Hi all,
About two years ago I ordered the plans for and began working on a Hugh Piggot 8' wind generator. The experience has been great and as I move toward actually flying the generator, I have been giving some thoughts to monitoring and load control.
Independently, I have begun some hobby work using the open source Arduino micro controller prototyping platform (http://www.arduino.cc/).
I am thinking of combining those two interests and creating a wind generator monitor/controller. Basically, I would use the analog inputs (8 total) and digital I/Os on the Arduino (14 total) to monitor the wind generator and weather conditions, and control load. My initial plans are to monitor:
- Generator RPM
- Anemometer RPM (wind speed)
- Tail furling angle
- Generator voltage output
It might also be possible (but more expensive sensor-wise) to to monitor:
- Wind direction
- Generator current output
- Diode temperature
- Generator vibration
- Generator noise
The device would be standalone, designed for off-grid use, and include a small LCD for real-time data output. Long-term monitoring would log data to a removable SD card or something similar. The Arduino includes 512 bytes of eeprom for storing preferences. The device would logically have three modes:
- A basic logging weather station for initial site analysis, etc.
- A wind generator monitor for determining optimal generator performance characteristics, such as appropriate loads for given wind speeds furling tail optimization.
- An intelligent wind generator load controller. Given the number of I/O pins, it would probably be possible to control 6 or more loads (I think way more than necessary for the generator I have built).
Total project cost (without the more expensive sensors) should be roughly $150.
Before I go to far with this, I am wondering if people have thoughts or advice with respect to this project. I have not used an actual wind generator controller before and so do not really have a sense for what is usual or useful.
I general, I have very much appreciated Hugh's "bullet proof" approach which trades off optimal output for reliability, and I realize a (relatively more complex) device like this somewhat goes against that ethos. However, to me this seems like an interesting and fun project, if it would end up being useful. I would open source the design and code.
Thanks much for the input,
Scott