Engine Control Unit (ECU) Fault? Diagnose Help.

Hello,

I'm having a problem with the engine control unit (ECU) of my AJM TDI. The engine starts, but then shuts off after one second. Initially, I suspected it was the immobilizer system, but when the key can't be read, the MFA usually displays a "immobilizer active" message. However, I tried using a different key without success.

"The engine control unit cannot be accessed, nor can the instrument cluster. According to the authorized workshop, both are connected on a CAN bus. Other devices such as the comfort control module, ABS/ESP, and airbags can be accessed. If the engine control unit is disconnected, the diagnostic function of the instrument cluster becomes available again."

I first encountered this problem about 2 years ago. Since the device had its chip tuned 8 years ago, I sent it to the tuner, who reprogrammed it, and it worked fine until yesterday. Now, it won't start anymore, exhibiting exactly the behavior described. I've taken it out again and am now considering what to do next. Another question is what else could be causing this.

I assume that reprogramming didn't fix the problem, but it was just a coincidence. I actually suspect that there might be something else in the wiring harness, and that the modification simply re-established a bad connection that has now reappeared due to the cold weather. It was also extremely cold during the last failure two years ago.

In the workshop, a new device costs 900 euros plus tax, installation, and recoding of the immobilizer. I think that's outrageous. I could get used ones for 100-300 euros, but then I would still have to have the immobilizer recoded at the workshop. Now the question is, where exactly is the coding stored? If it were just in an EEPROM, I could replace it. I would also have to replace the ROM from the chip tuning, otherwise it's a waste of time.

Where is it probably stored?

"I've also been thinking about repairing the device, but the soldering actually looks perfect, and the connector and internal components are dry. I'm also missing the pinout diagrams. If I knew which pins connect to the power supply and where the diagnostic line is, I could at least use cold spray and hot air here on the table to search for faulty areas by reading data through the diagnostic line and then see when the device fails."

Where can I find information about the pinout for that connector?

It's also strange that the car worked two more times. It even started when it was parked on the sidewalk overnight, but wouldn't start again in the afternoon. When it did start, it ran without any issues. To me, this almost looks like starting problems. Perhaps a capacitor that can no longer provide enough power during startup? However, what speaks against this is that the device doesn't allow diagnostics even when the engine isn't running. Still, it's strange because the engine runs perfectly for about 1 second...

Does anyone happen to have any useful information for me? (Pinout diagrams, physical location of the encoding, repair instructions).

Before you jump to conclusions: For professional reasons, I have enough knowledge of electronics to work on things, but I lack the specific background information about the details of this particular control unit.

Best regards,
Stephan.

What kind of car do you even have? Give me some details.

Quote:

For professional reasons, I have a good understanding of electronics.

Okay, then unscrew the engine control unit and take a look at how the flash memory is attached to the circuit board (poorly soldered, socketed).

Best regards, Rainer.

A Volkswagen Golf 4 PD-TDI, manufactured in 2000 and first registered in 2001. It has an AJM engine, as I already mentioned.

The flash module is very well soldered. I've actually never seen sockets for SOIC chips, except maybe in some prototyping tools or programming devices.

Outsider wrote:

It's also strange that the car worked two more times. It even started when it was left overnight on the sidewalk, but not again in the afternoon. When it did start, it ran without any issues. To me, this almost looks like starting problems. Perhaps a capacitor that can no longer provide enough power during startup? However, there's something against that because the device doesn't allow diagnostics even when the engine isn't running. Still strange because the engine runs perfectly for about 1 second...

"In my opinion, a defect in the diagnostic/CAN interface of the main system module (MSG) should be considered. The MSG communicates with the immobilizer via this interface. If the interface is faulty, the MSG would start, but then wouldn't be able to communicate with the immobilizer. Since no authorization from the immobilizer could be obtained, the MSG would then shut off the engine again."

The MSG has a 121-pin connector. What I've seen so far:
Pin 6 = CAN-L.
7 = CAN-H
Mass = 4.5
12V = 37.88

Good luck!
Sure.

PS: If you're considering swapping the MSG (Main System Module), the WFS (Vehicle Status Function) can always be programmed. However, if you want to replace the EEPROM and flash memory (chip), you need to make sure you get a control unit with the same suffix. The ones currently being sold by VW might be a newer hardware generation, which could mean your components wouldn't work.

I've also been thinking about the CAN bus. Based on the symptoms – the engine starts and immediately shuts off – it sounds like the permission signal from the instrument cluster isn't being received, which is why it's shutting down. But how can we explain that the diagnostic system isn't working, considering it should be communicating via the K-line?

I'm considering trying to power this device with just 12V to see if I can get it into diagnostic mode. However, I would need to figure out which pin is the diagnostic pin first.

Are all engine control units actually equipped with 121 pins? After all, this is a pump-injector control unit, and there are also versions for the 90/110 hp engines. They can't be compatible with each other...

How does the pairing of the WFS work exactly? Do you have to pair each key individually? I'm currently unable to find my spare key, and if I were to pair a new key now, I might have to pay the pairing fee again later .

K-Line is on pin 16.
For the old container, the WFS should still be routed through K-Line.

Best regards,
Micha

"In Prague, someone once tried to steal a car. The car had been broken into; the ignition switch was ripped out, and the diagnostic connector cover was removed. It seems they attempted to disable the immobilizer unsuccessfully. I read somewhere that cars from the model year 2001 onwards have a third-generation immobilizer system. Perhaps this contradicts the use of the K-line."

Outsider wrote:

... The control unit cannot be accessed, and neither can the instrument cluster. According to the authorized workshop, both are connected on the same CAN bus. ...

Either you've misunderstood something, or I would urgently recommend changing the workshop!
Because neither the MSTG (Mirror System Top View) nor the KI (Instrument Cluster) are connected to the CAN bus in a G4 with firmware version 01.
However, it's important to note that the AI (Artificial Intelligence) unit is located upstream of the MSTG (Modular System Technology Gateway) in the signal path. If the AI unit is defective, the MSTG cannot be accessed for diagnostics. Therefore, replacing the MSTG should be the last resort when troubleshooting this issue. The first step is to get the AI unit working again. Otherwise, the MSTG will also remain inaccessible.

The AI system routes the K-Line diagnostic data from the WFS chip (which is integrated into the AI) to the MSTG. Attempts to tamper with the WFS can damage this chip within the AI, resulting in the error mentioned above. I have seen this happen multiple times.
As a temporary workaround, I bridged the K-line from the instrument cluster (KI) to the main system control unit (MSTG). This allowed me to at least access the MSTG again, but it made the instrument cluster and thus the windshield wiper system (WFS) inaccessible. It also became impossible to reset the inspection interval via the diagnostic interface.


Adjust the AI and modify the WFS accordingly.

Okay, so if I measure the voltage between pins 5 and 6, I'm getting a reading on both the control unit and the wiring harness. So, I assume there's already a CAN bus present. What it's doing is another question entirely.

Why do you think the instrument cluster is faulty? That was also the first thought the workshop had back then. But when they disconnected the MSTG (Motor System Test Gateway), the instrument cluster diagnosis came back online. It's the same situation now. It looks very much like it's the MSTG, or why wouldn't it be?

If the AI (Artificial Intelligence unit) is actually defective, and you were able to make the MSTG (Main Service Terminal Gateway) accessible again by bridging the connection, then I would need to connect the MSTG directly to my desk with a K-line connection and a 12V power supply in order to access its diagnostics, correct?

Okay, here's something else that comes to mind: A neighbor of mine has a brand new instrument cluster (IC) with zero kilometers on it. It's not exactly the same one that fits in my car; it has these silver rings around the instruments. But if it's fundamentally similar, I should at least be able to determine if the IC is faulty by swapping it in and then trying to access the diagnostic system of the MSTG (Main System Test Group) again. If that works, it's probably really the IC. Then I would just need to have the IC programmed/adapted.

A Volkswagen Golf 4 PD-TDI, manufactured in 2000, first registered in 2001. It has an AJM engine, as I already mentioned.

The flash component is very well soldered. I've actually never seen sockets for SOIC chips, except in some prototyping tools or programming devices.

The vehicle should still be running on software version WFS II. I would recommend performing a battery reset and temporarily disconnecting any aftermarket radio to see if that resolves the issue, then try the diagnostic process again.

Base:

http://www.chip-service.de/product_info.php?info=p190_PSOP44%20Sockel%20Yamaichi.html
http://www.chip-service.de/product_info.php?info=p191_PSOP44%20Sockel%20Meritec.html

and you encounter them relatively frequently .

I might try testing the external radio. But after it worked for 7 years, I'm not sure why it should suddenly be a problem. Unless it has recently developed a defect. I specifically bought the correct adapter for VW, so the diagnostic cable connection should have been taken into account.

I'm not familiar with this type of socket, even though I work in the electronics industry. I only know about the extremely expensive Textool sockets used in programmers and similar devices. Always something new to learn.

Outsider wrote:

... So, I'm assuming there's already a CAN bus present. ...

Again... (for the last time!) ... This is not a CAN bus!
Neither the engine control unit (KI) nor the transmission control module (MSTG) communicates via the CAN bus. They use the K-line instead (see Wiesel's answer). This is a huge difference!

Quote:

If the AI is actually defective, and you were able to make the MSTG accessible again by bridging the line, then I would need to connect the MSTG to my desk with the K-line connected and a 12V power supply in order to access the diagnostics, correct?

Okay. This is how you can test it.

Quote:

Why do you think the AI is defective?

Because I've had exactly those kinds of mistakes multiple times before...
Here too, the MSTG was initially (pointlessly) replaced, and then the KI was replaced for even more money. After both were new, everything finally worked again.

Whether both were defective at the same time, I cannot say.
In the three cases I personally dealt with in my own garage, it was "only" the Keyless Infiltration (KI) system that was defective. Incidentally, in all of these cases, there had been prior attempts to tamper with the Window Film System (WFS) through break-ins.

For clarification. The diagnosis first goes through the AI system (specifically, the IMMO chip) and then to the MSTG. Why do you think that is?

If I remember correctly, the AI signal enters pin 25 of connector St1 (T32), and pin 5 of connector St2 (T32a) outputs the signal to pin 16 of the MSTG, which is the K-line.

If it were actually the AI module and I needed to replace it, I would have to take it to a Volkswagen workshop after the repair to have it adjusted. However, I would still like to know *what* needs to be replaced beforehand.

The current issue is that when the MSTG (Multi-Station Terminal) is connected, the diagnostic system for the KI (likely referring to a control unit or similar device) doesn't work, but it does work without the MSTG. If I install a new KI now, of course, nothing would start, but theoretically, the diagnostics should then work again, and I could be relatively certain that I have found the problem, right?

I might be able to temporarily test a different AI system; it's usually quite simple and quick. However, would that cause any problems where something gets altered, and the test AI stops working in the other car, or my own AI stops working in my car?