Q:
Do you you have any code examples for calculating the new CRC-16 option in the ADIS16448?
-----------------------------------------------------------------------------------------------------------------------------------------
A:
Yes, please see the attached file.
I have implemented a solution:
const uint16_t crc_table [ 256 ] =
{
0x0000, 0x17CE, 0x0FDF, 0x1811, 0x1FBE, 0x0870, 0x1061, 0x07AF,
0x1F3F, 0x08F1, 0x10E0, 0x072E, 0x0081, 0x174F, 0x0F5E, 0x1890,
0x1E3D, 0x09F3, 0x11E2, 0x062C, 0x0183, 0x164D, 0x0E5C, 0x1992,
0x0102, 0x16CC, 0x0EDD, 0x1913, 0x1EBC, 0x0972, 0x1163, 0x06AD,
0x1C39, 0x0BF7, 0x13E6, 0x0428, 0x0387, 0x1449, 0x0C58, 0x1B96,
0x0306, 0x14C8, 0x0CD9, 0x1B17, 0x1CB8, 0x0B76, 0x1367, 0x04A9,
0x0204, 0x15CA, 0x0DDB, 0x1A15, 0x1DBA, 0x0A74, 0x1265, 0x05AB,
0x1D3B, 0x0AF5, 0x12E4, 0x052A, 0x0285, 0x154B, 0x0D5A, 0x1A94,
0x1831, 0x0FFF, 0x17EE, 0x0020, 0x078F, 0x1041, 0x0850, 0x1F9E,
0x070E, 0x10C0, 0x08D1, 0x1F1F, 0x18B0, 0x0F7E, 0x176F, 0x00A1,
0x060C, 0x11C2, 0x09D3, 0x1E1D, 0x19B2, 0x0E7C, 0x166D, 0x01A3,
0x1933, 0x0EFD, 0x16EC, 0x0122, 0x068D, 0x1143, 0x0952, 0x1E9C,
0x0408, 0x13C6, 0x0BD7, 0x1C19, 0x1BB6, 0x0C78, 0x1469, 0x03A7,
0x1B37, 0x0CF9, 0x14E8, 0x0326, 0x0489, 0x1347, 0x0B56, 0x1C98,
0x1A35, 0x0DFB, 0x15EA, 0x0224, 0x058B, 0x1245, 0x0A54, 0x1D9A,
0x050A, 0x12C4, 0x0AD5, 0x1D1B, 0x1AB4, 0x0D7A, 0x156B, 0x02A5,
0x1021, 0x07EF, 0x1FFE, 0x0830, 0x0F9F, 0x1851, 0x0040, 0x178E,
0x0F1E, 0x18D0, 0x00C1, 0x170F, 0x10A0, 0x076E, 0x1F7F, 0x08B1,
0x0E1C, 0x19D2, 0x01C3, 0x160D, 0x11A2, 0x066C, 0x1E7D, 0x09B3,
0x1123, 0x06ED, 0x1EFC, 0x0932, 0x0E9D, 0x1953, 0x0142, 0x168C,
0x0C18, 0x1BD6, 0x03C7, 0x1409, 0x13A6, 0x0468, 0x1C79, 0x0BB7,
0x1327, 0x04E9, 0x1CF8, 0x0B36, 0x0C99, 0x1B57, 0x0346, 0x1488,
0x1225, 0x05EB, 0x1DFA, 0x0A34, 0x0D9B, 0x1A55, 0x0244, 0x158A,
0x0D1A, 0x1AD4, 0x02C5, 0x150B, 0x12A4, 0x056A, 0x1D7B, 0x0AB5,
0x0810, 0x1FDE, 0x07CF, 0x1001, 0x17AE, 0x0060, 0x1871, 0x0FBF,
0x172F, 0x00E1, 0x18F0, 0x0F3E, 0x0891, 0x1F5F, 0x074E, 0x1080,
0x162D, 0x01E3, 0x19F2, 0x0E3C, 0x0993, 0x1E5D, 0x064C, 0x1182,
0x0912, 0x1EDC, 0x06CD, 0x1103, 0x16AC, 0x0162, 0x1973, 0x0EBD,
0x1429, 0x03E7, 0x1BF6, 0x0C38, 0x0B97, 0x1C59, 0x0448, 0x1386,
0x0B16, 0x1CD8, 0x04C9, 0x1307, 0x14A8, 0x0366, 0x1B77, 0x0CB9,
0x0A14, 0x1DDA, 0x05CB, 0x1205, 0x15AA, 0x0264, 0x1A75, 0x0DBB,
0x152B, 0x02E5, 0x1AF4, 0x0D3A, 0x0A95, 0x1D5B, 0x054A, 0x1284
};
uint16_t get_crc16(const uint16_t *data, uint32_t data_length)
uint16_t crc = 0xFFFF;
uint16_t byte;
while ( data_length-- )
// Compute lower byte CRC first
byte = *data & 0x00FF;
crc = ( crc >> 8 ) ^ crc_table[ ( crc & 0x00FF ) ^ byte ];
// Compute upper byte of CRC
byte = *(data++) >> 8;
}
crc = ~crc; // Compute complement of CRC
return (uint16_t) ((crc << 8) | (crc >> 8)); // Perform byte swap prior to returning CRC;
Call it like this:
crc = get_crc16(&burst_read_data[2], 11)
What I mean is that the supplied loop driven CRC implementation is a bit slow (when run in a time restricted interrupt service routine).
That is why I wonder if you could also supply a table driven CRC implementation that would be faster.
Thanks for the comment. To be honest, I am not sure what you are asking for. This post offers the same process that the ADIS16448 uses to compute the CRC for each burst sequence.
Could you post a table driven runCRC() check?
For more explanation on this CRC-16, please see this link, until the ADIS16448 datasheet revisions are complete:
FAQ: ADIS16448 CRC In The Burst Mode!