Join Now
International Automotive Technicians Network
Understand How the System Works!
Posted to Technical Discussion Forum on 4/24/2013 31 Replies

As promised in a reply to another thread last week ( ) here is a writeup that I think illustrates the necessity of understanding exactly how the system you are dealing with works:


This client called stating that he was driving back from San Antonio when the "Coolant Low, Visit Workshop" warning popped up on the instrument cluster just before he got to Katy, where he lives. He pulled over and shut the car off and looked at the coolant level but it seemed okay. The car didn't seem to be overheating, so he got back in the car to limp it home and he was going to bring it in to have it checked out the following day, but it wouldn't start so he had it towed.

Furthermore, we got a call from the wrecker driver asking how to get the car out of park, as it would not do so. Apparently, the car eventually came out of park because nothing was disconnected.

When the car was dropped off, it naturally started right up. However, it hunted for idle, banged into gear and generally acted like it was possessed by a legion of very angry demons. This short video displays some of the symptoms. [2003 Mercedes-Benz E320, Video]

Scanning the car, I found that essentially nothing on CAN B was talking. However everything seemed to be working in some fashion: A/C worked, windows rolled up and down, seats moved, etc. (This is actually an important clue, as we find out later.) Looking at this print, notice all the exclamation points. [2003 Mercedes-Benz E320, Photo]

For those of you not familiar with SDS, a check mark means everything is okay, an "f" means a stored fault, an "F" means a current fault and a ! means no communication with that module.


So, where do I go from here? Do I start chucking control modules around? Do I "check all the fuses" (my all-time favorite diagnostic time waster found frequently in the TechHelp)? Do I get out the test light and start ripping the car apart and "checking for power" at every control module?

No. I'm going to start with the base electrical system. My rationale is that nothing electrical is going to work right unless the base electrical system is functioning properly. So I hook up my Midtronics to the battery and get started. It gives me the "System Noise. Check Consumers" message. No big deal, I think, I just shut the car off.

I go do something else and when I come back, I still get the same message. Hmmmm.... CAN buss is not going to sleep. (This is another clue.)

Alright, now I'm ready to start tearing stuff apart, right?

Wait, not so fast!

Now it is time to go sit down (in an ideal world, it would be with a Spaten Optimator or a Franziskanner Hefe-Weizen) and start reading.

What I am looking for is where any points of commonality among all those control modules are. What I discover is that what is called a Central Gateway is the hub that ties all the various communications busses together: CAN C, CAN B, MOST and CAN D. CAN D is the diagnostic buss through which the various networks communicate with each other and with X11/4 (the OBD II connector.)

Now, I was actually able to skip that step, because I already knew that information, but I included it for illustrative purposes. There are training documents from MBUSA out on Networking and it can also be pieced together out of AllData or WIS.

So I drop the under dash panel and go straight to the CGW. [2003 Mercedes-Benz E320, Photo] I unplug it and look what I find. Corrosion on one of the connectors. [2003 Mercedes-Benz E320, Photo] Lo and behold, the CAN buss goes to sleep as well. I think I've found the culprit.

So, with our client's approval we order up a new CGW and some stuff to make a new connector.

I make the new connector, put the new CGW in, get it programmed, and everything comes back up and everyone is talking now. I clean up all the fault memories and take the car for a drive. Beautiful. Everything works.

In the meantime, MB has a bulletin out concerning water intrusion through seams in the fenders near the A pillars that required 8 - 10 hours worth of work. Since the slight bit of corrosion on the CGW is the only real evidence of water intrusion -- no damp carpets or anything -- and it has taken 11 years for it to happen, (There's a surprise right there, an 11-year-old W211 that still has it's original CGW!) I elect instead to relocate it up off of the under dash panel and put a service loop in the wiring for any moisture to drip off of. [2003 Mercedes-Benz E320, Photo] That move is inspired by a bulletin for repairing water-damaged PSEs on the 170-body SLKs.

The whole diagnosis took less time than it did for me to write all this up.

How long would it have taken If I jumped right in and started tearing stuff apart without first studying and learning how the networking on this car functions?



And at what cost?

Byron from Texas

Files Referenced:

thumbnail thumbnail thumbnail thumbnail thumbnail

31 Replies Received (View Replies)


car Vehicle Data

2003 Mercedes-Benz E320 3.2L

Engine3.2 L
Trans5-speed Automatic (Electronic)