Posted by John Chappel on January 25, 1998 at 11:47:02:
In Reply to: m0re qu3stions about lab 1 posted by Junya Ho on January 23, 1998 at 17:47:47:
: Does this mean we do not need to synchronize at the
: beginning of each packet using the synchronization field
: in the header?
: If we are to synchronize on the header's sync field:
: - how big is the field, and what part of the header is it?
: - are the alternating bits always 0xa?
: Question 2:
: Do the header and CRC portions of the packet get sent
: LSB first, like the data portion? Interoperability
: would suffer if the choice was arbitrary, and the
: sheet isn't explicit.
The details on the header contents (synch field, etc.) are informational only
and shouldn't effect how you design your circuit. For this lab, the header
is just 32 bits with the hex value 55aa1234. It still has to go through
NRZI encoding and bit stuffing like all of the data packets. In fact,
for transmitting you should just consider the header to be a constant value
that always preceeds the data.
In between packets the D+=0 and D-=0. For NRZI encoding it makes since to
consider this a zero. The most significant byte of the header is 55.
The least significant bit is a 1, so D+ remains at 0 and D- goes to 1.