Faulty ECU, Bosch ME 2.1 from Mercedes CLK 200, W208,

[PiotrMisiuna]

Mercedes CLK 200, W208 a bunch of problems popping up and going away over time, now all those things are dead. finally the owner left the car to me to play with. In Cyprus where I'm from, quote for a repair by a specialist exceeds the value of the car.

Issue that immobilized the car was cranking and not starting because of the fuel pump not working. I figured, that if I bypass the relay in the trunk I am able to start the engine.
Also, the vacuum pump for the convenience lock, trunk does not work.
I bought a delphi 150e clone, but there is no communication with the car. I even made a makeshift break-out box. My ultra cheap DSO shell did not show any signals.

on this picture I show what measurements I took to get to conclusion that the fault is on board of the ECU.


I connected the ECU on the table to a power supply and I looked at the ECU with a thermal camera, but the only part that got remotely warmer was IC connected to the pump relay, I replaced it but nothing changed. I see no abnormalities around electrolytic capacitors, I didn't take them out thou.

I want to avoid experimenting with the only ECU I have without being able to backup the hash file. I thought about buying a programmer, but MC68HC11E9 chip needs more advanced techniques.

I usually don't whine too much but, PLEASE!!! is there anyone who could point me in the right direction, where to look for potential problems?

[iJoseph2]

a bunch of problems popping up and going away over time, now all those things are dead.

is there anyone who could point me in the right direction, where to look for potential problems?

check the flex wiring in the doors and boot lid .. maybe they're frayed and shorting.
check for water ingress
check ground points
check voltage drop from battery to main connections (eg when cranking look for voltage drop from battery negative to engine, and same for battery positive to major connection point)

a bunch of intermittent problems sounds to me like vehicle wiring

[narkeleptk]

a bunch of problems popping up and going away over time, now all those things are dead.

a bunch of intermittent problems sounds to me like vehicle wiring

As someone who works on ecu's daily, I'd have to agree with this. Unless there is liquid damage inside, random intermittent problems are not typical sounding problems from the engine controller.. The locks and trunk are 100% external of ecu functions.

I thought about buying a programmer, but MC68HC11E9 chip needs more advanced techniques.

I'd advise against this. I have read these ecus many times (including removing the hc11e9 from board and reading). For some reason this unit likes to stop accepting the chip properly after reading with a programmer. i work with many other hc11e9's with no problems, its just been on these units. I think its something to do with the security bytes getting damaged and then the master mcu on the units not happy about it.  I have better luck cloning these successfully using a device like alientech kess3 to read/write it through data lines. Its much better if you find someone who will clone it for you using one of these type tools.

[PiotrMisiuna]

I am no expert, but I already spent weeks on this car. I don't believe there is anything wrong with the harness, but of course I could be wrong. What makes me suspicious as to the ECU is the high resistance on all connections and voltages. Without the relay connected, the voltage that (I presume) should be pulled down to the ground sits above 4 volts, with the relay in circuit it is basically at  the battery voltage.

Moreover, I connected the ECU on the table to the power supply. It draws about 0.22A (not sure if it is nominal). The voltage at the same pin sits above 4V. I think it is suspicious, Is this normal?
The pin that in the car is at 12V, at the table is down to milivolts, I assume due to lack of ignition signal.

[darkflow]

I have no useful comment on the 4V observation, but I agree with other posters on the wiring suggestion - specifically ground / chassis straps. I'd be checking to make very sure that all ground straps in the vehicle are good low resistance connections. Potential rise across bad ground connections plays havoc with ECU functions.

[nctnico]

Check the CAN bus for defective wiring. Another common problem (in general) is that smart units like the instrument cluster may have developed faults. Unlike the ECU these typically contain cheap PCBs which can have bad solder joints on the connector. Recently I had a similar problem with my Ford and all it took to fix it was resoldering the connector on the instrument cluster.

The ECU looks like a quality build piece of electronics so don't touch that. It is highly unlikely that it is broken. Before anything else you have to do research about common problems; 99 out of 100 times other people have the same problem so see how they fixed it instead of guessing.

[Andreas]

What makes me suspicious as to the ECU is the high resistance on all connections and voltages. Without the relay connected, the voltage that (I presume) should be pulled down to the ground sits above 4 volts, with the relay in circuit it is basically at  the battery voltage.

Hello,

What you observe is part of the diagnostic function.
Usually each power stage has either a voltage divider or a constant current source of up to 1 mA to detect open load and short cirquits to ground within wiring.
A (switched off) open low side power stage has around 3.5V (between 2V and 5V) = open load error in fault code memory.

I also think that it is very unlikely that the ECU has a fault.

with best regards

Andreas

[vanders]

Many years ago had the same car ( cabrio ) but also W208.. and had once a non starter because key not recognized.. You already asked at digital-kaos.. or mhh auto forum?

[PiotrMisiuna]

I have no issue with the key. If I only bypass the fuel pump relay in the trunk I can start the engine and drive the car. The problem is related either in the ECU or it is affecting ECU.
I tried some automotive forums but they usually get into the same conclusion s I did.
The only way to solve it is to spend a couple of days with a multi meter and WIS to check every single wire. It is still a bit too hot outside in Cyprus thou

[Andreas]

Hmm,

is there any crash sensor which might have tripped?

with best regards

Andreas

[PiotrMisiuna]

could be, but this would came up on the scanner, so again - no communication to the ECU
Either there is a short in electrical affecting the ECU or it is inside the ECU. Since everyone here is telling me that the ECU is less probable to fail I need to exclude the first possibility.
 

[p.larner]

no comms with a scanner normaly means wrong obd protocol,get it scanned using a dealer spec tool,not sommat from a chinese xmas cracker.