Jerky Acceleration: Troubleshooting & Fixes

Hello,
My Passat, manufactured in 2002, with a PD engine, 96kW, AVF code, and 60,000 km on the odometer, has been experiencing the following problem in the last week:
Beim Hochbschleunigen bockte der Motor sehr heftig, als wenn Zündaussetzer stattfinden würden. Beim langsamen Steigern der Drehzahl bis ca. 2200 U/min lief er gleichmäßig. Plötzliches Gasgeben gab für ca. 0,5-1 s den gewohnten Antritt, dann setzte das Ruckeln ein und der Motor ließ sich in diesem unregelmäßigen Betrieb bis max. 3200/min hochbringen, was ich aber nur auf einer Testfahrt bis zu diesem Ausmaß nur 3x provoziert habe, um den Motor in diesem Ruckelbetrieb nicht zu beschädigen. Dabei ist die Steuerung 1x in den Notbetrieb gegangen (Fehler 17964, Unterschreitung Luftmenge). Von da an lief er wie ein Saugdiesel, zwar ruckelfrei, aber mit ca. 30% Leistung, bis zum Fehlerreset. Danach wieder das Ruckeln wie vorher. Das Logprotokoll zeigte, das der Ladedruck auf max.ca. 2400 mbar ausgeregelt wurde, die Luftmasse/Hub und der Saugrohrdruck jedoch sehr zackenartig verlief ( ca. 200mbar Diff. von Spitze-Spitze bei einem mittl. Level von ca. 2300mbar). Die Sollwerte im Leerlauf wurden lt. Tabelle AVF eingehalten, bis auf das Signal des Tastverhältnis AGR-Ventil, das zwar meist im 60%-Bereich war, jedoch auch sporadisch (bis zu 30s lang) auf den 4,8%-Wert (wie mit einem Rechtecksignal) fiel. To the best of my knowledge, this should only occur when accelerating or under higher load.
My question is: Is it possible that the mass airflow sensor (MAF) could occasionally experience intermittent failures, causing the air mass measurement to abruptly drop, which in turn would reduce the fuel injection amount? But could this trigger such jerky reactions?
I consider problems with the loader or its adjustment to be unlikely, as it seems to be operating normally and the adjustment mechanism is working.
"As of today, the engine is running normally again, but I don't know why. I'm expecting that the problems could reappear at any time. I had similar symptoms once in the winter after refueling, but they went away after about 200 km. I initially suspected water in the fuel, air in the system, or a clogged filter, but I'm ruling those out in this case."
Should one directly log the output signal of the mass flow meter using a multimeter to check for interruptions (which is complex), or is there a more straightforward way to check for this issue elsewhere?
I hope I haven't written too much, and I would appreciate any suggestions for further review or refinement.

Best regards,
Ralf

Given these severe misfires, I suspect a loose connection in the wiring. I would primarily examine the wiring for the mass airflow sensor (MAF) for corroded contacts and breaks. Alternatively, there could be issues with the diesel fuel supply, but you seem to be ruling that out (Why?).


Of course, there are also the usual 'common issues': Relay 109 (requiring resoldering) or NHG (hammer test).

Regards,
Private.

I had a similar problem with my 1Z engine. The car ran perfectly fine for normal driving. However, as soon as I reached higher RPMs (around 3100 and above), the engine would sputter and smoke quite a bit }.

Since I didn't have a complete understanding of the turbo's pneumatic system, I went to a Bosch diesel specialist workshop, and there we measured the boost pressure while driving (error code: intake manifold pressure difference). It was definitely not okay (I think it was too low) and it wasn't adjusting at all. That was the reason why the mixture was no longer correct at higher RPMs.

Then we looked for the cause.
- Turbo defective?
- Is the hose damaged or leaking?
...

Then we bypassed the magnetic valve (directly in front of the diesel filter) using hose connectors (for pressure hoses), and the engine ran smoothly (turbocharger uncontrolled). So, we installed a new magnetic valve, and so far, everything is quiet. See also: /viewtopic.php?t=3431&highlight=magnetventil+1z

Sure, here's the translation of the text from German to English:

'Cu müllo' is a Swiss German expression that means 'goodbye' or 'see you later.' It is often used in informal settings, such as when talking to friends or family.

Um... excuse me...

guys, we're talking about a PD. I wouldn't know if it has a needle injection system.

I would also be inclined to bet on a somewhat uncertain outcome, somewhere between LMM and STG.

Guys, we're talking about a PD. I don't know if it has a needle injector.
Ouch! When you once again make assumptions about other people based on their cars and don't bother to find out which car it is...
The NHG error is therefore ruled out... Thanks, Jörgs, for pointing that out!

I would also cautiously bet on something unstable, somewhere between lmm and stg.
At least the part of my answer that made sense...

Regards,
Private.

Hello.
First of all, thank you for your feedback.
Naturally, identifying the cause is more difficult because the error has not occurred in the past few days.
However, generally speaking, I would say that causes related to the fuel supply are unlikely for the following reasons:
-A filter/pipe is not temporarily blocked for 2-3 days, allowing only a minimal amount of flow, and then it clears up again.
-It was the same tank of fuel that initially caused the problems, and then the problems disappeared.
Leaks are also excluded for the aforementioned reasons.

Based on my experience with my previous diesel vehicles (Mercedes-Benz models from 1957, 1970, and 1976, all with inline injection pumps, purely mechanical), I also know that fuel shortages always manifest first as a slowly developing loss of power within 3 to 20 seconds. It's possible that this manifests more severely in an electronically controlled common-rail engine, especially if the fuel supply to the injector system is very tight. However, I find it hard to believe that this could cause such abrupt misfires. It would be interesting to know if anyone else has experienced this behavior related to fuel supply problems. I didn't want to run the tank completely empty for testing purposes.
"However, I can certainly imagine that the engine control unit could easily block (or significantly reduce) fuel injection signals from one cylinder to another. Furthermore, even with implausible input signals (e.g., mass airflow sensor, throttle position, etc.), this should ideally only be achieved with a more sophisticated ramp function programmed into the ECU. Ultimately, engine speed, turbocharger speed, and the amount of air entering the combustion chamber cannot change so drastically within just a few milliseconds."
In conclusion, the issue might be located on the signal input side (most likely the mass airflow sensor) if the engine control unit (ECU) is reacting so drastically. Based on the observed behavior, we must assume that the control signals for the injectors and fuel quantity are being transmitted correctly, otherwise the error would not consistently occur in the same operating conditions.
Could the measurement of the injection pressure (if any sensory data is being collected) also cause the injection to be reduced in such a way that, in the event of a suspected pressure overload due to a faulty measurement, the injection is reduced?

Thank you very much, and best regards.
Ralf.

The target values at idle, as specified in the AVF table, were met, with the exception of the signal from the EGR valve's duty cycle, which was mostly in the 60% range, but also occasionally (for up to 30 seconds) dropped to the 4.8% value (like a square wave). To my knowledge, this should only occur during acceleration or under higher load.
My question is: Is it conceivable that the LMM (Lambda Measuring Module) can occasionally have glitches, causing the air mass measurement to abruptly drop, which in turn would reduce the fuel injection amount? But could this trigger such jerky reactions?

Yes!

'Load pressure and air mass are directly related, and this relationship is reflected in the curves when the values are plotted in a table.'
If the mass airflow sensor (MAF) is currently defective or has a loose connection, it can cause severe jolts. Similarly, defects in the turbocharger pressure control or sensor system can also cause these issues.

The AGR doesn't really make any significant jumps, except when it's running at full load.
4.8% indicates a fully closed state, but this also happens in idle mode after 3 minutes.
Therefore, you must rule that out as a fundamental cause.

First, check the VTG (Variable Turbine Geometry) adjustment. You mentioned that the engine was running reasonably well without turbocharger assistance (in emergency mode).

Also, also check the EGR valve as well, as it can also become clogged and stuck.
Also, sometimes it's a good idea to take it offline and observe.

Then, check all the vacuum hoses. Disconnect the hoses from the control valves and apply vacuum to them, observing the corresponding components (turbocharger and EGR valve). Do they hold vacuum? Hairline cracks can be very problematic.
(See also my technical article on the basic settings for actuator testing and standardization.)

Check all cables and connectors. Are the control valves working properly?

CU Gremlin.

Gremlin wrote:

The AGR (Abgasrückführung) doesn't really make any significant jumps, except when you're running at full load. 4.8% means fully closed...

Hi Ralf,

Are you sure?

AFAIK, VAG has a special "logic" regarding the keying ratios: namely, small numbers indicate a high effective current, and vice versa.
Therefore, 4.8% likely indicates "AGR fully open"...?

Hello,

Thank you for the information. In particular, thank you to Gremlin for pointing out that the EGR valve control can jump to 4.8% during idle. I had previously considered this to be the clearest indication of a loose connection.
Alternatively, I have considered what relationship might exist between driving situations and error behavior if we assume a mechanically linked loading system.
During the inspection, I moved the adjustment rod, which may have freed it up. I can't say for sure whether it was stuck beforehand (force required is approximately 25 N in the axial direction with the hose disconnected).
Prior to the problem, the car had been driven for approximately 600 km with very low power output in order to test the minimum fuel consumption (shifting gears very early, with the engine speed most often between 1000-1400 RPM, and this over several days/trips). During this period, the average fuel consumption (on longer trips) decreased from approximately 6.5-7.5 to as low as 4.5 liters per 100 km. A similar driving style might have been employed in the winter due to slippery road conditions.
Can the bucket adjustment mechanism on the loader become stuck within a few days if it is practically never adjusted (within a few days) for a higher exhaust flow due to higher engine speeds?
This would mean that, even with the most economical driving style, one would occasionally need to engage a more powerful acceleration phase to keep the mechanical adjustment system working.
Furthermore, according to this theory, the angle of attack of the exhaust flow on the impeller blades would, with this hypothetical blockage of the guide vanes (due to soot or dirt), result in such a poor efficiency at higher exhaust flow rates/engine speeds that the boost pressure would collapse on the other side.
Furthermore, the problem could only be explained if a temporary boost pressure is generated, but it collapses very quickly because there is no corresponding power from the turbocharger driving it, or because the performance curve breaks down at a certain point. Consequently, the boost pressure decreases, the air-side power demand drops significantly, and the turbocharger's exhaust side can return to its previous operating state, etc.
Given that this process occurs in relatively short cycles, I fear that this behavior cannot be measured using the sampling rate during data logging.
But it would just be a theory.
If I can't find anything else, I might consider installing a microswitch on the adjustment rod that can be activated in the appropriate driving position.

Best regards,
Ralf

Hi,

Quote:

Can the bucket adjustment mechanism on the loader become stuck within a few days if it is practically never adjusted (within a few days) for a higher exhaust flow due to higher engine speeds? This would mean that, even with the most economical driving style, one would occasionally need to engage a more powerful acceleration phase to keep the mechanical adjustment system working.

I don't think it's likely to happen within a few days.

The approach has merit. For example, I drive my 10 km commute to work at a very low speed, and I have a spot just before my destination where the engine is sufficiently warm. If traffic allows, I try to accelerate from idle in second gear with full throttle to move the variable turbo geometry (VTG) through its entire range at least once per trip.

Best regards,

Jan.

The AGR (Abgasrückführung) doesn't really make any significant jumps, except when running at full load.
4.8% means fully closed...
Hi Ralf,

Are you sure?

AFAIK, VAG has a special 'logic' regarding the keying ratios: namely, small numbers indicate a high effective current, and vice versa.
Therefore, 4.8% would more likely mean 'AGR fully open' . . .?


no...

The valves are connected to +12V, and the ground connection is switched.
This results in the valve being fully open when the duty cycle is 100%.
At 0% duty cycle, the device is permanently closed.
is initially confusing .

Take a look at the log files from ALH .


CU Gremlin.

Gremlin wrote:

The valves are connected to +12V, and the ground connection is switched. This results in the valve being fully open when the duty cycle is 100%. At 0% duty cycle, the device is permanently closed. is initially confusing . Please check the log files for ALH

.
Please provide the text you would like me to translate from German to English.

Hm, that's strange... Your log with air mass + boost pressure:
When the planned output (1744) is significantly higher than the actual output (1244), a "low speed" setting (24.7%) is used to quickly increase the pressure.
0.8 seconds later: Target = 1948; Actual = 2050; Key press accuracy = 58.2%.

AGR: At low boost pressure, the AGR valve may be fully open -> measured value 84.5%.
At high boost pressure -> EGR on -> 4.8%.

So it seems you are right about the AGR, but I am right about the boost pressure.