Welcome

My virtual playground. Meer info

Servo controller

After a silent period (which lasted too long), I finally have some news.

I had to tackle some issues concerning the programming of my PICs, but now they work just fine.

This week, I’ve been working on a servo-controller which decodes the PPM pulse train from the receiver and sends pulses to the servos.
Again, I encountered the weirdest problems: My servo moved reeeaaal slow when I made a quick movement with the stick of my transmitter. After days of debugging my code, I found out that the quick movement made the servo-motor use a lot of current, triggering a voltage drop, and the Brown Out Reset of the PIC made the whole thing reset. But now it works fine :-)

The code is written in CCS/C. The most basic code (= 2 interrupt routines) is about 20 lines long, so I still have a lot of processing time left! I also added some basic filtering, which only sends valid frames (with the correct number of channels and a correct synchronization pulse) to the servo’s. This greatly improves the reliability of my old Mulitplex PICO5/6 receiver.

The next step is reading in the analog values of the thermophile sensors and adding a control mechanism to drive the servos. Exciting stuff!

Here some photos of my home-lab ;-)

22 October 2006, 10:22 | Link |

Melexis IR - thermophile - IR sensors arrived

Unfortunately I’ve been busy with other things, so I haven’t added a new Kalman tutorial. One of the reasons is that I ordered 4 melexis sensors (IR) for an easier stabilisation for my first autopilot attempt. Almost 2 months after ordering, they finally arrived! Get them yourselves from Scale Robotics if you have enough patience to wait 2 months.

The PIC’s shown are some I bought really cheap while i was in Turkey :-)

My first goal now is to create an autostabilizing platform using a PIC and the 4 thermophile sensors.

23 August 2006, 10:53 | Link |

B-Hunter crash

A few days ago, a B-Hunter UAV crashed in Congo. Six people were slightly injured.

The B-hunter is a moderately large UAV developed by Israel Aircraft Industries for the Belgian army.

Now comes the funny part: the UAV crashed due to a software failure. Being developed in the northern hemisphere, the software wasn’t able to interpret the GPS data correctly on the southern hemisphere.
Hilarious.

31 July 2006, 07:38 | Link |

Looking for GPS modules

I’ve been surfing the internet looking for ideas for my autopilot system. The most important sensor we need, is a GPS. Several important factors that influence my choice are:

  • Weight!
  • Accuracy
  • Update rate
  • Cost
  • Easy to use

Right now I found two candidates:
a Sirf Star III based module (1Hz update, 10m accuracy, 17 gram)

and some from µblox like the SAM-LS (4Hz update, 2.5mr accuracy, 23 gram) used by the paparazzi project. I prefered the SAM-LS because if it’s high update rate and great accuracy:

Unfortunately, it uses some weird flex cable to connect, so purchasing it could be a risk when I wouldn’t find the right connector. In the the RCB-4H (basically a development board; 17 grams and requires an active antenna) might be a better choice…

To be continued :-)

1 June 2006, 15:24 | Link | Comments [5]

| Newer articles