AVB zeitweise LuMa-Mangel (Lösung: VTG schließt zu weit) | Posts 32+

Okay, today I removed the AGR valve and bypass flap components again, definitely confirming that there's no temporary restriction caused by the flap. All air/turbo hoses and the plastic pipe were also removed and checked for any constrictions.

LMM voltage measured:

2 and mass 12 V --> OK
4 and 3, 5V --> OK.
5 and Mass: 1V when ignition is on --> OK.
5 and mass: 2.2V at idle speed --> OK (slightly higher than normal due to the EGR valve being closed).
"Voltage: 5V, Current: 3.9V at approximately 4000 RPM when stationary --> should also be okay."

Then I measured the vacuum directly at the sensor. At idle, with the throttle valve (TV) at 25%, I got a reading of 640 mbar, and the variable geometry turbocharger (VTG) was fully engaged. With the default setting, at 0% TV, the reading was 770 mbar. These values should also be okay.

Then I took it for another test drive, provoked the error again in second gear, and continued driving with full throttle. This time, at 33 seconds (in the log), it suddenly started working on its own without any input from me, and then returned to normal.

If I'm "happy" tomorrow (holiday), I will apply 12V to pin 2 and try it again. Where is the best place to get 5V? Are they available on another pin at the control unit?

Quote:

2 and mass 12 V --> OK 4 and 3 5V --> OK 5 volts and 1 amp when ignition is on --> OK. 5 and mass 2.2V at idle speed --> OK (slightly higher than normal due to the EGR valve being closed). "Voltage is 5V and current is 3.9V at approximately 4000 RPM when stationary – this should also be okay." Then I measured the vacuum directly at the can. At idle, with the throttle valve (TV) at 25%, it was 640 mbar, and the VTG (variable turbine geometry) was fully engaged. With the basic setting, at 0% TV, it was 770 mbar. These values should also be OK.

It would be insightful to see how the values behave during operation when the error occurs.

The vacuum level remained directly proportional to the duty cycle when the error occurred, and was approximately 550 millibars.

I'll also measure the voltage while driving next time; I realized today that it was already too late.

chli1976 wrote:

Then I took it for another test drive, and provoked the error by using the second gear, while continuing with full throttle. This time, at 33 seconds (in the log), it suddenly started working on its own without any gas pedal input, and then stopped again.

The log confirms my suspicion of a problem in the area of the mass airflow sensor (MAF) - which also includes the 5V power supply from the engine control unit (ECU).

Quote:

Where is the best place to tap into the 5V line? Are they available on another pin at the control unit?

IMO no.
It's better to build a small circuit board using, for example, a 7805 fixed voltage regulator, and include a small electrolytic capacitor (e.g., 100µF) between the input/output and ground on each side.
Connect the positive wire to the input of the circuit board (from pin 2 of the MAF sensor, or preferably pin 1 or 2 of the engine control unit), connect the output to pin 4 of the MAF sensor (by cutting the wire leading to the engine control unit), and connect the regulator ground to pin 3 of the MAF sensor.
Make sure the circuit board is well-insulated to prevent short circuits with vehicle components, and place it in a location within the engine compartment that doesn't get too hot.

Okay, I've now measured the supply voltages simultaneously.

Both will remain active if he doesn't leave, and they are based on the correct values. Pin 4: 4.5-5V, and Pin 2: battery voltage.
Just thinking about it now, I took the power from the glove compartment light. Should I also tap into pin 3 for that?

Hab auch die Signalspannung gemessen, wenn er funktioniert bis zu 4,2 V. If not 3.4V.

I also need to run it at idle speed for a certain amount of time; if I don't run it long enough at idle speed, it won't work.
You don't have to completely stop accelerating; reducing the throttle by about 30-40% is usually sufficient.

chli1976 wrote:

Both will remain at the correct values if he doesn't go, with pin 4 at ...5V and pin 2 measuring battery voltage. What just occurred to me is that I took the power supply from the glove box light, should I also tap into pin 3?

Actually, yes, but the signal voltage data clearly shows where the problem originates - namely between the supply voltages and the signal output:

Quote:

I've also measured the signal voltage; when it works, it goes up to 4.2 V. When it doesn't work, it's 3.4 V.

There's some kind of random generator there that's pulling down the signal voltage.
It doesn't necessarily have to be an electrical fault; it could also be a moving foreign object near the mass flow sensor that deflects the measuring airflow away from the sensor or restricts the flow in the measurement channel.

"Ulf, your assumption makes sense, but I can't find any errors in this area, so I've tried something else."

Tax curve slightly raised in the critical area.
The smoke value map has been completely set to 48 mg per injection. Therefore, your assumption that the signal voltage is being reduced should be incorrect, because even if it's taken into account, the fuel injection amount isn't being reduced.

The intake hose up to the turbo was removed and driven without it.

Same symptoms. And when it happens, it starts smoking excessively at a certain RPM, so I must be getting too little air into the cylinders.

I also blocked the VTG strut at the upper stop, and for some reason, it doesn't build up any load pressure there, and instead goes into emergency mode because the boost pressure regulation is exceeded.

I also installed a different software version, specifically version 1234 instead of 1263, but it still exhibits the same problem.

Okay, that pretty much confirms that the mass airflow sensor data is accurate, which means the air throttling must be happening through a hardware issue.

If you can rule out problems with the intake valve and issues related to the AgA system, another possibility is excessive oil pressure (if the short-circuit valve in the oil pump doesn't open). In this case, the hydraulic lifters could become overfilled, preventing the valves from closing properly.
If this happens with the intake valves, the compression blows back into the intake manifold, and the Lambda values decrease significantly.
The fact that it then produces excessive black smoke when running at full engine speed (if the compression is still sufficient for ignition) is obvious.

So, revert the software to its OEM settings, connect an oil pressure gauge up to approximately 15 bar, and try to see if there's a correlation between oil pressure and power loss .

I checked the oil pressure a few days ago, and it's okay; it regulates at 5.5 bar.

Yesterday, I also tried something else. I loosened the set screw for the VTG linkage by one full turn. The result was the same. You're not supposed to adjust it normally, but I couldn't think of any other solutions.

Today I reversed the process. I turned the adjustment screw two rotations further in from its original position, and now it seems to be working. So far, I haven't been able to reproduce the problem. I'll then take a test run and set the log back to its normal value. Hopefully, it will work now.

chli1976 wrote:

I measured the oil pressure a few days ago, and it's okay; it regulates at 5.5 bar.

Even while the error is occurring?

Quote:

Today I reversed the whole process, turned the set screw from its original position two rotations further in, and now it seems to be working. So far, I haven't been able to reproduce the problem. I'll then take a logging trip and put the log away. Hopefully, it will work now.

I'm curious to see how the LD (likely referring to a specific system or parameter) regulation behaves.

chli1976 wrote:

ulf wrote: chli1976 wrote: I measured the oil pressure a few days ago, and it's okay; it regulates at 5.5 bar. Even while the error is occurring? Yes, even if the error occurs

. That's mean; all my spectacular error theories are being ruined .

I was just thinking about a closed clamping VTG that constricts the gas flow on the AgA side to such an extent that the LuMa values collapse. But before that, the actual LD value would still have to significantly exceed the target value .

The way he operates makes it impossible to provoke him further.

I tightened the set screws of the VTG linkage by two rotations.

Thank you for your help, especially to Ulf.

Here's the log.

chli1976 wrote:

Since he functions in that way, he can no longer provoke it. I tightened the set screws of the VTG linkage by two rotations. Thank you for your help, especially to Ulf. Here is the log

.
It sounds like a refurbished charger from a professional.

Yes, I know that you're not normally supposed to tamper with things there. The charger was already replaced by someone else and it was a refurbished unit.

Thank you for the solution and an interesting story!
Now we know what log data looks like when the VTG closes too much.

What surprises me the most is that mistakes can be made by anyone, regardless of their expertise; therefore, I have definitely learned something.

To make it easier to find the thread later through search, I'm going to edit the title .