Technical Ducato x250 outages on the CAN bus.

[Coper]

I am solving a problem with a robotic gearbox that keeps saying speed is not available. Shifting can only be done in manual mode, and even with the neutral inserted. According to all tests and auto-calibration, the robot itself is fine.
The problem is in the communication between the robot ECU and the engine ECU. Apparently, when the robot's ECU does not have data from the engine ECU or about the position of the gas or brake for safety reasons, it claims that the speed is not available.

How did I figure it out?
Error P0856 appears sporadically, but it is related to CAN communication
In multiecuscan 4.8, if I let the pressure of the hydraulic fluid in the robot be displayed in the graph, the pressure in the graph regularly dropped to 0 and returned back.
Physically, the pressure did not drop anywhere, only the multiecuscan did not receive data or received incorrect (zero) data and wrote down the drop to 0 in the graph and read the correct data on the next reading.
In my opinion, the same phenomenon also occurs when the robot ECU read data fron the engine ECU, and I block the shifting when there are occasional outages.

The question is what about how to more closely identify and eliminate the problem on the CAN bus?

 

[theoneandonly]

thanks for your post it explains the graphs ive seen for a while. My zero sensor values are due to loss of communication not necessilary a dodgy sensor.
as to Can bus issues how to solve i'm afraid ive no clue other than an occilliscope is the first step. and identifying the pin outs on engine and robot ecus. Can hi and can lo are available on the OBD2 socket but will have too much data on it. I have seen a thread where someone has used a scope on the can bus but cant find it

 

[Coper]

In the meantime, I've researched the problem a bit and here's how it goes.
When I connect to the engine ECU and let multiecuscan graph the battery voltage read from the engine ECU, I get a straight line of 12.68V.
I connect to the robot's ECU and have the battery voltage plotted again. I will get this
As you can see, the voltage is only 12.0V (wrong, the battery was about 12.68V, the temperature was -2C) and the graph shows clearly visible blackouts. The same outages are visible even if I let the hydraulic fluid pressure from this unit be plotted.

Volt.jpg

• Coper
• Dec 6, 2023

Voltage readed from ECU

My plan is to get to the ECU, measure the voltage on the power pins1 GND and 27 KL30 Supply, if I measure 12V there will be a problem in the cables, if 12.68V I disassemble the ECU.
I don't have a better plan.

 

[theoneandonly]

what is 27 KL30 Supply?

 

[Coper]

Robot ECU have 80 pin connector
pin 1 and 2 = GND_PWR-Ground and
pin 27 is called KL30 Supply. I assume 12V power supply for the ECU
pin 28 is called KL15 Key Sense. I assume to 12V, but only when the key is in a position other than Stop
Based on the position and size of these pins, I assume they are power. In any case, the other power supplies are not

Conector.jpg

• Coper
• Dec 6, 2023

 

[theoneandonly]

Thanks for that.
Just checked you have an X250 and electrical info available on e-per. I was getting excited to think you might have ecu info on the X290.
you might find this guy useful on canbus stuff. this first one is

How To Test CAN BUS With A Multimeter​

if you end up needing to get a scope this thread shows what i bought and how little I know.

Injector Waveform Picoscope | FIAT Ducato

On a previous thread someone mentioned the use of an occiloscope in Auto diagnostics. Well i have a ducato and MES tells me the injector variances are greater than the FIAT spec of +-2 but under what conditions. So Ive watched a lot of you tube and bought a scope and an amp clamp probe. The...

www.fiatforum.com

 

[Coper]

Thanks, I'll look into it. I deduced the CAN bus error from the communication failures and from the P0856 error, which appeared twice.
Error on the CAN bus, but does not explain why the battery voltage measurement in the robot ECU shows 0.68V less than the reality and than the engine ECU correctly shows.
Plus, I remind you, outages in CAN bus communication are only when communicating with the robot ECU and not when communicating with the engine ECU. if at all it is a communication failure and not a measurement failure inside the ECU.
From the above I would say.
- the problem is either a bad ground or power line to the robot ECU, that's why the ECU only measures the battery voltage at 12.0V instead of the correct 12.68V.
* The second possibility is a faulty capacitor in the power supply inside the robot's ECU. If there is an electrolytic capacitor it could have lost capacity or has a high ESR, if there is a tantalum capacitor it could have broken and it has a cold junction.

 

[Coper]

So I disassembled the ECU of the robotic shifter.
There are three 47uF 35V electrolytic capacitors on the PCB, but they are fine (measured both capacitance and ESR), I added more capacitors in parallel to make sure, but the unit still measures lower battery voltage and has dropouts in communication.
The PCB is at least four layers.
12V power supply both directly from the battery and from the Key leads to a double diode STPS15H100C 2x7.5A 100V.
The voltage from the Key leads to a resistor divider of 1.5k and 2.7k to GND to a resistor of 27k. I assume that one is an input divider for battery voltage measurement and the other for key position detection. The detection then controls the PMOSFET through which the non-standby part of the ECU is powered.
There are two processors in the ECU, a small 8bit 5V ST72F324 and a large 16bit ST10F280.
others see picture

dmm.jpg

• Coper
• Dec 14, 2023

 

[Anthony489]

Hi

I do not know what the exact fault is, but here is some background information which might help.

The "15" supply is the DIN 72552 reference for an ignition switch controlled supply. The "30" is the reference for a permanent +12 volts from the battery. So I agree with your conclusions about the power supplies on Pins 27 and 28.

It is normal for control units like this to have an internal voltage regulator to drop the 12 volts down to a steady 5.0 volts for the electronics. This regulator will still work OK even with inputs as low as about 7 volts (e.g. during cold start) but if the input voltage drops below 7 for too long the electronics will not work properly. The capacitors will hold up the voltage during a very brief interruption, but not for a longer one.

It is possible that the ignition switch controlled supply is simply used to "turn on" a series FET to allow the permanent supply to power up part of the control circuitry when it is needed. So most of the current demand will be from the permanent "30" supply.

If the control unit is reporting 12.0 volts rather than 12.8, this could be a sign that there is a little bit too much resistance in the supply wiring, for example a fuse/fuse holder which is tarnished. I don't know what the current drain of the control unit is, but it may vary quite a lot depending on what it is doing. If it suddenly tries to draw more current (for example to operate a solenoid) this could cause sudden extra voltage drop in the supply line, and the 12.0 will dip down even further.

I suggest you double check the supply voltages going into the ECU, ideally with an oscilloscope so that you can see any short-term changes or dips. It may be that the reported error is fundamentally being caused by a DC supply voltage issue rather than anything related to the electronics. For completeness, check that the two ground connections are all OK.

 

[MSjoberg]

Thanks for the interesting contributions in this thread. I have not noticed any CAN bus outages or other Comfortmatic problems for my Ducato (fingers crossed), this is just a comment on voltage readings via OBD:

During the last 4-5 years i.e. since the Ducato was new, I have seen what I considered surprisingly low voltage readings most of the occasions I've connected MES to the OBD port. Usually this has been to check data from the Comfortmatic ECU, but sometimes some of the other modules also. The battery is of course under some load since the ignition key is turned on, but anyway.

Some weeks ago I browsed the parameter values from most modules. The ignition was constantly on throughout this operation, and the engine was not started. I've now looked at the log files and found these values:

So the/my question is what conclusions one really can draw from a low voltage reading from the Comfortmatic ECU?

(External temperature was +6 'C so perhaps also the glow plugs cretaed some load on the battery at this occasion.)