I found a neat approximation for an arctangent in radians of a value of X 
in [-1,1]:

atan(X) = X / { 1 + 0.28 X*X }     error < 0.005

If I use 8-bit fixed point, then the denominator will require 9 bits, which 
creates an annoying problem with the division.

I see several solutions to the problem.

1. I can shift the denominator one bit to the right, creating a temporary 8 
bit result, doing the division, and then shifting the quoteient one bit to 
the right to compensate. Sort of a limited floating point.

2. Do a 32 bit by 16 bit division, and then convert to an 8-bit answer.

Does anyone have experience with the best way to deal with this type of 
precision issue?

Thanks.
================================================================
Robert A. LaBudde, PhD, PAS, Dpl. ACAFS  e-mail: ral@lcfltd.com
Least Cost Formulations, Ltd.            URL: http://lcfltd.com/
824 Timberlake Drive                     Tel: 757-467-0954
Virginia Beach, VA 23464-3239            Fax: 757-467-2947

"Vere scire est per causas scire"
================================================================

-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist