HP-35 style trig functions

[Druzyek]

Thomas,
Thanks for the examples. I had not seen the digit by digit method. The last example looks like exactly what I need. I think I can eliminate a lot of cycles in the inner loop if sign changes on the first iteration or two, since I can skip iterations that would cancel out anyway.

Quote:If you are not converting to binary then what algorithms are you using for multiplication and division? Are you using CORDIC algorithms for these as well?

Multiplication is the one exception where I convert the bytes of one argument to binary and store them so they can be halved with the Russian peasant algorithm. I leave the bytes of the other argument in BCD, so doubling is just adding the byte to itself and halving is just a shift since the other byte is in binary. This is a lot faster than the quarter squares method. Division is just subtract and shift. I get a big speed up shifting by nibbles rather than bytes.