Aftermarket May 2022

BY Ryan Colley, Elite Automotive Diagnostics A t times, we encounter troubling vehicles from other workshops because other techs have failed to draw a correct diagnosis. Failures found in a controller area network (CAN) can be as simple as an open wire or as complicated as a noisy network resulting from outside interference. Regardless, a thorough understanding is required to ensure an accurate and swift diagnosis. The vehicle in question was a Renault Megane 2, with a 1.5DCI engine. It had quite a few local garages puzzled, keeping this vehicle off the road for over six months. Knowing how these systems operate along with the use of an oscilloscope is crucial for analysing any networking fault, including this one. The CAN bus network is made up of various control modules, also known as nodes, all connected via two wires which send data packets to each other. They communicate via a binary signal (signal in either a recessive or dominant state) and transmit data at an average rate of 500Kbps (which is the equivalent to 0.5Mbps). Often, the voltages of the binary signal range from the following: CAN-HI= 2.5v – 3.5v CAN-LOW=2.5v – 1.5v Anything ‘CAN’ be fixed, if it’s understood The customer complained that the vehicle would not crank when the push-to-start button was pressed. However, the ignition would come on. They also noted that the cooling fan would operate with the ignition ‘on’. They mentioned multiple scan tools had been used. However, communication with the engine computer could not be established. I started by confirming the fault and noted that the engine management light (MIL) did not illuminate on the dash display, with the ignition ‘on’. With no-start complaints, paying attention to the MIL status during ignition ‘on’ is a great first observation, as this will often tell us whether the engine computer is online or not. I carried out a full system scan to find no fault codes present in the vehicle, but did note the engine control module was not detected on the scan. This indicated a communication problem for the engine control module. If this is not communicating with the rest of the vehicle, it will not start as the immobilizer data will not be shared between modules. Knowing how vehicle networking operates, as described earlier, is critical and can speed up your diagnostic process. For example, I now know I cannot communicate with the engine computer via CAN, but older engine computers often have a single serial data line known as K-line, which connects directly from the data link connector (DLC) to the 36 AFTERMARKET MAY 2022 TECHNICAL www.aftermarketonline.net POWERTRAIN VEHICLE NETWORKING A Renault Megane that was experiencing issues has given Ryan the chance to discuss vehicle networking engine ECU. Therefore, my next diagnostic step was to see if I could communicate via the K-line to the ECM. If so, this would confirm that the engine ECU is receiving the powers and grounds it needs to communicate. This can be done with most scan tools by simply carrying out an emissions on-board diagnostic (EOBD) scan. The result of this was established communication. It is now likely a CAN bus-related issue for the ECM. See Fig.1. Following the stepping-stones Knowing I likely have a CAN Bus fault for the ECM, the next job was to access the engine computer and verify the wiring integrity. It is crucial that powers and grounds are verified, along with CAN communication line integrity, before ever condemning a control unit. Powers and grounds were verified as ‘good’. Although this was already assumed, because communication was established with the ECM via K-line, it had to be confirmed. Next, I had to check the CAN bus directly at the engine computer. This check can only be performed accurately with an oscilloscope to truly verify its integrity. I found, with the ignition ‘on’, and the engine computer connected, the CAN high and low signals were shorted to ground, thus confirming our suspicions of a CAN Bus problem. See Fig.2 You must carry out this test with the engine ECU connected as well as disconnected. In case the ECM is internally faulted and is the root cause of the CAN bus short. If the signal, with the ECM disconnected, returns to the expected waveform, then it is likely the issue is related to the ECM, or its internal circuitry. As you can see, referring to Fig.3, with the engine ECU disconnected, the CAN high signal has returned correctly. However, there still exists an obvious issue with the CAN low signal. When checking these signals with an oscilloscope you are looking for uniformity and a mirror image of them, as you will see. The proof is in the pudding I have now proven that the CAN low signal leading to the