• State
  • Locked Locked
  • Replies 3 replies
  • Subscribers 79 subscribers
  • Views 2642 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

Multiplying 16-bit number and 32-bit number using 16 x 16 hardware multiplier in MSP430F4784

SUNIL SAWANT
Intellectual 955 points
Other Parts Discussed in Thread: MSP430F4784

Hello,

 

I am developing application on MSP430F4784. It has a 16x16 hardware multiplier.

Development IDE is Code Composer Studiio v 5.4.

I want to perform following multiplication:

unsigned int temp1;

unsogned long temp2;

temp2 = temp1 * 100000;

 

16 x 16 hadrware multiplier cannot multiply numbers beyond 16-bit.

How can i perform above mentioned multiplication using 16 x 16 multiplier in MSP430F4784 ?

Is there any software library available to perform arithmetic calculations in Code Composer Studio v 5.4 ?

 

 

Thank you,

Sunil

  • 0
    Guru 25210 points

    Break the 32-bit number into 16-bit parts. Do the 2 multiplications with the 16-bit number, and then add the results (with 16-bit left shift of the upper product).

    Wrap this into it's own function, something like:

    unsigned long mulU16xU32(unsigned op16, unsigned long op32)
{
    unsigned long result = 0;

    // Your code here....

    return(result);
}

    If you make this a generic function, then you need to make sure to handle that the output could be 48-bits (16+32). So what to return if the calculation overflows?

  • 0 in reply to Brian Boorman
    Intellectual 955 points

    I have used the method you mentioned to multiply 2, 32-bit numbers.

    The method is working well for unsigned numbers.

    What about signed numbers ?

    I mean how to multiply 2, 32-bit signed numbers ?

  • 0 in reply to SUNIL SAWANT
    Guru 58965 points

    Keep it stupid and simple. (a) Figure out the sign of the result, (b) take the absolute values and multiply them (unsigned * unsigned), and (c) apply the sign of the result.

**Attention** This is a public forum