CRC16 - Any luck?

[Austin McElroy]

Hey all,

Has anyone had much luck with the CRC16 checking? I found this implementation from FLIR and ported it to Java, but the results are still off. 

https://github.com/groupgets/purethermal1-firmware/blob/master/Middlewares/lepton_sdk/Src/crc16fast.c

The manual is extremely vague, as there are numerous CRC implementations out there, but it says for a row of data from the Lepton, containing 164 bytes (zero indexing)

• Zero the first 4 bits of byte 0 (the frame check packet, upper bytes)
• Zero the 2nd and 3rd bytes (the CRC packets)
• Run a CRC16 on all 164 bytes, with the zeros from above

When using the CRC check, the frame rate drops to like 1 frame every 5 seconds, implying a huge number of bad packets. 

When running without CRC, 99.99% of the frames look good, but I need to ID the frames right before the SPI losses sync. 

So, either the CRC16 check is wrong or the data is being corrupted. I am using around 25cm cables right now to connect the Lepton to the Pi, and will be shortening these to 10cm to see if that helps. 

Austin

[Austin McElroy]

Bah, found the problem. Java doesn't have unsigned byte and I wasn't shifting one of the CRC packets incorrectly. Here is my Java code in case anyone else needs something that is known to work:

    int ccitt_16Table[] = {

        0x0000, 0x1021, 0x2042, 0x3063, 0x4084, 0x50A5, 0x60C6, 0x70E7,

        0x8108, 0x9129, 0xA14A, 0xB16B, 0xC18C, 0xD1AD, 0xE1CE, 0xF1EF,

        0x1231, 0x0210, 0x3273, 0x2252, 0x52B5, 0x4294, 0x72F7, 0x62D6,

        0x9339, 0x8318, 0xB37B, 0xA35A, 0xD3BD, 0xC39C, 0xF3FF, 0xE3DE,

        0x2462, 0x3443, 0x0420, 0x1401, 0x64E6, 0x74C7, 0x44A4, 0x5485,

        0xA56A, 0xB54B, 0x8528, 0x9509, 0xE5EE, 0xF5CF, 0xC5AC, 0xD58D,

        0x3653, 0x2672, 0x1611, 0x0630, 0x76D7, 0x66F6, 0x5695, 0x46B4,

        0xB75B, 0xA77A, 0x9719, 0x8738, 0xF7DF, 0xE7FE, 0xD79D, 0xC7BC,

        0x48C4, 0x58E5, 0x6886, 0x78A7, 0x0840, 0x1861, 0x2802, 0x3823,

        0xC9CC, 0xD9ED, 0xE98E, 0xF9AF, 0x8948, 0x9969, 0xA90A, 0xB92B,

        0x5AF5, 0x4AD4, 0x7AB7, 0x6A96, 0x1A71, 0x0A50, 0x3A33, 0x2A12,

        0xDBFD, 0xCBDC, 0xFBBF, 0xEB9E, 0x9B79, 0x8B58, 0xBB3B, 0xAB1A,

        0x6CA6, 0x7C87, 0x4CE4, 0x5CC5, 0x2C22, 0x3C03, 0x0C60, 0x1C41,

        0xEDAE, 0xFD8F, 0xCDEC, 0xDDCD, 0xAD2A, 0xBD0B, 0x8D68, 0x9D49,

        0x7E97, 0x6EB6, 0x5ED5, 0x4EF4, 0x3E13, 0x2E32, 0x1E51, 0x0E70,

        0xFF9F, 0xEFBE, 0xDFDD, 0xCFFC, 0xBF1B, 0xAF3A, 0x9F59, 0x8F78,

        0x9188, 0x81A9, 0xB1CA, 0xA1EB, 0xD10C, 0xC12D, 0xF14E, 0xE16F,

        0x1080, 0x00A1, 0x30C2, 0x20E3, 0x5004, 0x4025, 0x7046, 0x6067,

        0x83B9, 0x9398, 0xA3FB, 0xB3DA, 0xC33D, 0xD31C, 0xE37F, 0xF35E,

        0x02B1, 0x1290, 0x22F3, 0x32D2, 0x4235, 0x5214, 0x6277, 0x7256,

        0xB5EA, 0xA5CB, 0x95A8, 0x8589, 0xF56E, 0xE54F, 0xD52C, 0xC50D,

        0x34E2, 0x24C3, 0x14A0, 0x0481, 0x7466, 0x6447, 0x5424, 0x4405,

        0xA7DB, 0xB7FA, 0x8799, 0x97B8, 0xE75F, 0xF77E, 0xC71D, 0xD73C,

        0x26D3, 0x36F2, 0x0691, 0x16B0, 0x6657, 0x7676, 0x4615, 0x5634,

        0xD94C, 0xC96D, 0xF90E, 0xE92F, 0x99C8, 0x89E9, 0xB98A, 0xA9AB,

        0x5844, 0x4865, 0x7806, 0x6827, 0x18C0, 0x08E1, 0x3882, 0x28A3,

        0xCB7D, 0xDB5C, 0xEB3F, 0xFB1E, 0x8BF9, 0x9BD8, 0xABBB, 0xBB9A,

        0x4A75, 0x5A54, 0x6A37, 0x7A16, 0x0AF1, 0x1AD0, 0x2AB3, 0x3A92,

        0xFD2E, 0xED0F, 0xDD6C, 0xCD4D, 0xBDAA, 0xAD8B, 0x9DE8, 0x8DC9,

        0x7C26, 0x6C07, 0x5C64, 0x4C45, 0x3CA2, 0x2C83, 0x1CE0, 0x0CC1,

        0xEF1F, 0xFF3E, 0xCF5D, 0xDF7C, 0xAF9B, 0xBFBA, 0x8FD9, 0x9FF8,

        0x6E17, 0x7E36, 0x4E55, 0x5E74, 0x2E93, 0x3EB2, 0x0ED1, 0x1EF0

     };

  private int crc16rowBigEndian(final byte[] bytes) {

        int crcin = 0x0000;

        

        for (int b = 0; b < bytes.length; b += 2) {

            crcin = crc16(bytes[b], crcin);

            crcin = crc16(bytes[b + 1], crcin);

        }        

        

        return crcin;

    }

    private int crc16(byte value, int crcin){     

        int bb = value & 0xff;

        crcin = (crcin << 8) ^  ccitt_16Table[((crcin >> 8) ^ bb) & 255];  

        crcin &= 0xffff;

        return crcin;

    }

[Christian Gál Agnér]

Hi,

I've translated this to python, which is probably why I have issues, but I can't get it to work.

Can you post how you used this in your code? 

If possible please attach a packet and it's crc that I can verify against.

regards,

Chris