Shorted CAN lines
This 2012 GMC Savana van was towed in for no crank/ no start. On first check there was no communication from the PCM. Checking all modules, there was no comm from anything on the CAN lines, serial data was OK. I scoped the CAN signal and it showed the can signals (high and low) were identical, shorted together. Living in the rust belt all my life, I thought I should check under the van first. I disconnected the fuel pump module and BINGO, CAN signals come back normal. There was the nasty green stuff growing on one of the connector sockets and one pin in the module was corroded off completely. It looked like a slam-dunk with a new fuel pump module and a couple of repair terminals. The parts were installed and with the module programmed, the starter cranked and the engine fired right up.
With the job completed, the customer came and picked up the van. Two days later, the van returned on the wrecker, again no crank, no start. Rechecking my work, I checked the CAN signal. Again it appeared the lines were shorted together.
I disconnected the fuel pump module and immediately the signal returned to normal. Common sense tells me I have a bad fuel pump control module, but I fail to see why the replacement failed the exact same way as the first. The fuel pump power is fused and has a relay in the circuit. The way I see it, a bad fuel pump shouldn't have anything to do with shorted CAN lines, or am I wrong?
Any ideas out there?
In your first capture, the CAN lines don’t appear identical to me. They aren’t correct tho, I’d agree.
I’m wondering if something outside the FPCM is affecting it‘s Communication. I’d be unplugging the pump, and loading the power and grounds first.
BTW, the older vans had a FP ground wire that would fray at the bolt ring where it bolted to a body mount near the tank. Maybe the FPCM uses a similar ground ?
Are you sure about the power going through a relay I just looked on identifix and it looks like the 5.3 gets power and ground from the fuel pump flow control module. If the pump is going bad and drawing to much amps could be taking out your control module.
Walter, you are correct. There is a relay in the under hood fuse panel marked "fuel pump". This relay drives the fuel pump on models with a 4.8L or a 6.0L and those models do not use a FPCM. Go figure...
i had one like this but it was a yukon the module south of the FP mod caused my probelm when i unplugged the fp mod it was like unplugging the mod behind it on the line. maybe something to look at
Thanks for calling me to task, Justin. I assumed the CAN circuit was configured in parallel, when indeed it is in series. The TCM is the only module 'south' of the FPCM, and the proper signal returns when the TCM is unplugged and the FPCM is still connected.
I feel like the dog that chased the squirrel under the fence, and the dog sat and stared at that spot of the fence for days while the squirrel was a mile away!
A new Trans Control Module is ordered.
I agree with Justin above me, make sure you aren't "cutting off" another downstream module and that is actually the issue. I don't have 2012 in front of me, but a 2011 G-van with an LNF engine has the TCM /after/ the FPFCM.
I 100% hate silver bullets, but look to see if EBCM communication is available at any time, I have dealt with a very similar symptom on a chassis with similar network topology, looking at the information bus diagram. I am pretty positive(many RPO code variations) that the EBCM is down stream of the FPCM just looking at the terminating resistor location, this may be the issue. Also the EBCM can be bypassed, as it has can in and out, on the network to restore full use of vehicle to confirm the repair.
Okay, this story isn't done yet. I installed the TCM on this van but during programming I get the error code "E4399; server error; Reading VIN from device or controller." Hmmm, I could not communicate with any module using SPS. I disconnected the TCM and the message remains. I disconnected the FPCM, (the one I installed and programmed last week) and now I can read the other modules on the CAN bus. Figure 1 below shows the CAN bus with the FPCM unplugged, figure 2 shows the CAN bus with the FPCM plugged in; something is still not right... I jumpered the CAN terminals in the FPCM connector and reconnected the TCM and still get a scrambled CAN signal; TCM again unplugged and the signal straightens out, so the harness is okay. I sourced yet another FPCM (new from GM) and still get a scrambled CAN signal like the first. I don't believe that I getting a rash of bad modules from GM.
This vehicle will start and move with the TCM unplugged, and the scan tool shows no communication with the FPCM, but I can communicate with all the other modules. With the TCM plugged in there is no start and no communication with any module!
Is it possible there is for some reason something in the trans circuit feeding into the can lines? What is the voltage on the can lines? Ground circuit voltage drop?
I'd be loading powers and grounds on the FPCM and the TCM with a sealed beam.
Thanks Rusty and Scott. I seem to have a harness issue. If I would have remembered from CAN 101, we need to see 60 ohms resistance between pins 6 & 14; I get 120 ohms. I didn't test that until I replaced some parts (Cart/ Horse). One resistor is in the PCM the other is in the TCM at the other end of the line. Checking continuity on the CAN circuit between the FPCM and the TCM, I get 4ish megohms, and 6Kohms to ground. I have a little more digging to do, but I'll find it. ;-)
Oh yeah, the E4399 code is a failure to communicate message.