Seems that the reading of packets from the new protocol (CP2102) is resulting in broken readings - the packets are all there, just broken up. We need to be able to handle these by adjusting logic to be looking for our initial headers not just in one packet but in a number