McGMT said:
Ah, I see what is happening now.. I think... All the reference leads from certain assemblies head to the PCM (Or BCM whichever) from multiple sources with different "channels" or "PID's" over a single wire that comes from a block that all the ref wires go to.
The OBDII data bus is the one critical wire that goes to all modules. The PCM has a bunch of different and separated 5V reference (reference in this case just meaning electrically clean, accurate, and not messed with too much by other circuits). If the designers weren't limited by the number of pins available in the connectors, the "best" way (not the cheapest) would be to have one reference voltage for any sensor that needs one. Since they were constrained, they often grouped things together and used a shared reference voltage. This allows for unwanted interaction between circuits, though. The interaction means that multiple sensors can interact with each other and make each other look bad.
So when the clutch is flaking out and hitting the ref lead with 12V it is temporarily taking out the signals from whatever other assemblies are on that branch.
That's right, but the only sensors sharing this specific 5V reference are two on the accel pedal, two on the throttle body, and one on the fan clutch. The redundancy on the accel pedal and throttle body are to make sure that a single sensor or wiring fault won't cause a WOT behavior and the vehicle would runaway like a Toyota. That's why faults on this specific reference circuit give you a REP light and only a "limp home or to the side of the road" behavior. Yes, to put the fan clutch on the same circuit as the accel pedal and throttle body was a very poor design decision.
Unhooking the fan clutch takes that out of the equation to let the rest communicate freely without the "noise".
Pretty much, but the noise is really not a data communication message packet fault, but a true analog crosstalk or spike on the sensor reference. The sensors on the accel pedal and throttle body are basically potentiometers.
Which is why the engineers ran all the leads to a common block instead of daisy chain so that if one assembly was to completely lose com. you weren't disabling an entire group of systems because one system or even one wire at that went down.
Right ?
This is the slight misconception you have, and it isn't bad. The "blocks" you're thinking of are the "Splice Packs", two locations where the data communication wires all come together to a shorting plug. The messages are never really run as a daisy chain where each module has to retransmit the signal to the next one in the chain,
View attachment 22717
and it isn't wired as a common long wire (backbone) where each module touches the data comm wire but it's basically all of them always tied together.
View attachment 22718
They come together in the Splice Packs in case one module starts putting jibberish on the line and garbling comm on the wire for all modules, like one guy heckling a speech and then you can't hear the main speaker. This is called a star configuration:
View attachment 22719
You can pull the wire from one module at a time out of the Splice Pack at the center of the star and see if the gibberish disappears and the remaining modules can communicate properly. They actually did this SOLELY to make it possible to troubleshoot down the line, because it's a total PIA to get to each of the modules and just disconnect their data comm line to see who's talking when they shouldn't be.