Daniel

Manchester coding as I understand it:

The first 1/2 of the bit coming in is the value you want, the next 1/2 should be the inverse of this, if it isn't then an error arrived, this is easiest to see with a synchronous datastream as you clock in each 1/2 bit compare it with the next 1/2 bit for validity, then perform your action on this. (shift the valid bit into a byte or dump the transmission)

It would be possible to read in every second 1/2 bit and this would be the data - but this means you would miss any errors (a CRC in the message would pick it up of course)


Manchester Coding was designed to allow for an infinite length string of either 1's or 0's to be sent without losing track of where you are in the stream. NRZ (Non Return to Zero?) is the opposite of Manchester Coding, this where you simply send the bit.