Mercedes E300 | Warning lights and limp-home mode Part 1

In this ever-changing world, there is one constant in our trade: cars will break down. As we try to build our knowledge and the level of complexity increases, there are always things that can trip you up. This particular case was no exception, with twists and turns all the way!

We were asked to take a look at a 2014 Mercedes E300 BlueTec Hybrid. The reported issue was a large number of warnings on the dashboard accompanied by a warning message appearing on the instrument panel. The vehicle would then enter limp home mode before coming to a complete stop. Only switching off the vehicle and then restarting it returned the ability to drive the car.

When we got to the vehicle, the 12 V battery was almost flat and was unable to start the vehicle. We managed to connect a Bosch scan tool to do a quick global search and found a large number of fault codes, all relating to low voltage. There was also a couple of codes stored in the HV battery ECU for 12 V supply issues and P0A7F00 Battery Ageing Component Aged.

The first thing to do, though, was to get the battery off the vehicle and recharge it.

To recharge the battery, we used a CTEK MXS 10EC charger, which also reconditioned the battery before charging it back up to full capacity. There is a brilliant graphic for this charger which shows the different states of the recharging procedure.

We used PicoScope to monitor the charging process overnight after the charging program had started as I wanted to see how the battery performed.

We can see that over the course of nine hours, the battery received 8 A and towards the end of the charging procedure the current is gradually dropped as the voltage is brought up 14.48 V. When we compare the different stages of the charging program, I think we can say that the above capture was taken during the Bulk and Absorption phase. The battery was left on charge another night to complete the charging process and the next time I checked the charger it was displaying Float.

Next was to test the battery. For this, we have a rather antiquated device that basically puts a load on the battery. With PicoScope, we can visualise the effects the load has on the battery. I appreciate there are devices out there that will perform a battery capacitance test to verify if the battery is good or bad, but given how flat this battery was when the vehicle arrived, I would rather make sure that it’s capable of handling the types of load that it could encounter with a diesel hybrid vehicle.

This tester applied a load that pulled 170 A from the battery for 30 seconds. The battery dropped, as you would expect, to 10.79 V but the important thing to note here is how linear the voltage is when the load is applied. I’ve added another ruler in for Channel A to show how straight it is. If the battery was failing, I would expect this voltage to gradually fall away over time. For me, this was enough proof to say that this battery was good to go back on the vehicle to continue our testing.

As I wasn’t able to get back to the vehicle for some time, I thought I would utilise a PicoLog to monitor the battery voltage over a much longer period of time. Yes, graphing 12 V is extremely boring but it would show us if there was any deterioration. The beauty of PicoLog is that you can leave it running and with a recent update to PicoLog6 you can now use a cloud facility for captures and remote access to your PicoLog device when connected to a device with internet access.

Over another 12 hours, the battery voltage had dropped a little but was still at around 12.4 V.

With the battery back on the vehicle, we could now start the diagnosis of the warning lights. As the battery had been removed, the fault codes had all been erased, yet when carrying out another global scan with a scan tool, we still had P0A7F00 Battery Ageing Component Aged code.

At this point, we decided to see if the HV system was even working., We connected a 2000 A current clamp to the DC cable from the HV battery. Even if the vehicle couldn’t drive solely on electrical power, we should be able to see current going back to battery during regenerative braking.

A quick road test around the Pico car park gave us the result we were expecting.

In the above capture, positive current above the 0 A line is indicating that current is leaving the battery. Negative current is indicating current returning to the battery. This quick test might be all that’s needed to determine HV system operation. The test can be expanded to include brake and gas pedal position to give you a full overview. This can be found in the Electric vehicle section from the guided test menu in PS7. You can find more information on the recent updates to PicoScope 7 here.

Happy that the HV system appeared to do what we were expecting from it, we turned back to the fault code for battery ageing. After spending a lot of time searching for more details on this fault code, we weren’t getting any further and so we gave the helpful folk at HEVRA a call.

As always, they were both extremely helpful and knowledgeable and offered great advice. What is interesting is that this code is found by some scan tools and not by others. If there’s one thing we have learnt lately, it is that you almost need at least two scan tools to verify their results! Armed with this new information we hooked up a different scan tool and sure enough, there was now no fault for battery ageing. Just to double-check, we connected yet another scan tool and that also didn’t find a fault in the Battery ECU. As if the job isn’t hard enough, we also noticed that both these new scan tools reported that there are faults in the Engine ECU! Going back to our first scan tool we noticed that it hadn’t picked up the ECM at all and where we had so many fault codes, we did not notice the absence of this rather critical ECU. It’s an easy oversight but one we should have picked up on.

Clearing down all the fault codes and then carrying out yet another global scan left us with two fault codes in the engine ECU:

• P0106 MAP Sensor Range/Performance
• P0240 Turbo/Supercharger boost sensor “B” circuit Range/Performance

At last, some direction! The next stage was clearly to turn our attention to these codes and they would certainly explain the warning lights. Our first job was to use the serial data to determine if there was any output from the sensors during a snap-throttle test.

Focusing on the pressure sensors alone, we could see that the pressure sensor for the low-pressure turbo was not moving, despite the increase in engine speed. We could also see that the boost pressure sensor was reporting a change, indicating we had some boost pressure from the turbo. Next step was to locate the pressure sensors.

Images provided by AllData

Image provided by AllData

Based on accessibility alone, we took a look at the low-pressure sensor which was at the front of the engine. We used PicoScope to capture measurements at the sensor, again due to accessibility.

In the capture above, we have a power supply signal, a ground signal and a sensor signal, but how come the ECM wasn’t reporting any change during the road test.  

We advise that sensors should be measured at the ECU and actuators at the actuator, but I know from experience that this isn’t always straightforward. However, in this case, our next step was to locate the wire from the low-pressure sensor at the ECM.

In the above capture:

• Channel A is the signal output at the sensor
• Channel B is the sensor signal at the ECU
• Channel C is the crank sensor at the ECU with a filtered crank math channel to show WOT
• Channel D is the current at the HV battery

As you can see, the signal at the low-pressure boost sensor was reporting a change during the WOT test but the same signal at the ECU remained stationary. This could be associated with an open or short circuit. However, our fault code was for range/performance, and when we looked at the actual signal there was an output of around 3.8 V. This meant that there was something there but not the signal we were expecting. After doublechecking our wire diagram, we were confident that we had connected correctly to the low-pressure signal at the ECU. But how could we have a signal that didn’t match the sensor?

This was indeed a puzzle and one that took longer to solve than it should have, but it is always that way when you take on a job which has been passed through a number of hands.

It became clear that the low-pressure sensor was exactly the same as the one used for the downstream air filter pressure. The exact same connector and configuration and, best of all, it was more than possible to confuse the two connectors as they were next to each other!

I know this isn’t a particularly clear picture, but you can see the downstream air filter sensor and literally right alongside it is the low-pressure boost sensor. After swapping these connectors over we carried out a WOT test which gave us the following capture.

Now we were happy that the ECU was seeing the correct signal, we cleared the fault codes and everything appeared to be OK. Driveability was much improved and only when it was running correctly, did we realise what the performance should have been!

During the road test, due to this vehicle being a hybrid, the engine cut out to run on the electric motor. When slowing down on the approach to a junction and using the electric motor only, all was OK with no warning lights on the instrument cluster. However, when pulling away, the dashboard warnings all came back again and the vehicle suffered from a dramatic loss of power. We also noticed that the battery power gauge was stuck at MAX.

So, even though swapping the connectors helped the performance of the car and gave the sensors and the ECU the correct signals, we still hadn’t found the cause of all the issues in the initial complaint.

Read about the rest of this tricky but exciting case in Part 2 of the Mercedes E300 Hybrid case study