Black Box

by Just4Fun Media in Circuits > Tools

4177 Views, 87 Favorites, 0 Comments

Black Box

Slide1.jpg
5677a4d2937ddb6ebe0001af.jpeg
photo-2015-12-17, 6:36 PM.jpg

This black box tracker records magnetic and GPS heading as well as 3 axis GPS location, barametric altitude, and 3 axis tilt and acceleration. This allows for a wealth of information for both video overlays and for reviewing manoeuvres once you are on the ground.

Here is a video of the testing of an early prototype:

https://www.youtube.com/watch?v=POogGO4eM0E

For this project you will require:

- A study container

- Paperbacked foam board

- MediaTek LinkIt

- GPS Antenna

- Standard LinkIt Battery

- 9-axis + barometric sensor "boodle"

A huge thank you to instructables, autodesk, MediaTek, and Penolopy Bulnick for sending me the Link It One used in this instructable.

Check Out Our Other Accounts! Like, Subscribe, and Follow to keep up to date with our latest projects. :-)

Just4Fun Media Facebook Just4Fun Media Instagram Just4Fun Media Youtube Just4Fun Media Twitter Just4Fun Media Tumblr Just4Fun Media Website

Cutting the Base Plate

photo-2015-12-17, 6:36 PM.jpg
photo-2015-12-17, 6:36 PM.jpg
photo-2015-12-17, 6:36 PM.jpg
photo-2015-12-17, 6:36 PM.jpg

The first step is build a stable base to mount all of the components to. I used paper backed foam board left over from an architectural model but cardboard, balsa, or plastic board will also work.

I pressed the lid of my case into the foam to create an impression then cut it out with a utility knife. I made the piece of foam fit tight enough to friction fit into the lid.

Wiring

photo-2015-12-17, 6:36 PM.jpg
photo-2015-12-17, 6:36 PM.jpg
photo-2015-12-17, 6:36 PM.jpg
photo-2015-12-17, 6:36 PM.jpg
photo-2015-12-17, 6:36 PM.jpg

Now connect the GPS antenna, Battery, 9-axis Sensor, trigger switch, and insert a blank FAT formatted microSD card into the LinkIt.

Layout

photo-2015-12-17, 6:36 PM.jpg
photo-2015-12-17, 6:36 PM.jpg
photo-2015-12-17, 6:36 PM.jpg
photo-2015-12-17, 6:36 PM.jpg

Lay out all of the components neatly onto the foam board and trace the components as well as marking any screw holes. Then make pilot holes in the foam using a pin.

Mounting

photo-2015-12-17, 6:36 PM.jpg
photo-2015-12-17, 6:36 PM.jpg
photo-2015-12-17, 6:36 PM.jpg
photo-2015-12-17, 6:36 PM.jpg

Select screws that will pass through the component without penetrating all of the way through the foam. I have found that the best place to find small screws for mounting components is electronics, I salvaged most of my screws off of old printers. For components without screw holes I used rubberized mesh, as it will not conduct electricity, and screwed the mesh down the the foam.

This rigid mounting will prevent false sensor readings due to movement of the components.

Coding and Data

567609be4936d476f0000b55.jpeg

To read any of the data from the 9-axis sensor you will need to get the libraries from the git hub repository here:

https://github.com/Seeed-Studio/Grove_IMU_9DOF

Some general notes for setting up the sensors are:

The default I2C address is 0x68

Each of the value types (gyroscope, accelerometer, and magnetometer) are transmitted as their own 16bit integers

To find the vector of each value use the 3 dimensional pythagorean theorem to find the modules (m = (x^2 + y^2 + z^2)^1/2)

Things to keep in mind:

Barometric altitude will need to be calibrated in post (leave it raw), use your knee board values

GPS altitude is inaccurate an average of GPS and Barometric is advisable if calibration is unknown

Magnetometer data can be skewed by electronics (the alternator, magnetos, most everything electronic in the plane)

This sensor array is for informational purposes only and as you may note in the video at the start of the video is not always highly accurate

I will post a full source code file for this project on my website Just4FunMedia.TK under the "Apps" tab once I finish working out the bugs. Currently these bugs are inaccurate sensor data, randomly crashing, and corrupted text files. I believe this is due to inefficient coding and the large amount of data gathered during a flight.

I am currently working to find a balance in sensor update rate between quality and reliability, as I believe that my log rate is causing my errors.

The 9-axis sensor also logs x.y. and z axis magnetism which can yield some visually interesting results when graphed in 3 dimensions (see data above). NOTE: the graph is a "ball" because the sensor senses a certain magnetic declination that is broken down into a cartesian vector (X,Y,Z). Imagine the vector as the needle of a compass stemming from the origin to a point

Finishing Touches and Use

photo-2015-12-17, 6:36 PM.jpg
5677a4d2937ddb6ebe0001af.jpeg
5677a51067400ccf02000970.jpeg
photo-2015-12-17, 6:37 PM.jpg

All that is left is to secure the plastic cover and add some much needed stickers. To record cockpit audio I prefer to use a microphone placed in my headset. This way the recorder is fully electrically isolated from the plane, and the recorder can in no way interfere with radio system. Purpose made cables are also available to adapt the standard .25" and .206" headset plugs to 2.5mm audio jacks, although I prefer not to use them.

Just push the switch, Once the LED starts flashing, stow the tracker and forget it.

Feel free to post any comments, questions, or results in the comments.

Have a great day! :-)