-
Notifications
You must be signed in to change notification settings - Fork 3
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
Simplified readme. The detail moved to documentation.
- Loading branch information
Showing
1 changed file
with
32 additions
and
133 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -1,153 +1,52 @@ | ||
| # TFSLOT01A - Low-speed airspeed sensor for UAVs | ||
| # TFSLOT: Clogging-free airspeed sensor | ||
|
|
||
|  | ||
| TFSLOT is a state-of-the-art airspeed sensor designed specifically for use in Unmanned Aerial Vehicles (UAVs). Utilizing the principle of the [Venturi effect](https://en.wikipedia.org/wiki/Venturi_effect) with a differential pressure sensor for accurate airspeed measurement, this device is also equipped with an integrated [Inertial Measurement Unit](https://en.wikipedia.org/wiki/Inertial_measurement_unit) (IMU). The unique design and features of TFSLOT01 solve common issues associated with Pitot tubes, offering a robust solution for aerial applications. | ||
|
|
||
| [TFSLOT](https://github.com/ThunderFly-aerospace/TFSLOT01) sensor is an airspeed sensor designed to be used on UAVs to measure [IAS](https://en.wikipedia.org/wiki/Indicated_airspeed). Due to the 3D printed case, it is possible to optimize the characteristics according to the location of the sensor on the UAV and specific application requirements on measuring range. The first use of this sensor is on ThunderFly autogyro [TF-G2](https://github.com/ThunderFly-aerospace/TF-G2/). | ||
| <p align="center"> | ||
| <img src="doc/img/TFSLOT_1_small.jpg" /> | ||
| </p> | ||
|
|
||
| ## Main Features | ||
| ## TFSLOT01 features | ||
| - **High Resolution** at low airspeeds (below 10 m/s) | ||
| - **Configurable Sensitivity** by modifying the airfoil profile | ||
| - **Reduced Susceptibility to Clogging**, for instance, by mud or snow upon landing | ||
| - **Weatherproof**, it can be used in various conditions | ||
| - **Direct Differential Pressure Sensor Integration** without the need for additional tubing | ||
| - **Possibility of Direct Integration** into the drone's mechanical structure | ||
| - **Open Design** making the design documentation accessible | ||
| - **Integrated IMU Unit** for measuring vibrations, angle of attack, and more | ||
|
|
||
| * Improved low-speed resolution compared to traditional pitot tubes. | ||
| * Reduced tendency for clogging due to the absence of a stagnation point. | ||
| * Directly integrated differential pressure sensor, eliminating the need for additional tubing. | ||
| * Customizable 3D-printed case to fit specific applications. | ||
| * Integrated IMU to vibration diagnostics on AoA measurement. | ||
|
|
||
| TFSLOT is commercially available from [ThunderFly s.r.o.](https://www.thunderfly.cz/), write an email to info@thunderfly.cz or shop at [Tindie store](https://www.tindie.com/products/thunderfly/tfslot01a-drone-indicated-airspeed-sensor/). | ||
| ## How It Works? | ||
| The TFSLOT01 sensor uses the Venturi effect to measure airspeed, allowing precise tracking of airspeed even at low speeds. | ||
|
|
||
| ## Technical Parameters | ||
| <p align="center"> | ||
| <img src="/doc/img/tfslot_crossection.svg" /> | ||
| </p> | ||
|
|
||
| | Parameter | Value | Description | | ||
| |-----------|-------|-------------| | ||
| | Airspeed measurement range | 0 - 48 m/s | Assumes air density 1.29 kg/m³, negative airspeeds are also indicated but scaling is invalid | | ||
| | I2C Connector | 4-pin JST-GH | [Pixhawk connector standard](https://github.com/pixhawk/Pixhawk-Standards/blob/master/DS-009%20Pixhawk%20Connector%20Standard.pdf) | | ||
| | I2C address | 0x21 SDP3x sensor, 0x68 IMU | SDP3x sensor is visible only after proper init of IMU | | ||
| | I2C SCL clock frequency | Typ. 400 kHz | Operation above 400 kHz is possible, but unreliable| | ||
| | Operating and storage temperature | −20°C to +40°C | Limited by case material | | ||
| | Operational input voltage | +3.6V to +5.4V | Overvoltage internally protected by Zener diode, Undervoltage is not treated | | ||
| | Mass | 25g | Using the 3D printed case gcode included | | ||
| | Dimensions | 35x40x35mm | Default 3D printed Case | | ||
| | Weather resistance | [IP42](https://en.wikipedia.org/wiki/IP_Code) | External connectors fully occupied | | ||
|
|
||
| The TFSLOT uses the [TFASPDIMU02 electronics](https://github.com/ThunderFly-aerospace/TFASPDIMU02) which has many more options for how can be used, please look at its documentation for more details and parameters. | ||
| Air at a certain airspeed enters through an opening between two precisely defined airfoil profiles. As the depth within the sensor increases, the effective cross-section area decreases. Since the same volume of air must pass through a smaller cross-section, the flow velocity inside the sensor increases. As a result, based on the law of conservation of energy, the pressure decreases. | ||
|
|
||
| ## Connection to the Pixhawk-based autopilots | ||
| Inside the sensor, there are two sampling points connected to the differential pressure sensor by a channel. The measured pressure difference is then recalculated into the airspeed at the sensor's inlet. In addition to airspeed sensing, the sensor includes an integrated IMU unit that can also serve as an external magnetometer, because airspeed sensors are usually placed away from electromagnetic interference sources. | ||
|
|
||
| The I²C interface connectors respect the [Pixhawk connector standard](https://github.com/pixhawk/Pixhawk-Standards/blob/master/DS-009%20Pixhawk%20Connector%20Standard.pdf). The signal and color coding of the connector and supplied cable are described by following table (ThunderFly color scheme): | ||
| <p align="center"> | ||
| <img src="/doc/img/TFSLOT_4_small.jpg" /> | ||
| </p> | ||
|
|
||
| |Signal | Pixhawk Color | ThunderFly color | | ||
| |--------|------------------|---------------------| | ||
| | +5V | Red |  Red | | ||
| | SCL | Black |  Yellow | | ||
| | SDA | Black |  Green | | ||
| | GND | Black |  Black | | ||
| ## Where to get TFSLOT01? | ||
|
|
||
| The conductor colors in the cable are different from the Pixhawk standard to increase the visual distinction between multiple cables in the UAV. | ||
| <a href="https://www.tindie.com/products/thunderfly/tfslot01a-airspeed-sensor-with-integrated-imu/"><img src="https://d2ss6ovg47m0r5.cloudfront.net/badges/tindie-mediums.png" alt="I sell on Tindie" width="150" height="78" align="right"></a> | ||
|
|
||
| ### I2C Cable Twisting | ||
| TFSLOT can be bought directly from us via our contact [email](https://www.thunderfly.cz/contact-us.html). The email can also be used if there are specific requirements for custom modifications or if the product will be inquired in large quantities. TFSLOT is also available for online purchase through the [Tindie store](https://www.tindie.com/products/thunderfly/tfslot01a-airspeed-sensor-with-integrated-imu/). | ||
|
|
||
| To improve I2C bus reliability, the supplied cable is specifically twisted by following the scheme | ||
|
|
||
| - 10 turns for each pair SCL/+5V and SDA/GND per 30cm cable length | ||
| - The two pairs are turned again by 4 turns of pairs per 30cm cable length. | ||
| ## Support and Integration | ||
|
|
||
| These special cable conductors' winding method greatly improves the signal integrity by minimizing the crosstalk between the SDA and SCL signals. | ||
| ### Drone avionics | ||
| The TFSLOT01A sensor (PCB board [TFASPDIMU02A](https://github.com/ThunderFly-aerospace/TFASPDIMU02)) is currently supported by the PX4 autopilot, where it can be utilized both as an airspeed sensor and as an external IMU unit (external magnetometer). Instructions for integrating the sensor with the PX4 autopilot can be found in the official [PX4 documentation](http://docs.px4.io/master/en/sensor/airspeed.html#airspeed-sensors). The sensor is also hardware-compatible with other autopilots, such as Ardupilot, but currently lacks the computation code to convert differential pressure into airspeed. | ||
|
|
||
|
|
||
| ## Working Principle | ||
|
|
||
| The sensor basically uses the [Venturi Effect](https://en.wikipedia.org/wiki/Venturi_effect#Instrumentation_and_measurement) to measure fly velocity in the air. | ||
|
|
||
| The velocity and pressure of the fluid are required to meet [Bernoulli's principle](https://en.wikipedia.org/wiki/Bernoulli%27s_principle) | ||
|
|
||
|  | ||
|
|
||
| <!-- source: \frac{1}{2}\rho{v_\infty}^{2}+p_\infty=\frac{1}{2}\rho{v}^{2}+p --> | ||
|
|
||
| Then the velocities are in relation to cross sections in the plane of pressure measurement ports | ||
|
|
||
|  | ||
|
|
||
| <!-- source: \frac{v}{v_\infty}=\frac{A_D}{A_d} --> | ||
|
|
||
| Therefore equations for pressure difference and airspeed velocity could be derived. | ||
|
|
||
| ^2-1\right]\Rightarrow&space;&space;v_\infty=\sqrt{\frac{2\Delta&space;p}{\rho\left[\left(\frac{A_D}{A_d}\right)^2-1\right]}}) | ||
|
|
||
| <!-- source: \Delta p=\frac{1}{2}\rho{v_\infty}^{2}\left[\left(\frac{v}{v_\infty}\right)^2-1\right]\Rightarrow v_\infty=\sqrt{\frac{2\Delta p}{\rho\left[\left(\frac{A_D}{A_d}\right)^2-1\right]}} --> | ||
|
|
||
| Where | ||
|
|
||
| | Symbol | Explanation | | ||
| |--------|------------------| | ||
| | | Measured pressure difference | | ||
| |  | Air density | | ||
| | | Free air velocity e.g. measured airspeed | | ||
| |  | The velocity of air running through the sensor | | ||
| |  | Cross-section area at the position of the outer pressure port | | ||
| |  | Cross-section area at the position of internal pressure port | | ||
|
|
||
| ### Comparison to pitot-static tube | ||
|
|
||
| In contrast to classical [Pitot tube](https://en.wikipedia.org/wiki/Pitot_tube), the TFSLOT design is the perfect choice for low airspeed measurement (generally below 10 m/s). | ||
| The design has the following advantages compared to a pitot probe: | ||
|
|
||
| * Better resolution on low airspeeds (where kinetic pressure is small) | ||
| * Less tendency to clogging (due to the absence of stagnation point) | ||
| * Direct differential pressure sensor integration without any additional tubing | ||
|
|
||
| The improved resolution on low airspeeds is the principal difference from the pitot-static tube. This could be explained by comparing it with the classical equation for pitot-static tube: | ||
|
|
||
|  | ||
|
|
||
| The equation for TFSLOT venturi-based sensor (see equations above) differs from it by a factor | ||
|
|
||
| ^2-1) | ||
|
|
||
| That means the TFSLOT sensor is more sensitive (e.g. has a higher measurable pressure difference at the same airspeed) than the pitot-static tube (at the same air density) in every configuration where | ||
|
|
||
|  | ||
|
|
||
| The advantage of the higher sensitivity results in increased drag, but it could be negligible at low airspeeds. The best use case of the sensor is the integration of the sensing device into the relatively massive fuselage or other construction geometry. The user is supposed to find the best mounting option for the device or adapt the TFSLOT 3D printed case to fit the specific application. From that perspective, the TFSLOT01 is a fully functional development kit that is able to cover specific user needs. | ||
|
|
||
| # Usage | ||
|
|
||
| TFSLOT is equipped with [TFASPDIMU](https://github.com/ThunderFly-aerospace/TFASPDIMU02) electronics, which contains the required differential pressure sensor and an IMU unit through which the pressure sensor is connected. To activate I2C access to it, it is necessary to set the pass-through mode in the IMU unit. The [TFASPDIMU02](https://github.com/ThunderFly-aerospace/TFASPDIMU02) has many options for how can be used, please look at its documentation for more details. | ||
|
|
||
| ## PX4 autopilot | ||
|
|
||
| PX4 is able to initialize the sensor itself. All you have to do is add the following two lines to the [startup configuration file](https://docs.px4.io/master/en/concept/system_startup.html#replacing-the-system-startup) on the Autopilot's SD card. | ||
|
|
||
| ``` | ||
| icm20948_i2c_passthrough start -X -b 2 -a 0x68 | ||
| sdp3x start -X | ||
| ``` | ||
|
|
||
| The script assumes a connection to port _I2C 2_. This was set by the `-b` parameter in both commands. The correct bus has to be set by the user. | ||
|
|
||
| > The model for converting to the airspeed from an air pressure is not yet in the stable version of the PX4 autopilot source code. Therefore the master upstream needs to be used with the sensor. | ||
| Sensor requires firmware with modifications from the [`tf/aspdimu`](https://github.com/ThunderFly-aerospace/PX4Firmware/tree/tf/aspdimu) branch. The GitHub [pull-request](https://github.com/PX4/PX4-Autopilot/pull/18593) into the PX4 master is currently open to solve this issue. The airspeed recalculation model should be included in the next stable release. Alternatively, you can temporarily applicate [this PR changes](https://github.com/PX4/PX4-Autopilot/compare/master...ThunderFly-Aerospace:tf/aspdimu) or [apply patch](https://patch-diff.githubusercontent.com/raw/PX4/PX4-Autopilot/pull/18593.patch). | ||
|
|
||
| ### Configuration | ||
|
|
||
| TFSLOT01 contains a specific aerodynamic profile. Therefore it requires a different converting model. This is done by setting the parameter `CAL_AIR_CMODEL` to 3 (Venturi effect-based airspeed sensor). Setting the wrong profile causes the measurement of negative airspeed values, which results in an `airspeed sensor failure` condition during the attempt to fly. | ||
|
|
||
| ### PX4 Calibration process | ||
|
|
||
|  | ||
|
|
||
| The sensor itself is calibrated and the PX4 calibration process aims to verify that the sensor is properly mounted and connected. Thanks to the used sensor, this sensor is not sensitive to temperature changes and has zero offsets. It is enough to calibrate (verify) the sensor once it is mounted into the UAV. And there is no need to do it before each flight (as with other types of airspeed sensors). | ||
|
|
||
| The procedure is performed with the help of a calibration tool in a few steps: | ||
| 1. Place the calibration tool in the TFSLOT inlet | ||
| 1. Press the tool against the sensor (green arrow) with your hand or rubber band and hold it there during the whole process. | ||
| 1. Start the [callibration process](https://docs.px4.io/master/en/config/airspeed.html#performing-the-calibrationhttps://docs.px4.io/master/en/config/airspeed.html#performing-the-calibration) in PX4 | ||
| 1. Wait for the measure of static offset (no blowing) | ||
| 1. PX4 will command you to blow into the airspeed sensor | ||
| 1. Blow from the back of the sensor (blue arrow). If you blow too little, repeat the attempt. | ||
| 1. PX4 calibration should now be completed, confirmed by a beep, and in the QGC. | ||
|
|
||
|  | ||
| ### Robotics and measurement | ||
| For other applications, the sensor can be easily connected to any computer with an I2C interface, such as RaspberryPi or a generic desktop with a USB to I2C converter (e.g., [USBI2C02A](https://www.mlab.cz/module/USBI2C01/), [shop it on tindie](https://www.tindie.com/products/mlab-project/mlab-usbi2c01a-usb-to-smbusi2c-bridge/)), and the differential pressure values or values from the IMU unit can be read, for example, by a Python script. | ||
|
|
||
| # FAQ | ||
| Frequently asked questions are listed on [TFSLOT's documentation page](https://docs.thunderfly.cz/avionics/TFSLOT01#faq) |