End of the line?


  • Zoe works as best as she can, but she doesn’t have the performance to run the video above 480 x 480 pixels at 20 fps.  This means lateral motion over the IKEA play mat works, but not on the gravel on the drive, never mind the grass.  I have found a micro-USB GPS dongle which I may add but I’ve yet to find a micro-USB hub to allow power, WiFi and GPS to go into the single micro USB port (the other is blocked by the frame)
  • Hermione can process video macro-blocks at 640 x 640 pixels @ 20 fps (and perhaps higher) which may be enough for the gravel drive, but I’m completely out of ideas for the source of the I²C errors from the Garmin LiDAR-Lite V3 when it is connected even if it’s not run.

I’ll continue working on the I²C to the end of the month, but after that I’ll be downing drone development and move on to other things, probably still Pi related nonetheless.

4 thoughts on “End of the line?

    • I really don’t want to give up – if I can get this to work, that then opens up use of GPS for all sorts of options – not just fusing with the existing 3D distance / velocity sensors, but also setting up significant long distance flight plans to autonomously track between several sets of preset GPS locations.

      I’m currently looking to see if it’s possible to add another I2C bus or to hijack I2C-0. I didn’t know pigpio could sample an incoming PWM – that opens up another option too, thanks. A search for LiDAR lite problems does mention I2C problems when not the only sensor on the I2C bus, and suggests the solution as using PWM instead.

      How’s your project? I have the RSS feed from your site, but it all seems very quiet?

      • It definitely can sample PWM. I used it to sample CPPM from a Frsky receiver. See the code here: https://github.com/jeanleflambeur/silkopter/blob/master/silkopter/brain/src/source/CPPM_Receiver.cpp

        You have to install a callback on a pin and it will get called for any transition on that pin with an accurate timestamp. It’s accurate up to 5us, more than enough I guess.

        I got stuck with my project too several weeks ago. There’s just too much to do, now that I’ doing the RC system as well. Lately I had more progress and have to motivate myself to blog a bit more. I now have the Quad, the RC, the GS and an android H264 viewer, all working (theoretically). I’m playing with them in the simulator to see how everything behaves and fixing bugs left and right. Sometimes 2 monitors is not enough.

        How’s the Mavic?

        • Thanks to the PWM link, it sounds a strong possibility.

          I love the Mavic – it’s completely idiot proof! It has so many sensors, and so much protective code, my input from the RC almost seems like an afterthought in its processing – there are so many different flight settings that are autonomous flights tracking a selected visual object (i.e. me), or flying between a series of preset GPS locations etc. I’ve only played with it in the garden at the moment to get a good feel for the joysticks etc, but I live in the countryside, and the weather’s going to be sunny today, so I’m going to take it out into a local sheep field and play with videoing and photographing them.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.