b'Important clues, like module addressing, message verification from other modules (tracking) and other helpful data, are wholly and completely unavailable for technicians to obtain or use for diagnosing the vehicle. Strangely, this is even true on the OEM dealer level. Their techs cant make use of that data either. So, were all stuck figuring things out the hard way.So be it. At least were in it together!We can understand why things are done this way; its the OEMs only way to maintain engineering and software autonomy, deploy new technology, compete against other OEMs, and maintain some semblance of copyright and patent protection in a world where vehicle OEMs must now build their networks to a set of industry standards.But this paradigm doesnt make it any easier to getthese vehicles successfully diagnosed right the first time, when we suddenly find our diagnostic pathwayvaporized into the ether of coded data packets, algorithms, hex and binary processing, and ADC/DAC conversions. In other words, today there is a major information gap in this area for technicians, that leads to lots of misdiagnosis in the field. What to do?More often than not today, we find ourselves diagnosing more difficult cases, like multiple modules setting cascading CAN fault codes, long strings of UXXX-communication codes, and the worst of the worst in network diagnosis, the dreaded intermittents. The scan tool and network topology mapping, as weve learned, will only take a technician just so far, especially when a module or group of modules is offline to the scanner. So, we must rely on physical testing to isolate the fault.Next, well briefly describe the Physical Layer as ithow they interact to broadcast vehicle module dataThe Transparent Layer relates to the scope waveforms we can observe duringthrough the network helps give technicians anThe Transport Layer, which is the structure or the diagnosis of the bus. We often look at CANunderstanding of where we can and cannot play whenformatted carrier of the data transfer (the data waveforms, but if we can interpret what the waveformdiagnosing a CAN or network issue. This articlepacket itself), handles arbitration (who gets the is actually telling us, we can save massive amounts ofdiscusses using clues from the only two CAN layersbus first), routing, error, and control functions diagnostic time by using the characteristics of the that technicians can actually measure and decodeof the networks data frames. It also handles CAN waveform on the screen to guide our next steps(Physical and the Transport Layers), to quicklyhow they are moved and interpreted in relation in isolating the fault. narrow down the offender. to the electrical constraints and design of the The Physical Layer of Automotivephysical layer that transmits and carries the As we all know, it is also quite easy to out-techCAN and FlexRay Networks messages.ourselves straight into a misdiagnosis from time to time while solving complex networking issues likeThe Physical Layer of AutomotiveECUECUECU this. There are a lot of unwritten variables at playCAN and FlexRay Networks 1 2 nhere. The Physical Layer was summed up and simplified The realities of what we can really do in networknicely by a colleague recently, who provided a diagnosis: Where do we focus 0ur effortswell-rounded description of the Physical Layer of the productively? Some theory bears out the answers. CAN network as being the bus line wiresthe twisted pair, termination resistors, stubs, and CAN nodesR L R LCAN and other modulized communications networks(modules)the things you can physically touch, in are configured in functional layers. other words. Which is actually a really cool way toD Llook at the system.Understanding the basic flow of how the 7 CAN layers are constructed and configured (see Figure 2), and 8'