2007 Freightliner Columbia Series 60: ECM data drops

I have a 2007 Freightliner Columbia with a Detroit Series 60/DDEC. I have been chasing a dangerous intermittent electrical/ECM issue for several weeks and would appreciate input from anyone who has experienced this exact behavior.

Symptoms

The truck can operate normally for many hours. Without warning:

• Tachometer drops to zero.
• Oil-pressure gauge drops to zero.
• Dash voltage displays 0.0 V or “No Data.”
• Accelerator becomes completely unresponsive.
• The engine may continue idling smoothly for a period before eventually shutting down.
• The fuel gauge, air-pressure gauges, odometer and some warning lamps remain powered.
• After the event, my diagnostic scanner may be unable to connect to the engine ECM and displays the equivalent of “make sure ignition is on.”
• Cycling the key sometimes restores operation. Earlier incidents occasionally required pulling the ECM fuse.

In the attached video, the tachometer, oil-pressure gauge and ECM-reported voltage repeatedly disappear and recover on approximately a ten-second cycle. The engine continues sounding smooth while the tachometer reads zero, so the engine is clearly not physically stopping during each gauge dropout.

Earlier behavior

During the first incidents, the engine coughed, surged or ran unstable before shutting down. A broken/corroded ECM battery-positive ring terminal was discovered and repaired. That improved the violent coughing behavior, but the gauge-data loss, no-throttle condition and occasional shutdown remain.

One of the four ECM ground wires was also found damaged and had previously been repaired with a crimp connector. I verified continuity from the repair to the ECM terminal, found no measurable voltage drop, inspected the female terminal, and performed repeated wiggle tests without reproducing the fault.

Testing and repairs already completed

• ECM has four battery-positive and four ground connections.
• All ECM positives and grounds test correctly during normal operation.
• ECM connectors appear clean, tight and free of corrosion or heat damage.
• No measurable voltage drop has been found through the tested ECM circuits.
• Harness has been repeatedly flexed and wiggle-tested with no effect.
• Tapping the ECM currently causes no change.
• Ignition switch replaced.
• Damaged remote PDM replaced.
• Main frame, starter and PDM grounds cleaned and secured.
• Corroded three-wire connector near the ABS/PDM area replaced.
• Charging voltage remains approximately 13.5–13.7 V.
• Battery voltage does not actually collapse when the dash displays 0.0 V.
• Alternator ripple tested essentially zero.
• ABS ECU was disconnected for a stationary cold-start test. The truck started normally with only “NO ABS” and the expected ABS lamp. Because the fault is intermittent, that single test was inconclusive.
• Starter solenoid has been identified as an intermittent-duty starting contactor and is not active while driving.

Important detail

During the more recent failures, the engine can continue idling smoothly while the ECM-related gauges disappear and the accelerator stops responding. This makes me question whether it is:

1. An internal DDEC ECM communication or power-supply failure
2. Loss of the keyed ignition/wake input
3. A J1587/J1708 network problem or another module pulling down the data link
4. An intermittent OEM/engine harness problem that passes all stationary tests

I am now considering purchasing a programmed replacement ECM. The supplier permits returns with a $200 restocking fee, which would allow it to function as an expensive substitution test. A complete harness is the other option, but I do not want to replace the wrong harness section.

Questions

Has anyone seen this exact combination: engine keeps idling smoothly, tach/oil/voltage data disappear, scanner cannot connect, accelerator stops responding, and the engine later shuts down?

Would you substitute the DDEC ECM first, install temporary overlay power/ground/ignition circuits, or concentrate on the J1587/J1708 network?

Is there a specific DDEC pin, ignition circuit, network module or known Columbia junction that commonly causes this?