Vauxhall Corsa P1615/P1616 Cranked but Would Not Start: The ECU Was Isolated by One Corroded CAN Wire

P1615/P1616: Test Can At The Ecu

The 2004 Vauxhall Corsa arrived in pieces. The dashboard and engine trim had been removed, and the engine control module had already been condemned and taken away by a previous repairer. It cranked strongly but would not fire. P1615 and P1616 said the ECM was receiving the wrong vehicle identity from the body controller and instrument cluster. Both of those modules communicated with each other. The ECM was not necessarily bad; it was standing at the end of a damaged CAN branch.

An automotive meter/scope becomes useful when the network map identifies where to compare the bus. The AUTOOL DM303 can support accessible K/CAN voltage, waveform and continuity checks at known pins. DM303 can support accessible K/CAN waveform, voltage and continuity checks after the Corsa network map and pins are confirmed; it cannot decode every message, prove an ECU internally healthy or expose an open spur from a resistance reading at the diagnostic connector.

The defining idea is measurement location. A healthy waveform at the diagnostic socket or another module can clear the main trunk while completely missing an open spur. Work from the topology, preserve the twisted pair and never pierce or load CAN wires casually. Stop if the correct network diagram, breakout method or module pinout is unavailable; an improvised probe can create the communication fault being investigated.

P1615 and P1616 describe a failed conversation

The missing oil warning lamp alongside an illuminated engine lamp was an early behavioral clue on this model, but it was not proof. The two identity codes and a no-start with good cranking shifted attention away from fuel and position sensors toward the data exchange required for authorization. A scan tool could still communicate with the ECM and operate some outputs, showing that the module had at least some power and life. Partial function did not establish a clean network path.

Read the dashboard behavior before replacing the ECM

The body controller and instrument cluster exchanged data without their own communication codes. Both were accused only from the ECM’s perspective, making their shared route to the ECM the common question. A simple topology sketch separated the cabin CAN trunk from the engine-bay takeoff. That map was more valuable than a generic resistance number because it identified where a local break could isolate one node while the rest of the network continued normally.

Draw the network trunk and the engine-bay branch

At the ECM, CAN High retained recognizable structure while CAN Low sat low and lacked valid bit formation. The differential signal was therefore corrupted where the module needed to read it. Measurements at the ABS and body controller looked normal. The comparison localized the defect to the engine-bay section between the trunk splice and ECM rather than to the entire bus or two unrelated identity-producing modules.

Compare CAN at more than one node

Checking only 60 ohms at the diagnostic connector could have passed because the two terminating resistors and main trunk were still present. The ECM branch was a spur. An open or high-resistance conductor on that short takeoff may contribute little to the resistance measured elsewhere. The correct question is not ‘does the network measure 60 ohms somewhere?’ but ‘does this module see the same two-wire signal as a known healthy node?’

Test point	Observation	Meaning
ABS / body-controller area	CAN structure normal	Main trunk is functioning here
ECM pins	CAN-L malformed	Fault lies in ECM node or local branch
Engine-harness movement	Signal and starting return briefly	Physical branch fault localized
Opened conduit	Corroded strands nearly severed	Cause confirmed before ECM replacement

A good diagnostic-socket reading can miss an open spur

A controlled wiggle test on the engine harness changed CAN-L structure. At one point the waveform nearly normalized and the engine started—the first successful start in a long time. Movement was used as localization evidence, not as the repair. Leaving the loom in the position that worked would invite heat, vibration and water to reopen the connection without warning.

Use movement only to localize—not to finish—the repair

Opening the conduit in the localized area revealed corroded CAN strands worn almost through, likely after chafing inside the protective sleeve. The damaged pair was repaired with the correct conductor, splice method and twist rate, then resealed and supported away from the abrasion point. Both wires and nearby circuits were checked; CAN balance can be harmed by an untwisted or bulky repair even when continuity returns.

Open the conduit and rebuild the twisted pair correctly

After repair, CAN High and Low formed the expected mirror relationship at the ECM, communication remained stable and the Corsa started repeatedly. P1615 and P1616 cleared and stayed away. The trims were reassembled only after repeated heat, movement and restart tests, because access was part of the verification opportunity.

Verify start, communication and waveform at the ECM

Another Corsa with these codes may have programming, power, ground, connector, module or immobilizer problems. Do not order an ECM from this case alone. Compare module perspectives, draw the topology and measure at the node that reports the complaint. If the vehicle arrives dismantled, inventory connectors and previous interventions first; a new fault may have been added during the earlier search.

The Corsa did not need another computer opinion; it needed the same conversation listened to at two places. That is the lasting diagnostic habit. A bus can be healthy for most modules and unusable for one. Measure what the isolated module actually sees, repair the branch as a network component, and prove the waveform before putting the dashboard back together.