PIC Microcontoller Basic Math Division Methods

• 8 bit by 8bit to 24bit floating point AN575 format from Nikolai Golovchenko
• 16 bits by 8 from Lou Zher
• 16 bits by 8 with result as fraction rather than remainder by Nikolai Golovchenko
• div10byfxp7q8.htm Divide 10 bit integer using a fixed point divisor and result (7Q8 format)
• 24 bits by 8 from James Ashley Hillman
• 24 bits by 16 from Nikolai Golovchenko
• 24 bits by 16 from Andy David
• 24 bits by 24 from Tony Nixon
• 32 bits by 16 by Peter Hemsley +
• 32 bits by 16 via Tom Napier from FIG-Forth '79
• 32 bits by 16 using 16 bits by 16 by Nikolai Golovchenko
• 48 bits by 24 from Nikolai Golovchenko and Tony Kübek
• 48 bits by 24 from Andy Lee
• Binary Division by a Constant
• Tutorial
• Reciprocal
  • 32768/x by Nikolai Golovchenko
• Multiplication and/or division by a Constant
  • See: The PICList code generator by Nikolai Golovchenko
  • 48 bits by the constant 10 by VegiPete for PIC18F
  • 24 bits by the constant 6 by Hughe "can be used for any 8 bit constant divisor with 2 values changed."
  • 16 bits by the constant 15 by Dmitry Kiryashov
  • 16 bits by the constant 10 by Alex_t
  • PICList post "16bit divide by 10" +
  • List post "Divide by 10"
  • PICList post "coding challenge: divide by 7" +
  • 8 bits by the constant 10 on a PIC18
  • 8 bits by the constant 5 +by Ihsan Volkan Tore "can be used for constant divisors 5,10,20,40 and 80."
  • 8 bits by the constant 3 by Andy Warren
• (8 bits * 256) / 8 bits by Nikolai Golovchenko
• Archive
  • PICList post "Challenge" detecting divisibility by 3
  • PICList post "16fxdu/8 divide routine"
• see also
  • http://home.netcom.com/~fastfwd/answers.html#PIC00094 from Andrew Warren: 16-bit by 8
  • http://home.netcom.com/~fastfwd/answers.html#PIC00064 from Andrew Warren: 8-bit by 4 PIC16C5x
  • http://www.myke.com/PICMicro/16bit.htm 16 bit by 8
  • http://www.myke.com/16bit.htm#div 16fxdu/8 divide routine
  • http://www.dontronics.com/see.html#division 16 bit by 8

Archive:

• List post "16 bit Mod 7 Math" +
• PICList post "Divide 24/8 routine" ASM embedded in C +
• PICList post "16fxdu/8 divide routine" +
• PICList post "16/8 Bit Division with remainder"
• PICList post "Help needed with simple question"
• PICList post "24bit by 16bit Division" (actually 24 / 8)

Code:

• billyman3TakeThisOuT at gmail.com

  Any reciprocal 1/n where n - 1 = 2^m (and m is an integer)can be represented as a geometric series starting with 1/(n-1) and having a ratio of -1/(n-1). This is convenient because division by n is just a matter of shifting. To expand on that point, dividing by another number, q, when q's factors are made up of the sequence described earlier (3, 5, 9, 17, 33...) and the sequence 2^m (2, 4, 8, 16, 32...) is a breeze.

  +

Questions:

• neil_john_breakwellTakeThisOuT at hotmail.com asks: " Very helpful site. May I submit code routines of my own for possible inclusion ?" James Newton of MassMind replies: Of course. All submissions are moderated, although they don't appear to be.+

  +

• tomnapier3TakeThisOuT at gmail.com asks:

  I recently adapted the 32 bit division routine from a 1979 edition of 8080 FIG-Forth to run on a 16F876. As this uses only 16 loops it is inherently faster than Peter Hemsley's 32-loop version above. Would this be worth adding to your library and if so, how?
  Tom
  

  James Newton of MassMind replies: Sure. Just post it here, but be sure to select "preformate text" from the options list.+

  +

Comments:

  .