Having just coded up the VHDL for the CRC computations for both ethernet and HDLC then I would suspect that the internals of the 6702 are a state machine that depends upon the value written and the current value. The XOR operator and logical shifts (left or right) or rotates (left or right) will also come into the equation. So keep an eye out for these operators in the numbers that get spat out. As Segher has already hinted - convert the numbers to hex, octal or binary rather than decimal and you may see the pattern easier. If we can get a printout of the exact number sequence on this thread (and the values you poked in to obtain them) this would allow some degree of 'parallel processing' to the problem. Multiple hackers working in different time zones all with different ideas working on a single problem - either a recipe for disaster or success! Cheers, Dave > Message Received: May 10 2012, 07:42 AM > From: "Segher Boessenkool" > To: email@example.com > Cc: > Subject: Re: 6702 chip > > >> The 24 numbers I got on my first frequency distribution are: > >> > >> 3, 7, 34, 35, 38, 39, 81, 85, 112, 113, 116, 117, 130, 134, 162, > >> 163, 166, > >> 167, 208, 212, 240, 241, 244, 245 > > > > I don't know whether this is of any help, but after a quick look at > > the binary representations of these numbers the only pattern I find is > > that bit 3 is 0 in all of them. > > Oh, there is a lot more structure than that: those numbers > are exactly > > 01110000 + a0.00000001 + a1.01010010 + a2.00000100 + a7.10000000 > 11010000 + a0.10000001 + a1.01010010 + a2.00000100 > > (in binary, a0,a1,a2,a7 are 0,1, + is exclusive or). > > William, can you post those 24 numbers in the order you got them > out of the chip please? > > > Segher > > > Message was sent through the cbm-hackers mailing list > Message was sent through the cbm-hackers mailing listReceived on 2012-05-10 19:00:14
Archive generated by hypermail 2.2.0.