uch a device has merit. A squall curve diagram generally indicates the maximum angle of heel recommended for a given wind velocity, and the potential for gusts. At a basic level, assuming a prevailing wind of 25 knots (beaufort force 6), this device will measure heel (inclination) and indicate the level of gust that would present a knockdown risk based on that angle of heel, by flashing a LED for either 60 knots, 45 knots or 30 knots.
I based the design around an accelerometer, with the plan being to trial more than one. The sort we have in stock is a single axis, but Lindy had an ADXL335, 3-axis (one of the new ones, and it was around this that I designed the circuit. With three axis inputs, and three LED outputs, I needed to use the ATMega88 new for me. I designed the circuit so that the accelerometer resided on a separate board with three passive lo-pass filters that would connect to the main board via a 2x3 pin arrangement.
I set the board up this way so that I could experiment with several accelerometers, hoping to ultimately build my own using a flexure. The ADXL335 has resistors built in resistors so technically I donG•t need the ones in the above diagram, and they are zero ohm.
I included them on the advice of a former technical director at RIM who said she always made provision for them, so she could tweak the filter later if necessary. Alas, I could not get the AVRISP to recognize the board. I troubleshot every connection on the board, and they all seem to work. The easiest component to suspect is the ATMega maybe I cooked it while soldering, or maybe I shorted it out at some stage.
For coming weeks, (once I get this working) IG•d like to add an input from an anemometer (wind indicator) and have the device interpret both inputs against the squall curves. I may combine this with Output devices.