Interesting, John. I'll explore that a bit. I have some library code that
uses that binary-to-bcd algorithm; I should be able to play with it in
reverse. Thanks for the idea.

Larry

At 04:35 PM 12/21/2003 -0500, you wrote:
>After reading your first post, I started thinking about this. There does seem
>to be a method, but I need more time to refine it.
>There is a clever binary-to-BCD algorithm in the Microchip application note
>which describes utility math operations (AN526. appendix K). The method
>involves left shifting the BCD partial result and then "fixing up" the result.
>After repeating this N times for an N-bit binary number, the BCD result is
>the value of the original binary number. On close examination, you can
>see that the shift and fixup operation amounts to multiplying the BCD
>result by 2.
>It seems to me that performing this procedure in reverse will divide the
>BCD number by 2. In fact, it does. This means that if you start with
>50 00 00 00 ...
>in a BCD register, you can successively create
>25 00 00 00 ...
>12 50 00 00 ...
>06 25 00 00 ...
>03 12 50 00 ..
>If you add these BCD numbers to a BCD accumulator, using the binary
>fraction as a guide for when to add, the result will be the decimal fraction
>which is equivalent to the original binary fraction.
>To handle scientific notation, you must deal with large exponents. Doing
>this quickly and with optimized code is the part I am still thinking about.
>
>John Power
>
>--
>http://www.piclist.com#nomail Going offline? Don't AutoReply us!
>email listserv@mitvma.mit.edu with SET PICList DIGEST in the body

Larry Bradley
Orleans (Ottawa), Ontario, CANADA

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