W210 CAN bus issues - PeachParts Mercedes-Benz Forum

jay_bob · #1 06-20-2018, 04:47 PM

Ok now it’s my turn for more fun. Sold the wagon to VStech and sure enough 2 of my 4 cars decided to protest. At the same time.

On Saturday, the E320 has lost its Y3/8n2 sensor on the conductor plate. It’s going to the dealer tomorrow, can’t DIY this one. $$$$.

You have to do a 722.9 transmission service on a lift since it’s a bottom filler. I have no easy access to a lift. And to make matters worse, they won’t sell you a conductor plate, because you have to code the new plate to the car, and it requires legit Xentry that can phone home. I knew this going in, I pulled diagnostics when I bought the car, and the error was in the history. I got the seller to drop the asking price about what it will cost for the dealer to fix. Just bad timing.

But on to the main part of the story:
On Sunday I did an oil filter and air filter change on the E300. 248k miles, been practically trouble free the 5 years I have owned it.
Got all done and went to crank it and it would crank but not start.
Acting like it had no fuel. It would spin for a few seconds and sound like it was going to catch and then the auto crank would stop the starter.
WTF.

Dug into it and determined that the fuel valve was not opening. No voltage at pin 4 of X22 (the fuel valve drive signal).

I thought, the only thing I unplugged was the MAF to make it easier to take off the air filter lid to change the filter. I checked and I had +12 and +5 V at the sensor, so that ruled out losing the sensor voltages from the ECU.

Tried to open the trunk and the PSE pump was cycling but the trunk would not open.

Got the computer hooked up and the CAN bus system is throwing faults left and right. On both the engine bus and the body bus.

I was able to clear the faults and everything scanned clear (except for one door control module that was reporting failed). I was then able to start the car like nothing happened.

Took the car for a shakeout run and sure enough about 5 miles from the house, it just flat died going down the road. I coasted into a parking lot and hooked the computer back up. Same thing, practically every CAN related fault showing on both buses. Cleared again and got it back to the house.

Had to run to Texas for a work trip Monday morning through today. So luckily older brother is out of town so younger brother could drive the ML320.

Obviously something failed non related to the oil change.

Maybe EIS failure, that is the device in common to the two systems.

Anyone out there got any ideas?

I’m going to try to get on it after dinner, first thing I am going to check is system voltage. I wonder if the alternator is going south.

shertex · #2 06-20-2018, 04:56 PM

Yikes...interested to see how this plays out.

TMAllison · #3 06-20-2018, 05:31 PM

This is the E300 right? How old is the K40 relay.

jay_bob · #4 06-20-2018, 05:57 PM

Not sure, I haven’t replaced it. I can look for date codes when I get out there to troubleshoot in a little while.

Bolognablake · #5 06-20-2018, 07:21 PM

First, I'm not a Mercedes mechanic or a mechanic at all, really.

In my experience, when the can bus goes down, it's almost always one of a few things.

1) Broke/grounded wire. A broke or bad connection can be hard to diagnosis with a multimeter, you need to load the circuit. But if you're seen multiple faults, this is less likely.

2) A module on the can bus is bringing down the entire network. On a crank/no start where there's no CAN activity or faulty activity (i.e. low amplitude or the wave forms don't match) I've had a lot of success unplugging modules and seeing if the network comes back. I've seen many an ABS module take down a CAN network causing a no start. You pull the connector and brake fluid starts pouring out.

3) ECU. It happens.

shertex · #6 06-20-2018, 08:02 PM

Good to replace K40 "just because." Cheap and can cause all manner of electrical weirdness.

jay_bob · #7 06-20-2018, 08:57 PM

Well guess what. System voltage 12.1 V at idle (measured with my Fluke 87 at the jump start lug under the hood). Alternator time! I got 248k miles, not bad for the original alternator.

I suspect that the system voltage crashed while driving thus taking down the whole network.

With the car on a battery charger, all the engine CAN bus nodes (ABS, ETC, IFI, IC) report as healthy (ERE, Actual Values).

I do have an interior can bus issue that I have to run down but I don't think that's the issue with the engine control.

I found the screen in the HHT emulator (EZS, Actuations) that gives the run down on the whole body CAN network, and I have a CAN-H fault except when DCM3 (rear drivers door) module is unplugged from the harmonica.

97 SL320 · #8 06-21-2018, 10:05 AM

Quote:

Originally Posted by jay_bob

Well guess what. System voltage 12.1 V at idle (measured with my Fluke 87 at the jump start lug under the hood). Alternator time! I got 248k miles, not bad for the original alternator.

I suspect that the system voltage crashed while driving thus taking down the whole network.

12.1 isn't very low, cranking voltage will drop below that. I've had electrically noisy alternators cause odd problems on other cars, did you hang a scope on the alt output?

I did have an ignition module on a VW become overly sensitive to voltage causing a run problem. It would run fine with the alternator unplugged once battery voltage dropped a bit.

jay_bob · #9 06-21-2018, 12:37 PM

That 12.1 V was at idle and no appreciable load (climate fan off, lights off). I would expect at least 13.xx V.

I had the min/max function engaged so I could check the volt drop on crank but the lead fell off when the engine cranked so I didn’t get a good reading. I will try again later to pull this number.

I don’t have a scope but I did switch it to AC volts and had less than 100 mV so it’s not a blown diode.

What concerns me the most is we did not get a charge warning on the dash. They take the D+ to the dash but instead of the traditional bulb they have a resistor and a voltage input to the control logic.

I already have another alternator on order and it should be here tomorrow. I’m changing it on principle anyway (along with a new belt!) due to the age of the car and the fact that it will be at college with my son this fall.

97 SL320 · #10 06-23-2018, 06:36 PM

Quote:

Originally Posted by jay_bob

That 12.1 V was at idle and no appreciable load (climate fan off, lights off). I would expect at least 13.xx V.

Yes 12.1 is low for a running engine, but 12.1 should not prevent the car from continuing to run after initial start. Think about what happens when the alternator belt is removed to check for noises.

Quote:

Originally Posted by jay_bob

What concerns me the most is we did not get a charge warning on the dash. They take the D+ to the dash but instead of the traditional bulb they have a resistor and a voltage input to the control logic.

Perhaps the voltage regulator is working enough to energize the field enough to shut the light off but not enough for ample charging voltage.

jay_bob · #11 06-23-2018, 06:52 PM

That is what I’m thinking, regulator maybe going out. A new (rebuilt) alternator will be here Monday. I will be putting it in sometime next week when I get a work deadline out of the way.

jay_bob · #12 06-30-2018, 08:31 AM

Update:

Got the new alternator installed yesterday. System voltage 13.92 V at idle now.

If anyone following needs to do this, on a turbo engine, you will need to remove the charge air pipe from the turbo outlet to the intercooler rubber hose on the right side. One of the mounting bolts for this pipe is shared with the alternator bolt. It is an E-torx bolt and the mount point on the pipe is slotted. There are 2 more 10 mm bolts along the length of the pipe that have to come out and then it just drops out.

You can’t pull the alternator from the top, the water line connection to the radiator runs right over the top of the alternator.

Also you’ll need to remove the fan and shroud. It gets in the way just enough that it makes it impossible to extract the upper bolt for the alternator without it getting caught up in the fan.

And one more hint, you will need your lug wrench to release the belt tensioner. Well of course mine was locked in the trunk - if you disconnect the battery you cannot open the trunk on a W210. Luckily I had 3 other MBs around so we borrowed the wrench from one of the others.

Bolt torques: D+ lead 4 Nm, B+ lead 15 Nm, mount bolts 42 Nm. I think the bottom bolt was an E16 and the top an E14.

I also went to pull a part and found a left rear driver side window motor. I replaced this and now the system scans completely clean with no errors. The motor can be easily detached from the regulator - loosen two T25 screws and rotate the motor about 45 degrees and it pulls off the gear in the regulator frame.

I noticed the car that I had pulled the motor from that the regulator cable was a spaghetti pot in the bottom of the door. Not surprised. When I got the motor plugged into my car it would not run in one direction. On a hunch I put the car on diagnostics and found that the motor was inhibited due to being jammed previously. I was able to reset this and then the motor worked perfectly. I just had to retrain the up and down limits.

I will take the car out for a run today to make sure all is good but hopefully this is the end of this saga.

vstech · #13 06-30-2018, 11:00 AM

Yup... cars stick together

Ya get rid of their buddies and they protest.

Of course... they could be trying to get freedom as well...

Grass is always greener and all that.

97 SL320 · #14 06-30-2018, 11:10 AM

Did you try to run the car after the alternator change?

Was the window motor motor only or does it have the CAN box built in?

What does this system use to determine window position? ( potentiometer , motor shaft encoder. )

If changing the motor was solution to the engine run problem, I wonder if a window position sensor failure to electrical ground could drag the CAN box down. ( RE: Voltage supply to position sensor is also used for CAN box functions. )

Another possibility is one leg of motor was shorted to ground and pulling power down. RE: The system is likely designed to run a H bridge with power to both sides of the motor when it is not moving. If one leg is pulled to ground, voltage to CAN box will fall.

jay_bob · #15 06-30-2018, 12:37 PM

Yes I have run the car after the alternator change. I am going to take it for a drive this afternoon before releasing it back to my son. I will have my computer with me just in case it locks up again.

On the rear windows, the CAN box is built into the motor. It uses two Hall sensors looking at the rotor to count revolutions and look for a stall (for pinch protection, or if the mechanism self destructs as is so common with these).

The way you train the system is by running the window up by pushing and holding at the momentary position until it senses the jam from hitting the top of the track. Then you run it down to the bottom in the same manner. It counts the revolutions between the two points so that is how it knows how to do auto open/close.

I took the old one apart for fun.

The H bridge is a typical arrangement on motor controls like this. MB has been using that arrangement since the direct wired window switch days. Of course now they use relays or MOSFETS instead of the switch points.

The motor has the following leads:
Power and ground (fed from a fused circuit)
CAN-H and CAN-L
3 wires to the local window switch (uses multiple resistors to determine position of the switch).
2 wires to the courtesy light (switched within the motor control box).

All the sensing and control is contained within the window motor assembly.

Thread Tools
Show Printable Version Email this Page
Display Modes
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode