Has this subject been discussed enough yet? I like this kind of thing and would thus like to share some more information.

I made up a MathCad worksheet to re-organize the table from Raghu's sensor calibration data so that you end up with a look up table that can be used at run time to interpolate degrees using linear interpolation. The MM value from the sensor is used as an index to LUT and then this finds the N and N+1 values of degrees that need to be interpolated between. The MM interpolation range can be remainder from the divide that gives the index to the table.

Here is the look up table I produced with the MathCad worksheet:



Here are links with complete information on the MathCad worksheet including an error analysis of the original calibration points to see how good the cubic spline used to produce the above run time table will be.

The Mathcad Worksheet

A PDF Printout of the Worksheet

Michael Karas