Bit Shift difference between PIC16 and PIC18 ???

Hi,

This is my first post here, normally the depth of knowledge on this forum is so great - that a search always results in help.

But this problem has me stumped and searching has not pulled anything which will explain what is going on, probably user error.

My code is as follows

Code:

TempWord1 var WORD TempWord2 var WORD CRC16 var WORD CRC16 = $FFFF for j = 0 to 48 ' parse through each byte in serial data TempWord1 = (CRC16 >> 8) ^ SerialData[i] CRC16 = CRC16 << 8 TempWord2 = TempWord1 ^ (TempWord1 >> 4) CRC16 = CRC16 ^ TempWord2 TempWord2 = TempWord2 << 5 CRC16 = CRC16 ^ TempWord2 TempWord2 = TempWord2 << 7 CRC16 = CRC16 ^ TempWord2 next j

So my problem if I run this code on a PIC16F1936 then a PIC18F46K20, the resulting CRC16 value is different for the same data array (SerialData).

So I investigated further by watching the variables in a simulator and noticed this. (before anyone asks the simulator running the code on 16f1936 works ok.)

PIC16F1936
CRC16 = $FFFF
CRC16 >> 8 = $01FF

PIC18F46K20
CRC16 = $FFFF
CRC16 >> 8 = $FFFF

I also noticed the bit shift operation is not effecting the high byte of the word variables in the loop when running on the 18f46k20.

So am I going mad or is there something I do not understand that is different between the pic16 and pic18's.

Help please...

bambi123