• build and test

    turbinenreiter • 02/11/2016 at 20:49 • 0 comments

    A couple of days ago the boards arrived. I had them made at dirtypcbs and I'm pretty happy with them. It's the first time I actually spend time on cleaning up the silkscreen. Unfortunately, some of the silkscreen wasn't printed - probably because it was to small or in the wrong place. The SPI header should have pin names next to the through-holes.

    This is about the fifth board or so I ever SMD soldered. I almost did everything right - except that I used tape to fix the board on the desk - and with that tape I covered the landing pads of the 10k resistor for the switch. So after I applied solder paste with the stencil, I had a perfect spread, except for that resistor. You can see that I hand-soldered it on later with the soldering iron. The rest of the parts I could place and then solder with a heatgun.

    After building comes testing and it seems to work pretty good. The sensors work, LEDs work.

    However, the green LED does something shady. When the motherboard is connected via USB, it goes on, but as soon as the battery is connected, it goes out again. I have no idea why.

    Anyway, next step is to solder one servo header on and put this in my water rocket. The parachute deployment system is already done and tested with remote control, now it's time to automate. I already have flight data from earlier prototypes. The parachute should be triggered on the highest point of the flight. The barometric data isn't really good enough to do that, but the acceleration is. Basically, as soon as the thrust cut's out, the rocket is in free fall and therefore zero-g - except for the aerodynamic drag. This drag is a small negative acceleration that gets smaller the slower the rocket gets - until it's zero when the rocket is at the apogee with zero velocity. Due to noise I can't set the threshold to zero, so what I will do is to detect if the acceleration is below 0.1g and then trigger the parachute with a slight delay.

  • nRF24L01+

    turbinenreiter • 01/23/2016 at 20:05 • 0 comments

    I added another header that exposes SPI and two more pins, so it's easy to add a wireless module - or anything else SPI. It's basically some holes in the PCB where I had space left. Easiest way to add wireless is to just a Feather board that already has it.

  • Random Ramblings

    turbinenreiter • 01/22/2016 at 18:05 • 0 comments

    I've started working on a board for the Feather form factor that has:

    • 3-axis accelerometer
    • 3-axis gyroscope
    • 3-axis magnetometer
    • pressure sensor
    • boost converter to generate 5V
    • headers for 5 servos.

    At the moment, the design uses an MPU9150 and BMP180 - but I've decided to use the BMX055 and BMP280 instead. Bosch has good data-sheets and Open Source drivers, and the 280 is the 180s successor.



    It's one of my first designs and I plan to redo it a lot.



    You can find Eagle files and pdfs on Github: https://github.com/turbinenreiter/10DOF-FeatherWing



    Any feedback would be welcome. If you could use such a board, let me know what you would change or add. For example, I'm thinking about adding a connector for a GPS module.



    /edit:
    Now updated with version 0.2, using the Bosch sensors. I dropped the 5V step up and included an UART port, so it's easy to add GPS.
    /edit:
    Added 3 more servo headers, a button and two LEDs.