CRC 8 Calculator

[vinny 0]

Hi I am fairly new to TWINCAT and am trying to translate a bit of code for a checksum CRC8 calculator. I have the following code, which I think is in C# (which I have no experience in whatsoever) unsigned char CalcCRC8(unsigned char *Data) { unsigned char LoopCntr; unsigned char A; unsigned char B; unsigned char CRC8 = 0; B = 0; while (*(Data + B)){ A = *(Data + B); for (LoopCntr = 0; LoopCntr < 8; LoopCntr++, A >>= 1){ // 8 bit loop if ((A ^ CRC8) & 0x01){ // test bit 0 of (Data XOR CRC8) CRC8 ^= 0x18; // toggle bits 3 and 4 of CRC8 CRC8 >>= 1; // rotate right CRC8, 1 time CRC8 |= 0x80; // set bit 7 of CRC8 } else { CRC8 >>= 1; // rotate right CRC8, 1 time } } B++; } return CRC8; // return CRC8 } I would be most grateful if someone could translate it for me! Thank you very much

[JouniK 2]

Here's one version. In that code of yours, the function reads the data with the "data" pointer one byte at a time and does the CRC check until it hits 0x00 byte, which would be the end of the string. Also i do not know C#, but what i fould out is that there is no unsigned char in C#, rather you're supposed to use byte, so my version uses byte. Also there is no rotate in C# according to what i found, but the >> is a shift, so that comment in the code is not true. Also instead of using the pointer way of doing it, the data is passed to this function as a parameter (in this example, it's a byte array from 0 to 2, giving 3 bytes). You can change that however you need. You maybe should also rename the variables better than A and B, i just did that, so you could better compare the C# code and the ST code. Hopefully you have some way of verifying the result. I just quickly checked that it does what the comments say, but there could be bug(s). Here's your code with my added comments: unsigned char CalcCRC8(unsigned char *Data) { unsigned char LoopCntr; unsigned char A; unsigned char B; unsigned char CRC8 = 0; B = 0; // B is the index for pointing to new byte while (*(Data + B)){ // while pointer points to data that isn't 0x00, which would mark the end of the "string" A = *(Data + B); // copy the data to A for (LoopCntr = 0; LoopCntr < 8; LoopCntr++, A >>= 1){ // 8 bit loop if ((A ^ CRC8) & 0x01){ // test bit 0 of (Data XOR CRC8) CRC8 ^= 0x18; // toggle bits 3 and 4 of CRC8 CRC8 >>= 1; // rotate right CRC8, 1 time CRC8 |= 0x80; // set bit 7 of CRC8 } else { CRC8 >>= 1; // rotate right CRC8, 1 time } } B++; // Byte handled, increment index for next byte } return CRC8; // return CRC8 } Here's the code for twincat: FUNCTION CRC8 : BYTE VAR_INPUT Data : ARRAY[0..2] OF BYTE; END_VAR VAR B : BYTE; (* data index *) LoopCntr: BYTE; (* Bit loop counter *) A: BYTE; (* Temporary data byte *) END_VAR CRC8:= 0; (* reset CRC *) B:= 0; (* Reset index *) WHILE ( B < 3 ) DO (* while not end of data *) A:= data[B]; (* Read data to A *) LoopCntr:= 0; (* Reset loop counter *) WHILE ( LoopCntr < 8 ) DO (* 8 bit loop *) IF (((A XOR CRC8) AND 16#01) > 0) THEN (* test bit 0 OF (Data XOR CRC8) *) CRC8:= CRC8 XOR 16#18; (* toggle bits 3 AND 4 OF CRC8 *) CRC8:= SHR(CRC8, 1); (* shift RIGHT CRC8, 1 TIME *) CRC8:= CRC8 OR 16#80; (* set bit 7 OF CRC8 *) ELSE CRC8:= SHR(CRC8, 1); (* shift right CRC8, 1 time *) END_IF; LoopCntr:= LoopCntr + 1; (* Increase loop counter for next bit *) A:= SHR(A,1); (* Shift data 1 bit right *) END_WHILE; B:= B + 1; (* Increment index for next byte *) END_WHILE; Edited 18 Apr 2013 by JouniK

[vinny 0]

You Sir are a Gentleman and a Scholar! Yeah that code was provided in a fairly useless manual and I couldn't get my head around the source but thank you so much for doing what you have! In ST I can at least make sense of what is happening and try to solve the issue.. And yes I do have a few commands and their predefined crc8 values which i can refer against. If you are ever in England, I owe you a pint of beer! Thanks again!