AVR Butterfly GPS logger

Introduction

The original goal of the project was to build lightweight and cost effective GPS solution for RC plain. The AVR Butterfly board was chosen as cheap but quite functional hardware platform that already have 512 Kb on board flash memory to store log. The board weights only 20 grams without battery. The GPS receiver GlobalSat BR-355 shown on Fig.1 and connectors add another 35 grams so the device may be easily mounted even on the parkflier. The code was derived from excellent gcc port of the original ATMEL AVR Butterfly Application by Martin Thomas. The data logging part of the code were inspired by AVR Butterly Logger though the code was written from scratch to ensure high degree of portability and extensibility. Now adding new data entries to the log is as simple as adding new entry to the log descriptor tables in the C source and in the python script for log download (btget.py).

Getting sources

You can

  • access SVN repository directly:
    svn co https://butterflygps.svn.sourceforge.net/svnroot/butterflygps butterflygps
  • download latest tarball
  • download latest code release

How to build

You need avr-gcc compiler in order to build the project. The WinAVR is perfectly suitable on windows. To build just type make in the project directory. If you don't like building just use butterflygps.hex file located in the project directory.

How to install

To install the program onto the AVR Butterfly board you can either use the built in boot loader or ISP programmer. The first approach may not work out of the box depending on the CPU fuses configuration. I used ARVISP mkII as a more flexible solution.

Connecting to GPS receiver and powering

Any NMEA compatible GPS receiver may be connected directly to the AVR Butterfly serial port. I used GlobalSat BR-355 receiver. Note that GPS receiver typically requires 5V power. While the Butterfly CPU may be powered from 5V the on board data flash does not. Powering from 5V source does not damage the flash but data reading/writing without errors becomes impossible. So it's necessary to convert voltage down to 3.3V. I used single chip regulator LM2937-3.3 from National Semiconductor. To power device during log downloading you may either use battery or power it from the external source. I used power cable connected to the UCON232 USB to serial bridge (see Fig.2).

Running

Upon entering GPS sub menu the device will start receiving GPS data stream. You can monitor various data stream items, start/stop collecting min/max values of the speed, altitude and the number of satellites available as well as start/stop GPS data log. The log is capable of storing 1 data frame per second during about 11 hours. The data buffer is being written to the log every 22 seconds. Unexpected powering the device off doesn't destroy log data except the last unsaved data buffer. The log may contain more than one track so you can start/stop log multiple times and all corresponding tracks will be stored separately. In case the log running time exceeds the storage capacity the oldest data will be overwritten by the newest one so the log will keep the data collected during the last 11 hours of running. You may reset the log any time to logically clear current log content or erase the log physically. Unless the LOG_DEBUG macro is undefined you also have the possibility to run log test sequence which is useful for checking the storage for data errors. The log format has checksums embedded in the data stream so errors will likely to be detected.

Log manipulation tools

There are 3 tools written on python:

  • btget.py
    Download the log and split it onto the individual tracks. Start the script first than choose GPS/LOG/TRANSFER on the Butterfly board. The script will do everything and quit. The log is being transferred at 9600 baud via the same serial port as used to connect to GPS receiver.
  • btplot.py
    Track viewer (see Fig.3). It also prints track statistics to the console. You can place cursor on the track by mouse clicking, move it by arrow keys and monitor track data at the cursor position on the console.
  • btkml.py
    Export one or more tracks to KML format that can be viewed in Google Earth as long as other available geo-information such as maps, satellites photo ets. (see Fig.4).
All tools except btget.py may work either in 2D or 3D mode. Pass -2 or -3 command line option to the script to choose mode. To run tools you need python and the following modules:

Links

  1. Project main page
  2. Browse project source code
  3. ATMEL AVR Butterfly product page
  4. Martin Thomas AVR-Projects page
  5. The AVR Butterly Logger page
  6. Useful info and links on AVR Butterfly Primer page
  7. NMEA protocol information page
  8. KML Tutorial
  9. Beginning Python Visualization book by Shai Vaingast

Copyright (c) 2009 OlegV

Butterfly GPS on Maxi-Swift with BR-355 receiver

Fig.1 Butterfly GPS on Maxi-Swift with BR-355 receiver

Firmware development setup

Fig.2 Firmware development setup


RC plane flight record in btplot viewer

Fig.3 RC plane flight record in btplot viewer


RC plane flight record exported to Google Earth

Fig.4 RC plane flight record exported to Google Earth