Raghunathan R wrote:
-------------------------------
Incidentally how did you manage to fit the pattern into a single function ?


i donno what ou mean by single function but what paul has done is Mr. karas and i said interpolation. linear interpolation is you assume all the points between two descrete points in a line. higher degree polynomial represets other more complex curves. every curve can be fitted by a polynomial aproxly, unless it has large degree of randomness between its points.

if flash space is no problem and you can increase the resolution of your table. you can built a two or three level LUT , by distribution of data, giving you enogh precision.

mov a,value
mov dptr,table1
movc a,@a+dptr
mov dptr,table2
movc a,@a+dptr


abhishek