Intermittent stall then no crank

I have some interesting problem with a 2012 GMC Sierra, 5.3L gas. At times (maybe 15 minutes after a cold start or 5 minutes after hot start) the engine will stall and cut off. This is usually preceded with some dash warnings like reduced engine power, service stabilitrak, etc. just prior to shutoff. After the shutoff occurs, I will encounter a no crank condition for a while. If it sits a while, it will start back up with a strong sounding crank (meaning the crank isn't weak/slow). I can bypass the cool down period by removing the battery cable and reconnecting the cable. This leads me to believe its some sort of module resetting itself. I was curious about the CAN bus and if the power/ground/comm integrity of the lines were good, so I put a scope on them to monitor. I got some interesting signals from the CAN that I am not sure if they are good or not. The CAN signals were taken at the ECM comm wires and the ground was hooked up to the ECM ground. I also used battery ground, but the signal looked the same. I attached a screen capture showing some communication packets look good and others have some funky behavior.

When I monitor the CAN bus with engine off, it looks fine (nice square waves, not sure about actual data sent) but with the engine running, there is some erratic behavior. It tends to get worse as the engine runs (maybe heat related?). Where I am at now is if the erratic CAN signal could be a sign of a problem or the behavior is actually normal. I don't have too much experience with CAN issues so I don't know if the erratic looking signal is typical and I am chasing ghosts or it is a sign of a problem and I should pursue further.

Definitely looks like a bad signal - does the ECU perhaps get its ground from the engine? I've seen quite a few loose ecu earths on engines.

Stored codes in any modules? The Stabilitrak message means you're gonna find at least one in the EBCM. That waveform definitely looks weird, but GM's are usually on point about setting U-codes when there's an issue. To be clear, you're connected to the HSLAN pins at the ECM, right?

After the stall, when you turn the key on, does the little __ mark show up beneath any of the letters of the PRNDL on the dash? No mark would back up a network issue. Seeing the mark would suggest the HSLAN is OK. This isn't a fool proof test, just a quick indicator.

While we're talking about the dash, do all the other dash lights come on as they normally would during the issue? Head lights come on? I've seen more than one of these trucks with intermittent negative battery cable issues, usually right at the crimp for the cable end. That might explain why disconnecting and reconnecting the battery fixes the issue. :huh: Is the negative cable end hot to the touch?

Thank you both for the replies. That's why I really appreciate this forum, very helpful and thoughtful responses.

Desmond, so I did check the engine grounds. Actually, when I bought the truck it had a green engine to body ground strap and green corroded negative battery cable. While those were replaced prior to this scope capture, I wonder if those issues may have played a part in the problems.

Tyler, previously I had no engine codes when the symptoms were present. After about 5 more times stalling, one did appear: a P0601 which is some sort of internal memory error for the ECM. I checked the powers and grounds and they all were maintained during a stall event. Eventually I did replace the ECM and solve my problems. To answer your other questions, I can't recall if the _ under the PRNDL display on the IPC was good, but I have a feeling it was. I found a really great video by Pico that discussed CAN diagnostics and serial decoding that helped me perform some additional diagnosis and check on the weird looking waveform on the CAN bus
Pico CAN testing .
One of my issues is I don't have access to a scan toolanymore, but using their method they had some good ideas about checking for communication issues with modules. Another interesting note is regarding that funky CAN signal. Apparently, the module takes the high minus low (or A - value, so it essentially turns into a 0-2 volt square wave. So even though it had that funky signal, since it happens equally on both the high and low CAN, it still turns out to a 0-2 V square wave. Using the math channel on Pico, I created the high minus low on the plot below which is the black line.

Even with the new working ECM, I still get that erratic plot. Perhaps some interference or a shaky ground on a particular module providing that specific data packet. One interesting thing the Pico video provides is some great ideas on how to isolate which module is creating that packet with serial decoding. The video is an hour long, but it is very instructive on some troubleshooting techniques and ideas.