {
fill: none;
}
.a {
stroke: #d5d5d5;
stroke-miterlimit: 10;
stroke-width: 31px;
}
.b {
fill: #ff9400;
}
.c {
opacity: 0.8;
}
.d {
stroke: #d80100;
}
.d, .f, .j {
stroke-linejoin: bevel;
stroke-width: 13px;
}
.e {
fill: #d80100;
}
.f {
stroke: #57ed00;
}
.g {
fill: #57ed00;
}
.h {
fill: #c1c1c1;
}
.i {
opacity: 0.12;
}
.j {
stroke: #0076ba;
}
.k {
fill: #0076ba;
}
.l {
font-size: 50px;
font-family: Tahoma;
}
.m {
letter-spacing: -0.01em;
}
.n {
letter-spacing: -0.01em;
}
.o {
letter-spacing: 0em;
}
</style>
</defs>
<g>
<g>
<line class="a" x1="207.39" y1="402.28" x2="480.71" y2="309.67"/>
<line class="a" x1="259.93" y1="282.6" x2="429.86" y2="430.83"/>
<ellipse class="b" cx="259.59" cy="282.24" rx="88.68" ry="47.94"/>
<ellipse class="b" cx="480.77" cy="309.86" rx="88.68" ry="47.94"/>
<ellipse class="b" cx="427.96" cy="430.14" rx="88.68" ry="47.94"/>
<ellipse class="b" cx="207.45" cy="402.47" rx="88.68" ry="47.94"/>
</g>
<g class="c">
<g>
<line class="d" x1="344.89" y1="356.27" x2="422.02" y2="172.47"/>
<polygon class="e" points="440.89 187.79 435.01 141.5 397.86 169.74 440.89 187.79"/>
</g>
</g>
<g class="c">
<g>
<line class="f" x1="345.01" y1="356.5" x2="79.27" y2="319.17"/>
<polygon class="g" points="89.28 297.01 46.01 314.5 82.78 343.23 89.28 297.01"/>
</g>
</g>
<ellipse class="h" cx="344.6" cy="356.05" rx="47.21" ry="25.52"/>
<path class="i" d="M346.25,345.11l-14.41,12.95a4.32,4.32,0,0,0,2.34,7.5L351,367.67a4.33,4.33,0,0,0,4.8-5l-2.39-15.06A4.31,4.31,0,0,0,346.25,345.11Z"/>
<g class="c">
<g>
<line class="j" x1="345.01" y1="355" x2="345.01" y2="62.08"/>
<polygon class="k" points="368.35 68.91 345.01 28.5 321.67 68.91 368.35 68.91"/>
</g>
</g>
<text class="l" transform="translate(-0.76 281.71)">y/left</text>
<text class="l" transform="translate(371.65 38.02)">z/up</text>
<text class="l" transform="translate(455.23 152.62)">x/<tspan class="m" x="43.87" y="0">f</tspan><tspan x="59.3" y="0">or</tspan><tspan class="n" x="104.47" y="0">w</tspan><tspan x="141.14" y="0">a</tspan><tspan class="o" x="167.38" y="0">r</tspan><tspan x="185.16" y="0">d</tspan></text>
</g>
</svg>

After

Width:  |  Height:  |  Size: 2.7 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 46 KiB

+38
View File
@@ -0,0 +1,38 @@
<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 734.86 378.46">
<defs>
<style>
.a {
fill: none;
stroke: #d5d5d5;
stroke-miterlimit: 10;
stroke-width: 31px;
}
.b {
fill: #c1c1c1;
}
.c {
fill: #ff9400;
}
.d {
fill: #d5d5d5;
}
</style>
</defs>
<g>
<g>
<line class="a" x1="448.78" y1="294.23" x2="648.77" y2="93.24"/>
<line class="a" x1="449.78" y1="94.24" x2="649.77" y2="295.23"/>
<circle class="b" cx="549.27" cy="193.73" r="41.5"/>
<circle class="c" cx="449.35" cy="93.74" r="77.95"/>
<circle class="c" cx="648.89" cy="93.53" r="77.95"/>
<circle class="c" cx="647.89" cy="294.51" r="77.95"/>
<circle class="c" cx="448.9" cy="294.51" r="77.95"/>
</g>
<path class="d" d="M8,161.93q0-17.85,22.36-17.85h4.81V100.57Q35.17,8.49,131.23,8h1q36.87,0,47.31,7.48L181,17.41l1,2.41V50q0,12.32-5.82,12.31-2.43,0-7.4-4.64t-14.19-9a48.63,48.63,0,0,0-21.22-4.41q-12,0-19.17,3.5a18.85,18.85,0,0,0-9.82,10.62,67.35,67.35,0,0,0-3.52,12.06,82.52,82.52,0,0,0-.85,13.39v60.32h27.28q10.86,0,16.05,5.43a17.52,17.52,0,0,1,5.19,12.42,22.5,22.5,0,0,1-1.21,7.36q-1.22,3.51-6.64,7.24t-14.36,3.74H101.64V344.82a56,56,0,0,1-.61,9.65,37.8,37.8,0,0,1-2.56,7.6,11.83,11.83,0,0,1-6.94,6.4q-5,1.93-13.51,1.93H57.08q-10.47,0-15.7-4.71t-5.73-8.44a77.31,77.31,0,0,1-.48-9.53V180.27H29.4Q8,180.27,8,161.93Z"/>
<path class="d" d="M201.21,348.2V37q0-23.16,20.86-23.4h22.81q22.8,0,22.8,22.44V346.27a68.92,68.92,0,0,1-.49,9.41,22.59,22.59,0,0,1-2.42,7.12,11.48,11.48,0,0,1-6.67,5.43,47.78,47.78,0,0,1-12.74,2.17H225q-11.4,0-17.58-4.47T201.21,348.2Z"/>
<path class="d" d="M284.9,61.08V36.47a40.39,40.39,0,0,1,1.7-12.91,11.36,11.36,0,0,1,6.18-7,25.27,25.27,0,0,1,6.68-2.3c1.45-.15,4-.4,7.76-.72h25.23q10.43,0,16.73,4.7t6.31,18.22V62.05a27.94,27.94,0,0,1-.85,7.11,23,23,0,0,1-2.06,5.43,20,20,0,0,1-2.91,4,10,10,0,0,1-3.52,2.54c-1.21.48-2.54,1-4,1.44a11.53,11.53,0,0,1-3.4.73,13.71,13.71,0,0,0-3.15.48H307.7q-10.43,0-16.61-4.7T284.9,61.08Zm1.94,284.71V166.28q0-22.2,21.83-22.2h20.38q10.92,0,16.62,4t6.67,8.45a60.47,60.47,0,0,1,1,12.18V348.2q0,22.2-21.83,22.2H308.67q-10.43,0-15.64-4.95t-5.7-8.81A90.93,90.93,0,0,1,286.84,345.79Z"/>
</g>
</svg>

After

Width:  |  Height:  |  Size: 2.2 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 101 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 33 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 18 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 18 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 17 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 17 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 10 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 9.6 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 10 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 10 KiB

Binary file not shown.

Before

Width:  |  Height:  |  Size: 38 KiB

After

Width:  |  Height:  |  Size: 46 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 61 KiB

+89
View File
@@ -0,0 +1,89 @@
<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 356.79 357.11">
<defs>
<style>
.a, .e {
fill: none;
stroke-miterlimit: 10;
}
.a {
stroke: #d5d5d5;
stroke-width: 31px;
}
.b {
fill: #c1c1c1;
}
.c {
fill: #ff9400;
}
.d {
opacity: 0.12;
}
.e {
stroke: #cc5200;
stroke-width: 8px;
}
.f {
fill: #cc5200;
}
.g {
font-size: 50px;
fill: #fff;
}
.g, .h {
font-family: Tahoma;
}
.h {
font-size: 20px;
}
.i {
letter-spacing: 0em;
}
</style>
</defs>
<g>
<g>
<line class="a" x1="77.4" y1="278.67" x2="277.4" y2="77.67"/>
<line class="a" x1="78.4" y1="78.67" x2="278.4" y2="279.67"/>
<circle class="b" cx="177.9" cy="178.17" r="41.5"/>
<circle class="c" cx="77.97" cy="78.17" r="77.95"/>
<circle class="c" cx="277.52" cy="77.95" r="77.95"/>
</g>
<path class="d" d="M174.29,163.6l-8.45,26.05A4.32,4.32,0,0,0,170,195.3h16.9a4.32,4.32,0,0,0,4.11-5.65l-8.45-26.05A4.32,4.32,0,0,0,174.29,163.6Z"/>
<g>
<path class="e" d="M307.47,122.53a52.66,52.66,0,1,0-72.08-76"/>
<polygon class="f" points="228.68 38.51 227.44 59.21 245.99 49.94 228.68 38.51"/>
</g>
<g>
<path class="e" d="M48.11,122.22a52.66,52.66,0,1,1,72.08-75.95"/>
<polygon class="f" points="109.59 49.63 128.14 58.91 126.9 38.2 109.59 49.63"/>
</g>
<text class="g" transform="translate(64.89 98.77)">3</text>
<text class="g" transform="translate(260.92 98.8)">2</text>
<text class="h" transform="translate(66.06 129.25)">p<tspan class="i" x="11.05" y="0">r</tspan><tspan x="18.16" y="0">op A</tspan></text>
<text class="h" transform="translate(232.55 128.95)">p<tspan class="i" x="11.05" y="0">r</tspan><tspan x="18.16" y="0">op B</tspan></text>
<circle class="c" cx="278.83" cy="277.92" r="77.95"/>
<g>
<path class="e" d="M249,322a52.66,52.66,0,1,1,72.09-76"/>
<polygon class="f" points="310.45 249.38 329 258.66 327.76 237.95 310.45 249.38"/>
</g>
<text class="g" transform="translate(265.76 298.52)">1</text>
<text class="h" transform="translate(266.92 329.01)">p<tspan class="i" x="11.05" y="0">r</tspan><tspan x="18.16" y="0">op A</tspan></text>
<circle class="c" cx="77.95" cy="277.92" r="77.95"/>
<g>
<path class="e" d="M107.9,322.49a52.66,52.66,0,1,0-72.08-76"/>
<polygon class="f" points="29.11 238.47 27.87 259.18 46.42 249.91 29.11 238.47"/>
</g>
<text class="g" transform="translate(61.35 298.76)">0</text>
<text class="h" transform="translate(32.98 328.92)">p<tspan class="i" x="11.05" y="0">r</tspan><tspan x="18.16" y="0">op B</tspan></text>
</g>
</svg>

After

Width:  |  Height:  |  Size: 2.8 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 16 KiB

File diff suppressed because one or more lines are too long

Before

Width:  |  Height:  |  Size: 18 KiB

After

Width:  |  Height:  |  Size: 18 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 60 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 52 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 56 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 62 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 48 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 50 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 17 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 62 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 44 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 55 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 68 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 35 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 32 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 60 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 38 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 50 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 49 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 41 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 56 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 60 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 54 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 69 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 58 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 50 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 105 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 34 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 36 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 108 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 93 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 65 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 28 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 31 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 42 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 43 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 76 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 28 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 44 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 32 KiB

Binary file not shown.

After

Width:  |  Height:  |  Size: 65 KiB

+1 -5
View File
@@ -2,11 +2,7 @@
Flix quadcopter uses RAM to store flight log data. The default log capacity is 10 seconds at 100 Hz. This configuration can be adjusted in the `log.ino` file.
To perform log analysis, you need to download the log right after the flight without powering off the drone. Then you can use several tools to analyze the log data.
## Log download
To download the log, connect the ESP32 using USB right after the flight and run the following command:
To perform log analysis, you need to download the flight log. To to that, ensure you're connected to the drone using Wi-Fi and run the following command:
```bash
make log
+21 -11
View File
@@ -4,34 +4,44 @@
Do the following:
* **Check ESP32 core is installed**. Check if the version matches the one used in the [tutorial](usage.md#firmware).
* **Check libraries**. Install all the required libraries from the tutorial. Make sure there are no MPU9250 or other peripherals libraries that may conflict with the ones used in the tutorial.
* **Check ESP32 core is installed**. Check if the version matches the one used in the [tutorial](usage.md#building-the-firmware).
* **Check libraries**. Install all the required libraries from the tutorial. Make sure there are no MPU-9250 or other peripherals libraries that may conflict with the ones used in the tutorial.
* **Check the chosen board**. The correct board to choose in Arduino IDE for ESP32 Mini is *WEMOS D1 MINI ESP32*.
## The drone doesn't fly
Do the following:
* **Check the battery voltage**. Use a multimeter to measure the battery voltage. It should be in range of 3.7-4.2 V.
* **Check if there are some startup errors**. Connect the ESP32 to the computer and check the Serial Monitor output. Use the Reset button to make sure you see the whole ESP32 output.
* **Check the battery voltage**. Use a multimeter to measure the battery voltage. The fully charged battery should have about 4.2V.
* **Check the battery you use has enough discharge current**. The battery should be able to provide 15A of current. So the C-rating for a 1000 mAh battery should be at least 15C (higher is better).
* **Check if there are some startup errors**. Connect the ESP32 to the computer and check the Serial Monitor output. Use the Reset button or `reboot` command to see the whole startup output.
* **Check the baudrate is correct**. If you see garbage characters in the Serial Monitor, make sure the baudrate is set to 115200.
* **Make sure correct IMU model is chosen**. If using ICM-20948/MPU-6050 board, change `MPU9250` to `ICM20948`/`MPU6050` in the `imu.ino` file.
* **Check if the CLI is working**. Perform `help` command in Serial Monitor. You should see the list of available commands. You can also access the CLI using QGroundControl (*Vehicle Setup* ⇒ *Analyze Tools**MAVLink Console*).
* **Check if the console is working**. Perform `help` command in Serial Monitor. You should see the list of available commands. You can also access the console using QGroundControl *(Vehicle Setup**Analyze Tools**MAVLink Console)*.
* **Configure QGroundControl correctly before connecting to the drone** if you use it to control the drone. Go to the settings and enable *Virtual Joystick*. *Auto-Center Throttle* setting **should be disabled**.
* **If QGroundControl doesn't connect**, you might need to disable the firewall and/or VPN on your computer.
* **Make sure correct IMU model is chosen**. If using ICM-20948/MPU-6050 board, change `MPU9250` to `ICM20948`/`MPU6050` in the `imu.ino` file.
* **Check the IMU is working**. Perform `imu` command and check its output:
* The `status` field should be `OK`.
* The `rate` field should be about 1000 (Hz).
* The `accel` and `gyro` fields should change as you move the drone.
* **Check the IMU orientation is set correctly**. If the attitude estimation is rotated, set the correct IMU orientation as described in the [tutorial](usage.md#define-imu-orientation).
* **Calibrate the accelerometer.** if is wasn't done before. Type `ca` command in Serial Monitor and follow the instructions.
* **Check the attitude estimation**. Connect to the drone using QGroundControl. Rotate the drone in different orientations and check if the attitude estimation shown in QGroundControl is correct.
* **Check the IMU orientation is set correctly**. If the attitude estimation is rotated, make sure `rotateIMU` function is defined correctly in `imu.ino` file.
* **Check the attitude estimation**. Connect to the drone using QGroundControl. Rotate the drone in different orientations and check if the attitude estimation is shown exactly as on the video below:
<a href="https://youtu.be/yVRN23-GISU"><img width=200 src="https://i3.ytimg.com/vi/yVRN23-GISU/maxresdefault.jpg"></a>
* **Check the IMU output**. Connect to the drone using QGroundControl on your computer. Go to the *Analyze* tab, *MAVLINK Inspector*. Plot the data from the `SCALED_IMU` message. The gyroscope and accelerometer data should change according to the drone movement.
* **Check the motors type**. Motors with exact 3.7V voltage are needed, not ranged working voltage (3.7V — 6V).
* **Check the motors**. Perform the following commands using Serial Monitor:
* `mfr` — should rotate front right motor (counter-clockwise).
* `mfl` — should rotate front left motor (clockwise).
* `mrl` — should rotate rear left motor (counter-clockwise).
* `mrr` — should rotate rear right motor (clockwise).
* **Check the remote control**. Using `rc` command, check the control values reflect your sticks movement. All the controls should change between -1 and 1, and throttle between 0 and 1.
* If using SBUS receiver, **calibrate the RC**. Type `cr` command in Serial Monitor and follow the instructions.
* **Check the IMU output using QGroundControl**. Connect to the drone using QGroundControl on your computer. Go to the *Analyze* tab, *MAVLINK Inspector*. Plot the data from the `SCALED_IMU` message. The gyroscope and accelerometer data should change according to the drone movement.
* **Check the propeller directions are correct**. Make sure your propeller types (A or B) are installed as on the picture:
<img src="img/user/peter_ukhov-2/1.jpg" width="200">
* **If using an SBUS receiver**:
* **Define the used GPIO pin** in `RC_RX_PIN` parameter.
* **Calibrate the RC** using `cr` command in the console.
* **Check the controls** using `rc` command. All the controls should change between -1 and 1, and the throttle between 0 and 1.
+283 -137
View File
@@ -1,130 +1,67 @@
# Usage: build, setup and flight
To use Flix, you need to build the firmware and upload it to the ESP32 board. For simulation, you need to build and run the simulator.
To fly Flix quadcopter, you need to upload the firmware to the ESP32 board, and set up the drone for flight.
For the start, clone the repository using git:
## Uploading the firmware
You can either use the **prebuilt binaries** or **build the firmware** from sources — this will let you modify the firmware and add new features.
### Prebuilt binaries (the easiest way)
1. Download the latest firmware file using the following links:
<!-- markdownlint-disable MD044 -->
|Type|Boards|Link|
|-|-|-|
|ESP32|DevKit, D1 Mini|[`quadcopter.dev/flix.esp32.merged.bin`](https://quadcopter.dev/flix.esp32.merged.bin)|
|ESP32-S3|Most S3 based|[`quadcopter.dev/flix.esp32s3.merged.bin`](https://quadcopter.dev/flix.esp32s3.merged.bin)|
|ESP32-S3 (2MB PSRAM)|S3 Super Mini, S3 Zero (2MB PSRAM)|[`quadcopter.dev/flix.esp32s3.qspi.merged.bin`](https://quadcopter.dev/flix.esp32s3.qspi.merged.bin)|
|ESP32-S3 (8/16MB PSRAM)|S3 Zero (8MB PSRAM)|[`quadcopter.dev/flix.esp32s3.opi.merged.bin`](https://quadcopter.dev/flix.esp32s3.opi.merged.bin)|
|ESP32-C3|C3 Super Mini|[`quadcopter.dev/flix.esp32c3.merged.bin`](https://quadcopter.dev/flix.esp32c3.merged.bin)|
|Flix2|Flix2 board|[`quadcopter.dev/flix.flix2.merged.bin`](https://quadcopter.dev/flix.flix2.merged.bin)|
<!-- markdownlint-enable MD044 -->
2. Flash your ESP32 board using [ESP32 Web Flasher](https://www.espboards.dev/tools/program/):
<img src="img/web-flasher.png" width="400">
* Connect the board to your computer, press *Connect to ESP*, choose the serial port.
* Go to the *Flash* tab.
* Choose the downloaded firmware file, set *Flash address* to *0* (important).
* Click *Program* button and wait until the process is finished.
### Building from sources (flexible)
You can build and upload the firmware using either **Arduino IDE** (easier for beginners) or **command line**.
Get the sources using git:
```bash
git clone https://github.com/okalachev/flix.git
cd flix
git clone https://github.com/okalachev/flix.git && cd flix
```
## Simulation
Beginners can [download the sources as a ZIP archive](https://github.com/okalachev/flix/archive/refs/heads/master.zip).
### Ubuntu
#### Arduino IDE (Windows, Linux, macOS)
The latest version of Ubuntu supported by Gazebo 11 simulator is 20.04. If you have a newer version, consider using a virtual machine.
1. Install Arduino CLI:
```bash
curl -fsSL https://raw.githubusercontent.com/arduino/arduino-cli/master/install.sh | BINDIR=~/.local/bin sh
```
2. Install Gazebo 11:
```bash
sudo sh -c 'echo "deb http://packages.osrfoundation.org/gazebo/ubuntu-stable `lsb_release -cs` main" > /etc/apt/sources.list.d/gazebo-stable.list'
wget https://packages.osrfoundation.org/gazebo.key -O - | sudo apt-key add -
sudo apt-get update
sudo apt-get install -y gazebo11 libgazebo11-dev
```
Set up your Gazebo environment variables:
```bash
echo "source /usr/share/gazebo/setup.sh" >> ~/.bashrc
source ~/.bashrc
```
3. Install SDL2 and other dependencies:
```bash
sudo apt-get update && sudo apt-get install build-essential libsdl2-dev
```
4. Add your user to the `input` group to enable joystick support (you need to re-login after this command):
```bash
sudo usermod -a -G input $USER
```
5. Run the simulation:
```bash
make simulator
```
### macOS
1. Install Homebrew package manager, if you don't have it installed:
```bash
/bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/HEAD/install.sh)"
```
2. Install Arduino CLI, Gazebo 11 and SDL2:
```bash
brew tap osrf/simulation
brew install arduino-cli
brew install gazebo11
brew install sdl2
```
Set up your Gazebo environment variables:
```bash
echo "source /opt/homebrew/share/gazebo/setup.sh" >> ~/.zshrc
source ~/.zshrc
```
3. Run the simulation:
```bash
make simulator
```
### Setup
#### Control with smartphone
1. Install [QGroundControl mobile app](https://docs.qgroundcontrol.com/master/en/qgc-user-guide/getting_started/download_and_install.html#android) on your smartphone. For **iOS**, use [QGroundControl build from TAJISOFT](https://apps.apple.com/ru/app/qgc-from-tajisoft/id1618653051).
2. Connect your smartphone to the same Wi-Fi network as the machine running the simulator.
3. If you're using a virtual machine, make sure that its network is set to the **bridged** mode with Wi-Fi adapter selected.
4. Run the simulation.
5. Open QGroundControl app. It should connect and begin showing the virtual drone's telemetry automatically.
6. Go to the settings and enable *Virtual Joystick*. *Auto-Center Throttle* setting **should be disabled**.
7. Use the virtual joystick to fly the drone!
#### Control with USB remote control
1. Connect your USB remote control to the machine running the simulator.
2. Run the simulation.
3. Calibrate the RC using `cr` command in the command line interface.
4. Run the simulation again.
5. Use the USB remote control to fly the drone!
## Firmware
### Arduino IDE (Windows, Linux, macOS)
<img src="img/arduino-ide.png" width="400" alt="Flix firmware open in Arduino IDE">
1. Install [Arduino IDE](https://www.arduino.cc/en/software) (version 2 is recommended).
2. Windows users might need to install [USB to UART bridge driver from Silicon Labs](https://www.silabs.com/developers/usb-to-uart-bridge-vcp-drivers).
3. Install ESP32 core, version 3.2.0. See the [official Espressif's instructions](https://docs.espressif.com/projects/arduino-esp32/en/latest/installing.html#installing-using-arduino-ide) on installing ESP32 Core in Arduino IDE.
2. *Windows users might need to install [USB to UART bridge driver from Silicon Labs](https://www.silabs.com/developers/usb-to-uart-bridge-vcp-drivers).*
3. Install ESP32 core, version 3.3.10. See the [official Espressif's instructions](https://docs.espressif.com/projects/arduino-esp32/en/latest/installing.html#installing-using-arduino-ide) on installing ESP32 Core in Arduino IDE.
4. Install the following libraries using [Library Manager](https://docs.arduino.cc/software/ide-v2/tutorials/ide-v2-installing-a-library):
* `FlixPeriph`, the latest version.
* `MAVLink`, version 2.0.16.
5. Clone the project using git or [download the source code as a ZIP archive](https://codeload.github.com/okalachev/flix/zip/refs/heads/master).
6. Open the downloaded Arduino sketch `flix/flix.ino` in Arduino IDE.
7. Connect your ESP32 board to the computer and choose correct board type in Arduino IDE (*WEMOS D1 MINI ESP32* for ESP32 Mini) and the port.
* `MAVLink`, version 2.0.25.
5. Open the `flix/flix.ino` sketch from downloaded firmware sources in Arduino IDE.
6. Connect your ESP32 board to the computer and choose correct board type in Arduino IDE (*WEMOS D1 MINI ESP32* for ESP32 Mini, *ESP32S3 Dev Module* for ESP32-S3 Super Mini) and the port.
7. Set *Tools**Core Debug Level* to *Error* to see the errors in the serial console. Set *Tools**USB CDC on Boot* to *Enabled* for ESP32-S3/ESP32-C3 boards.
8. [Build and upload](https://docs.arduino.cc/software/ide-v2/tutorials/getting-started/ide-v2-uploading-a-sketch) the firmware using Arduino IDE.
### Command line (Windows, Linux, macOS)
#### Command line (Windows, Linux, macOS)
1. [Install Arduino CLI](https://arduino.github.io/arduino-cli/installation/).
On Linux, use:
On Linux, install it like this:
```bash
curl -fsSL https://raw.githubusercontent.com/arduino/arduino-cli/master/install.sh | BINDIR=~/.local/bin sh
@@ -149,19 +86,163 @@ The latest version of Ubuntu supported by Gazebo 11 simulator is 20.04. If you h
make upload monitor
```
See other available Make commands in the [Makefile](../Makefile).
For ESP32-S3/ESP32-C3 boards, set the appropriate [FQBN](https://docs.arduino.cc/arduino-cli/FAQ/#whats-the-fqbn-string) using `BOARD` parameter:
```bash
make BOARD=esp32:esp32:esp32s3:FlashSize=4M,CDCOnBoot=cdc upload
```
See other available Make commands in [Makefile](../Makefile).
> [!TIP]
> You can test the firmware on a bare ESP32 board without connecting IMU and other peripherals. The Wi-Fi network `flix` should appear and all the basic functionality including CLI and QGroundControl connection should work.
> You can test the firmware on a bare ESP32 board without connecting IMU and other peripherals. The Wi-Fi network `flix` should appear and all the basic functionality including console and QGroundControl connection should work.
### Setup
## Before first flight
### Connect using QGroundControl
QGroundControl is a ground control station software that can be used to monitor and control the drone.
1. Install mobile or desktop version of [QGroundControl](https://docs.qgroundcontrol.com/master/en/qgc-user-guide/getting_started/download_and_install.html).
2. Power up the drone.
3. Connect your computer or smartphone to the appeared `flix` Wi-Fi network (password: `flixwifi`).
4. Launch QGroundControl app. It should connect and begin showing the drone's telemetry automatically.
> [!TIP]
> If QGroundControl doesn't connect, try to disable the firewall and/or VPN on your computer, as they may block the connection.
### Access console
The console is a command line interface (CLI) that allows to interact with the drone, change parameters, and perform various actions. There are two ways of accessing the console: using **serial port** or using **QGroundControl (wirelessly)**.
To access the console using serial port:
1. Connect the ESP32 board to the computer using USB cable.
2. Open Serial Monitor in Arduino IDE (or use `make monitor` in the command line).
3. In Arduino IDE, make sure the baudrate is set to 115200.
To access the console using QGroundControl:
1. Connect to the drone using QGroundControl app.
2. Go to the QGroundControl menu ⇒ *Analyze Tools* ⇒ *MAVLink Console*.
<img src="img/cli.png" width="400">
> [!TIP]
> Use `help` command to see the list of available commands.
### Access parameters
The drone is configured using parameters. To access and modify them, go to the QGroundControl menu ⇒ *Vehicle Setup* ⇒ *Parameters*:
<img src="img/parameters.png" width="400">
You can also work with parameters using `p` command in the console. Parameter names are case-insensitive.
### Configure the IMU
1. Configure the following parameters for the IMU:
* `IMU_MODEL` — IMU model (1 for MPU-9250/MPU-6500, 2 for ICM-20948, 3 for MPU-6050, 4 for ICM-40609-D).
* `IMU_BUS` — communication bus (0 for SPI, 1 for I²C).
* `IMU_PIN_SCK`, `IMU_PIN_MISO`, `IMU_PIN_MOSI`, `IMU_PIN_CS` — SPI pin numbers.
* `IMU_PIN_SCL`, `IMU_PIN_SDA` — I²C pin numbers.
* `IMU_PIN_INT` — IMU data ready pin number (-1 if not used).
2. Reboot the drone.
3. Check the IMU is working using `imu` command in the console (should print `status: OK`).
### Define IMU orientation
The IMU orientation (relative to the drone's axes) is defined using the parameters: `IMU_ROT_ROLL`, `IMU_ROT_PITCH`, and `IMU_ROT_YAW`.
The drone has *X* axis pointing forward, *Y* axis pointing left, and *Z* axis pointing up, and the supported IMU boards have *X* axis pointing to the mounting holes side and *Z* axis pointing up from the component side:
<img src="img/imu-axes.png" width="200">
Use the following table to set the parameters for common IMU orientations:
|Orientation|Parameters|Orientation|Parameters|
|:-:|-|-|-|
|<img src="img/imu-rot-3.png" width="180">|`IMU_ROT_ROLL` = 0<br>`IMU_ROT_PITCH` = 0<br>`IMU_ROT_YAW` = 0 |<img src="img/imu-rot-7.png" width="180">|`IMU_ROT_ROLL` = 3.142<br>`IMU_ROT_PITCH` = 0<br>`IMU_ROT_YAW` = 0|
|<img src="img/imu-rot-2.png" width="180">|`IMU_ROT_ROLL` = 0<br>`IMU_ROT_PITCH` = 0<br>`IMU_ROT_YAW` = -1.571|<img src="img/imu-rot-6.png" width="180">|`IMU_ROT_ROLL` = 3.142<br>`IMU_ROT_PITCH` = 0<br>`IMU_ROT_YAW` = -1.571|
|<img src="img/imu-rot-1.png" width="180">|`IMU_ROT_ROLL` = 0<br>`IMU_ROT_PITCH` = 0<br>`IMU_ROT_YAW` = 3.142|<img src="img/imu-rot-5.png" width="180">|`IMU_ROT_ROLL` = 3.142<br>`IMU_ROT_PITCH` = 0<br>`IMU_ROT_YAW` = 3.142|
|<img src="img/imu-rot-4.png" width="180"><br>☑️ **Default**|<br>`IMU_ROT_ROLL` = 0<br>`IMU_ROT_PITCH` = 0<br>`IMU_ROT_YAW` = 1.571|<img src="img/imu-rot-8.png" width="180">|`IMU_ROT_ROLL` = 3.142<br>`IMU_ROT_PITCH` = 0<br>`IMU_ROT_YAW` = 1.571|
### Calibrate accelerometer
Before flight you need to calibrate the accelerometer:
1. Open Serial Monitor in Arduino IDE (or use `make monitor` command in the command line).
1. Access the console using QGroundControl (recommended) or Serial Monitor.
2. Type `ca` command there and follow the instructions.
#### Control with smartphone
### Setup motors
If using non-default motor pins, set the pin numbers using the parameters: `MOTOR_PIN_FL`, `MOTOR_PIN_FR`, `MOTOR_PIN_RL`, `MOTOR_PIN_RR` (front-left, front-right, rear-left, rear-right respectively).
#### Brushless motors
If using brushless motors with ESCs:
1. Set the appropriate PWM using the parameters: `MOT_PWM_STOP`, `MOT_PWM_MIN`, and `MOT_PWM_MAX` (1000, 1000, and 2000 is typical).
2. Decrease the PWM frequency using the `MOT_PWM_FREQ` parameter (400 is typical).
> [!CAUTION]
> **Remove the props when configuring the motors!** If improperly configured, you may not be able to stop them.
### Battery voltage monitoring (optional)
ESP32 ADC can measure only up to 3.3 V, so you need to use a voltage divider to monitor the battery voltage. To enable voltage measurement, set the following parameters:
1. `PWR_VOLT_PIN` — GPIO pin number where the voltage divider is connected (*-1* to disable).
2. `PWR_VOLT_SCALE` — voltage divider coefficient (*2* for two equal resistors).
After this setup, you should see the battery voltage in QGroundControl top panel or using `pw` command in the console.
### Important: check everything works
1. Check the IMU is working: perform `imu` command in the console and check the output:
* The `status` field should be `OK`.
* The `rate` field should be about 1000 (Hz).
* The `accel` and `gyro` fields should change as you move the drone.
* The `accel bias` and `accel scale` fields should contain calibration parameters (not zeros and ones).
* The `gyro bias` field should contain estimated gyro bias (not zeros).
* The `landed` field should be `1` when the drone is still on the ground and `0` when you lift it up.
2. Check the attitude estimation: connect to the drone using QGroundControl, rotate the drone in different orientations and check if the attitude estimation shown in QGroundControl is correct. Compare your attitude indicator (in the *large vertical* mode) to the video:
<a href="https://youtu.be/yVRN23-GISU"><img width=300 src="https://i3.ytimg.com/vi/yVRN23-GISU/maxresdefault.jpg"></a>
3. Perform motor tests. Use the following commands **— remove the propellers before running the tests!**
* `mfr` — rotate front right motor (counter-clockwise).
* `mfl` — rotate front left motor (clockwise).
* `mrl` — rotate rear left motor (counter-clockwise).
* `mrr` — rotate rear right motor (clockwise).
Make sure rotation directions and propeller types match the following diagram:
<img src="img/motors.svg" width=200>
> [!WARNING]
> Never run the motors when powering the drone from USB, always use the battery for that.
## Setup remote control
There are several ways to control the drone's flight: using **smartphone** (Wi-Fi), using **SBUS remote control**, or using **USB remote control** (Wi-Fi/ESP-NOW).
### Control with a smartphone
#### Using Mavlink Joystick app (Android)
<img src="https://github.com/goldarte/mavlink-joystick/blob/master/app_screen.png?raw=true" width="400">
1. Download and install [Mavlink Joystick app](https://github.com/goldarte/mavlink-joystick/releases/latest).
2. Power the drone using the battery.
3. Connect your smartphone to the appeared `flix` Wi-Fi network (password: `flixwifi`).
4. Open Mavlink Joystick app. It should connect and begin showing the drone's telemetry automatically.
5. Use the virtual joystick to fly the drone!
#### Using QGroundControl app
1. Install [QGroundControl mobile app](https://docs.qgroundcontrol.com/master/en/qgc-user-guide/getting_started/download_and_install.html#android) on your smartphone.
2. Power the drone using the battery.
@@ -171,17 +252,19 @@ Before flight you need to calibrate the accelerometer:
6. Use the virtual joystick to fly the drone!
> [!TIP]
> Decrease `TILT_MAX` parameter when flying using the smartphone to make the controls less sensitive.
> Decrease `CTL_TILT_MAX` parameter when flying using the smartphone to make the controls less sensitive.
#### Control with remote control
### Control with a remote control
Before flight using remote control, you need to calibrate it:
If using SBUS-connected remote control you need to enable SBUS and calibrate it:
1. Open Serial Monitor in Arduino IDE (or use `make monitor` command in the command line).
2. Type `cr` command there and follow the instructions.
3. Use the remote control to fly the drone!
1. Connect to the drone using QGroundControl.
2. In parameters, set the `RC_RX_PIN` parameter to the GPIO pin number where the SBUS signal is connected, for example: 4. Negative value disables SBUS.
3. Check if the receiver is working using `rc` command in the console.
4. Open the console, type `cr` command and follow the instructions to calibrate the remote control.
5. Use the remote control to fly the drone!
#### Control with USB remote control (Wi-Fi)
### Control with a USB remote control
If your drone doesn't have RC receiver installed, you can use USB remote control and QGroundControl app to fly it.
@@ -190,12 +273,9 @@ If your drone doesn't have RC receiver installed, you can use USB remote control
3. Power up the drone.
4. Connect your computer to the appeared `flix` Wi-Fi network (password: `flixwifi`).
5. Launch QGroundControl app. It should connect and begin showing the drone's telemetry automatically.
6. Go the the QGroundControl menu ⇒ *Vehicle Setup**Joystick*. Calibrate you USB remote control there.
6. Go to the QGroundControl menu ⇒ *Vehicle Setup* ⇒ *Joystick*. Calibrate your USB remote control there.
7. Use the USB remote control to fly the drone!
> [!NOTE]
> If something goes wrong, go to the [Troubleshooting](troubleshooting.md) article.
## Flight
For both virtual sticks and a physical joystick, the default control scheme is left stick for throttle and yaw and right stick for pitch and roll:
@@ -214,13 +294,16 @@ When finished flying, **disarm** the drone, moving the left stick to the bottom
<img src="img/disarming.svg" width="150">
> [!NOTE]
> If something goes wrong, go to the [Troubleshooting](troubleshooting.md) article.
### Flight modes
Flight mode is changed using mode switch on the remote control or using the command line.
Flight mode is changed using mode switch on the remote control (if configured) or using the console commands. The main flight mode is *STAB*. In order to change modes using SBUS remote control, set the parameters: `CTL_FLT_MODE_0`, `CTL_FLT_MODE_1`, and `CTL_FLT_MODE_2` to required mode numbers (0 for *RAW*, 1 for *ACRO*, 2 for *STAB*, 3 for *AUTO*).
#### STAB
The default mode is *STAB*. In this mode, the drone stabilizes its attitude (orientation). The left stick controls throttle and yaw rate, the right stick controls pitch and roll angles.
In this mode, the drone stabilizes its attitude (orientation). The left stick controls throttle and yaw rate, the right stick controls pitch and roll angles.
> [!IMPORTANT]
> The drone doesn't stabilize its position, so slight drift is possible. The pilot should compensate it manually.
@@ -229,24 +312,87 @@ The default mode is *STAB*. In this mode, the drone stabilizes its attitude (ori
In this mode, the pilot controls the angular rates. This control method is difficult to fly and mostly used in FPV racing.
#### MANUAL
#### RAW
Manual mode disables all the stabilization, and the pilot inputs are passed directly to the motors. This mode is intended for testing and demonstration purposes only, and basically the drone **cannot fly in this mode**.
*RAW* mode disables all the stabilization, and the pilot inputs are mixed directly to the motors. The IMU sensor is not involved. This mode is intended for testing and demonstration purposes only, and basically the drone **cannot fly in this mode**.
#### AUTO
In this mode, the pilot inputs are ignored (except the mode switch, if configured). The drone can be controlled using [pyflix](../tools/pyflix/) Python library, or by modifying the firmware to implement the needed autonomous behavior.
In this mode, the pilot inputs are ignored (except the mode switch). The drone can be controlled using [pyflix](../tools/pyflix/) Python library, or by modifying the firmware to implement the needed behavior.
If the pilot moves the control sticks, the drone will switch back to *STAB* mode.
If the pilot moves the control sticks and mode switch is not configured, the drone will switch back to *STAB* mode.
## Adjusting parameters
## Wi-Fi configuration
You can adjust some of the drone's parameters (include PID coefficients) in QGroundControl app. In order to do that, go to the QGroundControl menu ⇒ *Vehicle Setup**Parameters*.
You can configure the Wi-Fi using parameters and console commands.
<img src="img/parameters.png" width="400">
The Wi-Fi mode is chosen using `WIFI_MODE` parameter in QGroundControl or in the console:
## CLI access
* `0` — Wi-Fi is disabled.
* `1` — Access Point mode *(AP)* — the drone creates a Wi-Fi network.
* `2` — Client mode *(STA)* — the drone connects to an existing Wi-Fi network (may cause additional delays, so generally not recommended).
* `3` — ESP-NOW mode — the drone uses ESP-NOW protocol for communication.
In addition to accessing the drone's command line interface (CLI) using the serial port, you can also access it with QGroundControl using Wi-Fi connection. To do that, go to the QGroundControl menu ⇒ *Vehicle Setup**Analyze Tools**MAVLink Console*.
The SSID and password are configured using the `ap` and `sta` console commands:
<img src="img/cli.png" width="400">
```
ap <ssid> <password>
sta <ssid> <password>
```
Example of configuring the Access Point mode:
```
ap my-flix-ssid mypassword123
p WIFI_MODE 1
```
Disabling Wi-Fi:
```
p WIFI_MODE 0
```
### Using ESP-NOW
[ESP-NOW](https://docs.espressif.com/projects/esp-idf/en/stable/esp32/api-reference/network/esp_now.html) is a low level wireless communication protocol. It can provide lower latency, better reliability, and longer range than Wi-Fi. However, it requires a second ESP32 board to be used as a proxy for the computer.
<img src="img/espnow-connection.jpg" width="600">
To setup ESP-NOW communication:
1. Flash the second ESP32 board with ESP-NOW proxy sketch: [`tools/espnow-proxy/espnow-proxy.ino`](../tools/espnow-proxy/espnow-proxy.ino). Use Arduino IDE or command line: `make upload_proxy`.
2. Open Serial Monitor in Arduino IDE or use `make monitor` command. The ESP32 will print its MAC address and generated encryption key, for example:
```
espnow 7a:c8:e3:eb:bf:e9 &PiuSysxP9+$L&5E
```
Run this line as a console command on each drone you want to bind to this proxy board. [The maximum number](https://github.com/espressif/esp-idf/blob/e95cab4be8fd293e3f3323181e7a2280874da6f7/components/esp_wifi/include/esp_now.h#L32-L33) of simultaneously connected drones is 20 (unencrypted) or 6 (encrypted).
3. Set the `WIFI_MODE` parameter to `3` on the drone:
```
p WIFI_MODE 3
```
4. Go to the QGroundControl menu ⇒ *Application Settings* ⇒ *Comm Links*, add new link with the following settings:
* Name: ESP32.
* Type: Serial.
* Serial Port: choose the port of the proxy ESP32 board, e. g. `/dev/cu.usbserial-0001`.
* Baud Rate: 115200.
5. Click *Save*, click *Connect*. QGroundControl should connect to the drone using ESP-NOW and begin showing the telemetry.
> [!TIP]
> Make sure Arduino IDE is not running when using ESP-NOW proxy board, as it may block the serial port.
## Flight log
After the flight, you can download the flight log wirelessly for analysis. Use the following command on your computer for that:
```bash
make log
```
See more details about log analysis in the [log analysis](log.md) article.
+159 -1
View File
@@ -4,12 +4,170 @@ This page contains user-built drones based on the Flix project. Publish your pro
---
Author: [Oleg1405](https://t.me/Oleg1405).<br>
Description: ESP32 Mini, MPU-6500 IMU, boost converter, BT2.0 power connector, 65 mm props, BetaFPV ELRS Lite Receiver, Radiomaster Pocket + Mavlink Joystick (Android) control.
<img src="img/user/oleg1405/1.jpg" height=300>
[Flight video](https://www.youtube.com/shorts/rbXV4sHbpso).
---
Author: Alican Erüst.<br>
Description: QX95 mm frame, 55 mm propellers, 3.7 V 25C 1050 mAh LiPo battery, MPU6050 IMU, Logitech F310 gamepad controller, with a total quadcopter weight of 66 g.
<img src="img/user/alicanerus/1.jpg" height=200> <img src="img/user/alicanerus/2.jpg" height=200> <img src="img/user/alicanerus/3.jpg" height=200>
[Flight video](https://drive.google.com/file/d/1k0WeWTKnCAfaugkX7LcmNxsUuq79RL8Z/view?usp=sharing).
---
Author: [Неруш Михаил](https://t.me/NerushMV).<br>
Description: custom frame made of 4 mm plywood, 8520 brushed motors, 75 mm propellers, MPU-6500. FlySky FS-i6X with ESP32-based adapter for ESP-NOW communication (using PPM output).<br>
Sources and materials: [link](https://drive.google.com/drive/folders/1uWiDcuorLrtVs_IIR7Y13omij-7Q1nx8).
<img src="img/user/nerush/1.jpg" height=200> <img src="img/user/nerush/2.jpg" height=200>
[Flight video](https://drive.google.com/file/d/1jRXeGx34lJpUfw0GKLQeIzkWZvooQJSE/view?usp=sharing).
---
Author: [Konstantinos Paraskevas](https://github.com/Frapais).<br>
Description: drone with a custom single-boarded airframe, extending the [Sprig-C3 module](https://github.com/Frapais/Sprig-C3).
ESP32-C3 microcontroller, ICM-20948 IMU, on-board fuel-gauge, status LED indicator.<br>
Repository with all the code and PCB sources: https://github.com/Frapais/Sprig-Drone.
<img src="img/user/kostas/1.jpg" height=150> <img src="img/user/kostas/2.jpg" height=150>
Detailed video about making the drone:
<a href="https://youtu.be/82Q-uBq6s48"><img width=400 src="https://i3.ytimg.com/vi/82Q-uBq6s48/maxresdefault.jpg"></a>
---
Author: [Awab Anas](http://t.me/AW_VENOM).<br>
Description: ESP32 D1 Mini, MPU-6050, 8520 3.7V brushed motors, 55 mm propellers, battery li-po 1200 mAh, controlling via [Mavlink Joystick app](https://github.com/goldarte/mavlink-joystick/releases/latest).<br>
[Flight validation](https://drive.google.com/file/d/12z0jfctZDBA6b5UKCG0Uje5rAxj6DhF-/view?usp=sharing).
<img src="img/user/aw_venom/1.jpg" height=200>
---
Author: [Ina Tix](https://t.me/ina_tix).<br>
Description: XR2981 based DC-DC converter, ELRS MINI 2.4GHz RX SX1280 receiver (SBUS interface), Radiomaster TX12 remote control.<br>
[Flight validation](https://drive.google.com/file/d/1yqkKNuz4R_yxGqUNQxVpixJbXqEEcUSj/view?usp=share_link).
<img src="img/user/ina_tix/1.jpg" height=200> <img src="img/user/ina_tix/2.jpg" height=200> <img src="img/user/ina_tix/3.jpg" height=200>
---
Author: Oleg Kalachev.<br>
Description: the first attempt on making an official PCB based Flix drone (Flix2 board). The IMU is not working on this version, so an external MPU-6050 board was used, therefore considered as **Flix version 1.5**.<br>
[Flight video](https://drive.google.com/file/d/1R7tuUsFmPY0CGcOCFfMFaCp9kR49K3bl/view?usp=sharing).
<img src="img/flix1.5.jpg" width=300>
---
Author: [FanBy0ru](https://https://github.com/FanBy0ru).<br>
Description: custom 3D-printed frame.<br>
Frame STLs and flight validation: https://cults3d.com/en/3d-model/gadget/armature-pour-flix-drone.
<img src="img/user/fanby0ru/1.jpg" height=200> <img src="img/user/fanby0ru/2.jpg" height=200>
---
Author: Ivan44 Phalko.<br>
Description: custom PCB, cusom test bench.<br>
[Flight validation](https://drive.google.com/file/d/17DNDJ1gPmCmDRAwjedCbJ9RXAyqMqqcX/view?usp=sharing).
<img src="img/user/phalko/1.jpg" height=200> <img src="img/user/phalko/2.jpg" height=200> <img src="img/user/phalko/3.jpg" height=200>
---
Author: **Arkadiy "Arky" Matsekh**, Foucault Dynamics, Gold Coast, Australia.<br>
The drone was built for the University of Queensland industry-led Master's capstone project.
**Flight video:**
<a href="https://drive.google.com/file/d/1NNYSVXBY-w0JjCo07D8-PgnVq3ca9plj/view?usp=sharing"><img height=300 src="img/user/arkymatsekh/video.jpg"></a>
<img src="img/user/arkymatsekh/1.jpg" height=150> <img src="img/user/arkymatsekh/2.jpg" height=150> <img src="img/user/arkymatsekh/3.jpg" height=150>
---
Author: [goldarte](https://t.me/goldarte).<br>
<img src="img/user/goldarte/1.jpg" height=150> <img src="img/user/goldarte/2.jpg" height=150>
**Flight video:**
<a href="https://drive.google.com/file/d/1nQtFjEcGGLx-l4xkL5ko9ZpOTVU-WDjL/view?usp=sharing"><img height=200 src="img/user/goldarte/video.jpg"></a>
---
Author: [malagis](https://oshwhub.com/malagis).<br>
A Chinese custom PCB version of Flix with a big community of users, lots of materials and modifications.
Main project's page: https://oshwhub.com/malagis/esp32-mini-plane.<br>
Video about the project: https://www.bilibili.com/video/BV14vyqBFEJn/.
<img src="img/user/malagis/1.jpg" height=200> <img src="img/user/malagis/2.jpg" height=200> <img src="img/user/malagis/3.jpg" height=200>
---
## School 548 course
Special course on quadcopter design and engineering took place in october-november 2025 in School 548, Moscow. The course included UAV control theory, electronics, drone assembly and setup practice, using the Flix project.
<img height=200 src="img/user/school548/1.jpg"> <img height=200 src="img/user/school548/2.jpg"> <img height=200 src="img/user/school548/3.jpg">
STL files and other materials: see [here](https://drive.google.com/drive/folders/1wTUzj087LjKQQl3Lz5CjHCuobxoykhyp?usp=share_link).
### Selected works
Author: [KiraFlux](https://t.me/@kiraflux_0XC0000005).<br>
Description: **custom ESPNOW remote control** was implemented, modified firmware to support ESPNOW protocol.<br>
Telegram posts: [1](https://t.me/opensourcequadcopter/106), [2](https://t.me/opensourcequadcopter/114).<br>
Modified Flix firmware: https://github.com/KiraFlux/flix/tree/klyax.<br>
Remote control project: https://github.com/KiraFlux/ESP32-DJC.<br>
Drone design: https://github.com/KiraFlux/Klyax.<br>
<img src="img/user/school548/kiraflux1.jpg" height=150> <img src="img/user/school548/kiraflux2.jpg" height=150>
**ESPNOW remote control demonstration**:
<img height=200 src="img/user/school548/kiraflux-video.jpg"><a href="https://drive.google.com/file/d/1soHDAeHQWnm97Y4dg4nWevJuMiTdJJXW/view?usp=sharing"></a>
Author: [tolyan4krut](https://t.me/tolyan4krut).<br>
Description: the first drone based on ESP32-S3-CAM board **with a camera**, implementing Wi-Fi video streaming. Runs HTTP server and HTTP video stream.<br>
Modified Flix firmware: https://github.com/CatRey/Flix-Camera-Streaming.<br>
[Telegram post](https://t.me/opensourcequadcopter/117).
<img src="img/user/school548/tolyan4krut.jpg" height=150>
**Video streaming and flight demonstration**:
<a href="https://drive.google.com/file/d/1KuOBsujLsk7q8FoqKD8u7uoq4ptS5onp/view?usp=sharing"><img height=200 src="img/user/school548/tolyan4krut-video.jpg"></a>
Author: [Vlad Tolshinov](https://t.me/Vlad_Tolshinov).<br>
Description: custom frame with enlarged arm length, which provides very high flight stability, 65 mm props.
<img src="img/user/school548/vlad_tolshinov1.jpg" height=150> <img src="img/user/school548/vlad_tolshinov2.jpg" height=150>
**Flight video**:
<a href="https://drive.google.com/file/d/1zu00DZxhC7DJ9Z2mYjtxdNQqOOLAyYbp/view?usp=sharing"><img height=200 src="img/user/school548/vlad_tolshinov-video.jpg"></a>
---
## RoboCamp
Author: RoboCamp participants.<br>
Description: 3D-printed and wooden frames, ESP32 Mini, DC-DC buck-boost converters. BetaFPV LiteRadio 3 to control the drones via Wi-Fi connection.<br>
Features: altitude hold, obstacle avoidance, autonomous flight elements.<br>
Some of the designed model files: https://drive.google.com/drive/folders/18YHWGquKeIevzrMH4-OUT-zKXMETTEUu?usp=share_link.
Some of the designed model files: see [here](https://drive.google.com/drive/folders/18YHWGquKeIevzrMH4-OUT-zKXMETTEUu?usp=share_link).
RoboCamp took place in July 2025, Saint Petersburg, where 9 participants designed and built their own drones using the Flix project, and then modified the firmware to complete specific flight tasks.
+1 -1
View File
@@ -14,7 +14,7 @@ Flix version 0 (obsolete):
|Motor|8520 3.7V brushed motor (**shaft 0.8mm!**)|<img src="img/motor.jpeg" width=100>|4|
|Propeller|Hubsan 55 mm|<img src="img/prop.jpg" width=100>|4|
|Motor ESC|2.7A 1S Dual Way Micro Brush ESC|<img src="img/esc.jpg" width=100>|4|
|RC transmitter|KINGKONG TINY X8|<img src="img/tx.jpg" width=100>|1|
|RC transmitter|KINGKONG TINY X8|<img src="img/kingkong.jpg" width=100>|1|
|RC receiver|DF500 (SBUS)|<img src="img/rx.jpg" width=100>|1|
|~~SBUS inverter~~*||<img src="img/inv.jpg" width=100>|~~1~~|
|Battery|3.7 Li-Po 850 MaH 60C|||