First Simple test of the lSM303DLHC
Procedures:
The digital compass was kept in my basement which was 67 Fahrenheit. The compass was attached to a tripod by means of a nut glued to the base of the compass for this purpose. The compass was pointed across a room at the corner of the house where two walls meet approximately 20 feet away. Then the compass was turned on. The operator waited at least two seconds between measurements. The operator pressed the pause button on the heading screen. The heading was recorded, then the pause button pressed again to begin the next measurement. After a delay greater than 1 second, the procedure was repeated for 48 data points.
Observations:
The digital compass was kept in my basement which was 67 Fahrenheit. The compass was attached to a tripod by means of a nut glued to the base of the compass for this purpose. The compass was pointed across a room at the corner of the house where two walls meet approximately 20 feet away. Then the compass was turned on. The operator waited at least two seconds between measurements. The operator pressed the pause button on the heading screen. The heading was recorded, then the pause button pressed again to begin the next measurement. After a delay greater than 1 second, the procedure was repeated for 48 data points.
Observations:
- The laser on the far wall moved significantly when the button was pressed. This is due to the mount not being extremely sturdy and the requirement to press the button hard. This would be indicative of normal use, but could be a factor in the test results.
- The first 3 measurements varied greatly from the rest of the data set. It was not determined whether this was due to movement of the compass or "warm-up" time for the compass. This will need to be evaluated more closely.
Results without accounting for the 3 outlying points at the beginning of the run
Without accounting for the initial 3 points being major outliers, we get a result that shows a variation of 4.7 degrees. This is clearly not sufficient for a cave compass and other changes would have to be made to the circuit or the test in order to improve upon this. However, we should first determine ideal results without the startup points.
Results after removing the first 4 points from the calculations
These results are much more promising. The variation was 2.3 degrees which is closer to what we would see in the field. Through testing this will need to be improved upon in future firmware releases.
Next Steps:
- Build a more stable platform to determine how much of this variation is due to movement on the tripod.
- Test the device as it changes temperature (put it in the fridge for 20 minutes and then test it over the next hour to identify temperature drift)