Hi,
This thread takes me back ![]()
Rob's state machine was based on testing the 4 bit nibbles, so you'd have to rewrite that too, not just the getNibble() function. I'd be concerned though that while the previous logic could be out of sync with the stream by up to 3 bits, this could get lost by up to 7. This would lead to potentially more than double the data loss on those occasions while the sketch found the correct preamble again.
Why would you want to do it byte by byte? Didn't you have this working as at your previous post?
Geoff