Check Sum Question - for you math people

JohnMichnr
JohnMichnr Junior Member
in AMX Technical Discussion
Hey All - I have a Hitachi LCOS video wall cube that says it's check sum is a "twos complement" check sum. Looking Two's compliment up on Wikipedia I find that it is a inversion of the bit in a byte plus one. So I am thinking that I can calculate the checksum by doing this:

sum (addition of the Bytes in question)

CheckSum = !Sum + 1

Will that work ? or do I have to go into the bits in the byte and invert each one? before I add one.

Anybody got any suggestions?
· Share on FacebookShare on Twitter

Comments

  • DHawthorne
    DHawthorne Junior Member
    JohnMichnr wrote:
    Hey All - I have a Hitachi LCOS video wall cube that says it's check sum is a "twos complement" check sum. Looking Two's compliment up on Wikipedia I find that it is a inversion of the bit in a byte plus one. So I am thinking that I can calculate the checksum by doing this:

    sum (addition of the Bytes in question)

    CheckSum = !Sum + 1

    Will that work ? or do I have to go into the bits in the byte and invert each one? before I add one.

    Anybody got any suggestions?

    !Sum will make Sum = 1 if it's currently 0, otherwise it will make it 0. So, no matter what's it's value, you are going to wind up with 0 or 1. You need to do a bitwise operation.
    · Share on FacebookShare on Twitter
  • JohnMichnr
    JohnMichnr Junior Member
    Yeah now that I go back and re-read the exact defintion of Not, I gues all it does is produce a true/false result.
    So it is back to the bitwise not on each bit.
    · Share on FacebookShare on Twitter
  • mpullin
    mpullin Obvious Troll Account, Marked for Deletion
    CheckSum = ~Sum + 1 is what you're looking for, I believe.
    · Share on FacebookShare on Twitter
  • JohnMichnr
    JohnMichnr Junior Member
    Yeah - that is what I am going to try - the other thing that works in teh Windows calculator is taking sum and subtracting it from 0. Cause the twos compliment is just the inverse of the number (negative. I think I'll try both and see what happens.
    · Share on FacebookShare on Twitter
  • yuri
    yuri Junior Member
    isnt it BNOT? binary not? :)

    edit:
    damnit... BNOT == ~ :P
    · Share on FacebookShare on Twitter
  • JohnMichnr
    JohnMichnr Junior Member
    I was asked to post the description of the command protocol for this issue. I am going to try to do this.

    I don't know how to do the little quote boxes

    FORMAT of COMMANDS TO THE PROJECTOR
    (A) When individually specifying a projector
    STX + BC + CMD + PARA + CS + SN
    STX - Start command (fixed to 02H) 1 Byte
    BC - Byte count from CMD to the last byte of PARA 1 Byte
    CMD - Command code 1 Byte
    PARA* Parameter Variable
    CS - Value when the sum of the lower 8 bits from BC to CS is 0. Two?s complement. 1 Byte
    SN - Projector No.. FFH corresponds to all projectors. 1 Byte

    Obviously the area I am looking at is the CS. And I think the BNOT (~) is the way to go about it. I'll be trying that this morning with some strings.
    · Share on FacebookShare on Twitter
  • Chip Moody
    Chip Moody Junior Member
    Make "count" or somesuch equal to the sum of all the bytes from BC to the last PARA byte. BAND it with $FF to make sure it's 255 or less. Then...

    CS = 256 - count.

    - Chip
    · Share on FacebookShare on Twitter

Categories