At 08:49 PM 11/26/02 -0800, you wrote:
>Hi every one
>
>Do any one have any experience in use sine and cos maths function on PIC.
>like the one stated in the microchip AN660 document(using minimax
>approximations). Can any one kindly enlighten me what method they use cos
>I am current running short of memory and I need to replace their fixed
>point multiplication routine with my standard multiplication routine. I am
>not able to understand the flow of the program is there any material that
>I can read on?

Microchip is just using a polynomial to approximate a segment of the
function, and very simple trig transformations to get the operand into the
domain where the polynomial is relatively accurate. Minmax just refers to the
criteria for finding the "optimal" polynomial coefficients. If you are
going to be using their coefficients, then you don't need to know how
to do that.

I didn't delve further into the AN (it looks pretty lucid, actually), but
polynomials are typically evaluated using Horner's method, which is
fast and yields minimum error.

P(x) = a*x^n + b*x^(n-1)... p*x + q

      = (..((x * a) + b) * x) +... + q

So you need only n multiplications and n additions to get the answer.
When using fixed point multiplication, you have to be especially careful
of overflow and of letting the numbers get too small so accuracy suffers.
You might have to scale at each operation to keep the numbers in the
right range, and adjust the coefficients appropriately.

If you don't need much performance speed-wise, you might consider
re-writing the routine to use your pre-existing (?) FP multiplication
routine, which should save space.

Best regards,

Spehro Pefhany --"it's the network..."            "The Journey is the reward"
speff@interlog.com             Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog  Info for designers:  http://www.speff.com

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email listserv@mitvma.mit.edu with SET PICList DIGEST in the body