Boost pressure drop at 1500 rpm - why?

Hello,

"With my turbocharged vehicle, I repeatedly observe dips in the boost pressure during full-throttle acceleration at low RPMs in higher gears."
Sure, please provide the text you would like me to translate from German to English. I will only provide the translation without any explanations.

Around 1500-1600 rpm, this almost always happens, but sometimes it also occurs at higher speeds, up to about 2000 rpm. At those higher speeds, the dips are not as pronounced as in the image.

I thought the charger was malfunctioning.
But then, as far as I know, it would only start building pressure again after the fuel injection volume has decreased significantly, along with the turbocharger speed, allowing the compressor to "engage" instead of "over-revving."
The image shows that the drop in line pressure during the load dump event only slightly affects the rotational speed, so the motor torque hardly decreases. The pop-o-meter indicates virtually unchanged acceleration during the sag, and I also hear no unusual (pump) noises.

In some cases of sagging, the cloudiness remains consistently above the rotating part, while in others, the motor sinks deeper into the cloudiness.

The dips in performance occur alongside the serial and tuning capabilities, even without exceeding the specified limits for the lambda sensor (LD), which suggests that overly aggressive control interventions can be ruled out.
Contrary to that, the trend of the key pressure (TV) indicates that the Electronic Control Unit (ECU) is attempting to compensate for the error (low TV = VTG remains closed = increased pressure build-up).

Who can make sense of all this?
It's not really a serious problem, more of a minor flaw, but perhaps there's an explanation for it.

Aber Ulf, so eine Frage von dir.....

Das sind doch die üblichen Unter-Überschwinger vom Einregelvorgang, die je nach Angebot der Abgasenergie (Drehzahl/Menge/Temperaturen zum Zeitpunkt des Erreichens des Sollwertes) und Güte der TV-Vorsteuerung (Toleranzlage deines Ladesystems) mehr oder weniger stark ausfallen. Bei niedrigen Drehzahlen ist das besonders kritisch für die Regelung, da bei so kleinen Drehzahlen im Normalfall nicht mit dem Erreichen des Sollwertes gerechnet wird (Höchster Gang-bergauf - da erreicht man den Solldruck bzw. höchste Ladedrücke bei niedrigster Drehzahl weil viel Zeit für den Ladedruck/Temperaturaufbau bleibt). BeBelow 1400 RPM, the LD (likely referring to a specific component or system) is controlled but not regulated. This results in excessively high torque values in the pre-control stage, which hinders the settling process.badly.

Best regards,

bafische wrote:

These are the usual sub-routines of the registration process,

Hi Bavarian,

For me, "undershoot" always means that the target value has been exceeded or fallen below.
In this log/image, it's only briefly touched (Expected Value = Actual Value = 1980 MB), and already is the EDC "overreacting"?

Why not also at 1800 rpm? But there, it is "calmly" accepted that the actual value is greater than the target value .

Quote:

Bei Drehzahlen unter 1400U wird der LD gesteuert und nicht geregelt, daher stehen aus Dynamikgründen viel zu hohe TV in der Vorsteuerung, die die Einregelung verschlechtern.

I also see that on the TV tuner display. But my power dips occur later, around ~1500 rpm.

Why not also at 1800 rpm? But there, it is 'calmly' accepted that the actual value is greater than the target value .



I don't quite understand this sentence - your actual value is still below the target value, even though it's already at 1800 RPM.

But it would be worth investigating this phenomenon.

I recently encountered a similar problem with my car, although it's much more pronounced. However, when this happens, I don't experience any boost overruns, but the engine has significantly less power in that range.
Unfortunately, I don't have a graph to include right now, but here's a description similar to yours: The actual value reaches approximately 1500 RPM and achieves its target pressure, without exceeding it, then drops significantly. Like a triangle-shaped curve, the load (LD) reaches its target value up to 2200 RPM, but not beyond. This was observed in 6th gear at 100% load. However, this is also somewhat strange – when the LD target is reached, there are no significant overshoots, but rather an optimal settling behavior (everything like yours, but more pronounced).
greetings

Devilseye503 wrote:

Your current speed is 1800 RPM, which is still below the target value.

^^}
However:
"Is" = dark blue, "Should be" = light blue.

To the right of the red arrow, a section shows that the setpoint is exceeding the actual value (approximately 2200 millibars).
At the same time, further down, the green RPM curve runs within the range of 1750-1900.

The graphic was created in Excel (only data series can be logged) and then completed using an image editing program.

ulf wrote:

For me, "undershoot" always means that the target value has been either exceeded or not reached. In diesem Log / Bild wird er aber nur mal eben berührt (Soll = Ist = 1980 mb), und schon wird durch die EDC "überreagiert"?

Generally, yes, but specifically not in this case. The controller always only adds to the pre-set value; that is, the controller output is zero when the target value is reached (although there are exceptions). Therefore, the value in the pre-set then becomes relevant. This can lead to overshooting or undershooting by a smaller or larger amount than the target value.
The tuning of boost pressure regulators is extremely complex because each load and engine speed point would require its own optimal regulator configuration, as the system being controlled is not constant. However, this is not feasible in practice, so the design is always a compromise, and in certain areas, the control cannot achieve the optimum performance.
Ein weiterer Punkt ist, das die Regelung eine Art Gedächtnis hat. Sie merkt sich Abweichungen vom Idealverhalten und korrigiert sich selber (Das meinte ich oben mit Ausnahmen). Je nach Vorgeschichte wird sich das Regelverhalten ändern, d.h. die Regelung kann auch gelernte Werte wieder vergessen oder anders gewichten. Das kannst du messen, indem du aus unterschiedlichen Lastzuständen (lange Schubphase oder konstante hohe bzw. niedrige Last, danach Fuss aufs Pedal bei immer der gleichen Drehzahl/Gang) beschleunigst. Das Regelverhalten wird sich jedesmal ändern. Vor allem wenn du vorher in einem Regelbereich fährst und die AGR "defekt" ist. D"Then, 'as' realizes the situation and thinks to themselves: 'Oh dear, something is definitely wrong here,' and completely loses it."If you accelerate after that, the control system will be disrupted for a certain period of time.

Hello Ulf,
I can only provide answers in English.
The pressure drop before or during the achievement of the target charging pressure is due to the pre-regulation of the duty cycle.
As mentioned previously, the VTG/VNT valve is opened before the target value is reached to prevent overshooting. An overshoot limit of 150 mbar is permissible.
Depending on the application, overshoots may be desirable for improved transient behavior.
With your PQ24, which is relatively light, you don't need the overswing feature, as your "butt meter" proves in practice.

bafische wrote:

The behavior of the system will change each time. Especially if you were previously driving within a certain operating range and the EGR valve is "malfunctioning".

... oder die VTG temporär klemmt (Fachartikel) ...


Ulf has already provided the explanation himself.

"Quote:"
"The fluctuations occur with the serial and tuning performance, even without the LD (lambda) value exceeding its limit, which means that excessive control interventions can be ruled out, in my opinion."
On the contrary, the trend of the key pressure (TV) shows that the EDC is trying to compensate for the error (low TV = VTG remains closed = greater pressure build-up)."

When do such contradictory control interventions occur? When the feedback loop is interrupted somewhere. In the case of a load-dependent (LD) control system, this is often caused by a stuck or jammed variable transmission gear (VTG).

The effect that Ulf observed will continue to intensify and may eventually lead to a failure at some point in the future.

Otherwise, this drop in voltage can be observed in all VTG chargers (see explanation from "bavische"). However, new vehicles typically show a much smaller drop. But Ulf's car should be like new in every other respect .

schneirk wrote:

der Abfall im ist-Ladedruck vor bzw. beim Erreichen des Sollladedrucks liegt an der Vorsteuerrung des Tastverhältnisses.

Yes, that could be the key.

With the values obtained from the pre-tax engine control unit (ECU), the TV logarithm values would be higher, and therefore the load demand (LD) would be even lower; in other words, the controller already adjusts the LD upwards to above 1500 rpm.

If the target of approximately 1500 is reached, the regulator seems to relax and thinks "okay, done," loosening its corrective measures. As a result, the LD (likely referring to a specific system or process) crashes because it starts operating more like the pre-tax KF (likely another system or process) again.
Okay, so the regulator needs to be adjusted again, which will correct the TV picture and pull the laserdisc player back up.
-> A regular oscillation due to an improperly adjusted offset...?

@bavarian:
The behavior (within this driving cycle) is practically the same with a functioning and a defective EGR system.

Thank you for your answers .

Bertil wrote:

When do such contradictory rule interventions occur? When the control loop is interrupted somewhere. In the case of a load-discharge (LD) regulation system, this is often due to a jammed variable transformer gear (VTG).

Hi Bertil,

Although my car has been showing this low fuel consumption reading at higher speeds since it was new (currently around 18,000 km), and it hasn't changed significantly, I just performed a fuel injector test as a precaution.
Result: The VTG (Vertical Translation and Guidance system) can be controlled with extreme precision in all directions within the hub, allowing for smooth and minute adjustments.

It doesn't chatter on the closed end like your old spool did, and I've never been able to observe excessive overspin on the LDA.

Das Ganze sieht also bis jetzt nicht hiernach aus:

Quote:

Der von Ulf beobachtete Effekt wird sich immer weiter Verstärken und irgendwann viel später möglicherweise zum Ausfall führen.

As I said, I'm now even more inclined to believe there's a factory-installed misconfiguration between the pre-tax vehicle configuration (likely referring to software settings) and the hardware, at least for this particular driving profile.
Let's see if I can figure out how to manage this without the over-revving causing excessive wheelspin when accelerating quickly in lower gears.

research now ulf

Is the entire issue explainable by the larger drop due to the bigger loader?

Devilseye503 wrote:

Is the entire issue explainable by the larger drop caused by the bigger loader?

As long as the loader isn't pumping, then no.
If, for example, it already achieves 1 bar at 1500 rpm and full load, there's no reason why it should only achieve 0.8 bar at 1700 rpm - as long as the variable geometry turbocharger (VGT) control is working correctly and the fuel injection quantity hasn't been reduced.

I have managed to reduce the described oscillation around 1600 rpm to the point where the LDA no longer fluctuates erratically (I haven't had a chance to log the data yet).
Currently, there's still a slight dip detectable around 1800 rpm, but that should probably also be fixable...

According to the LDA (likely referring to a data logging or diagnostic system), the acceleration behavior in 3rd gear with full throttle from approximately 1200 rpm has not changed. Tests in 1st and 2nd gear are still pending.

ulf wrote:

... although my car has been showing this LD drop at high gears since new (currently around ~18,000 km) and it hasn't changed significantly...

Okay, then your charger is already showing that characteristic to a pronounced degree. However, I haven't encountered many vehicles of that age with such issues.
However, it could be within the manufacturing tolerances. If your loader reacts a bit slower than others, you might experience the same effect.

Quote:

It doesn't chatter like your old spool did, and I've never been able to observe excessive over-extrusion on the LDA.

You don't necessarily have to experience extreme mood swings.
My container didn't leak; instead, it was completely sealed. This wasn't noticeable either with the syringe or by hand. Only when it was held under full vacuum for a few seconds could this effect be observed.
By the way, I now know of three more PD TDIs that had this same problem. They were all manufactured around the same time period.
Based on your description, it doesn't seem to be due to other mechanical issues.

Quote:

Wie gesagt, ich gehe (jetzt noch mehr) von einer werksmäßigen Fehlanpassung zwischen Vorsteuer-KF und Hardware aus, zumindest für dieses Fahrprofil.

No, I don't believe it was a misconfiguration.
Your charger is right at the upper tolerance limit, and it seems to be a bit more sluggish than usual, which is why your load dump regulation system is trying to intervene so strongly. I think you'll have to live with that. Observe these dips and overshoots; you'll see that they will become larger (sooner or later... probably around 50,000 km).

Is the entire issue explainable by the larger drop due to the larger loader? As long as the loader isn't pumping, then no.
If, for example, it already achieves 1 bar at 1500 rpm and full load, there's no reason why it should only achieve 0.8 bar at 1700 rpm – as long as the variable geometry turbocharger (VGT) control is working correctly and the fuel injection quantity hasn't been reduced.

I have managed to reduce the described fluctuation around 1600 rpm to the point where the LDA no longer rises erratically (I haven't had a chance to log it yet).
Currently, there's a slight dip detectable around 1800 rpm, but that should probably also be fixable...

.

I checked it today – it reaches approximately 2 bar of absolute pressure at around 1500 RPM. However, the pressure doesn't increase further; it stagnates. At speeds around 2200 RPM, the pressure rises again abruptly without overshooting the target pressure. The fuel injection volume is definitely not being reduced.

What did you change to reduce the hole?


However, that's not always the case.
Sometimes the engine runs like this, with large overshoots – in this case, with significantly more pressure up to 2000 RPM (see image: /files/6tergang.jpg).
what I already showed in the other thread.

'...as can be seen here, the target pressure is reached at 1500 - however, in one case it exceeds the target, while in the other case, there is a drop.'

Devilseye503 wrote:

bei 1500 wird der solldruck erreicht - nur in dem fall übersteigt er das soll - im anderen fall erfolgt ein einbruch.

This could be due to the most recently saved learning values (see bafische's answer) or to random mechanical jams in the VTG unit, the pressure sensor, and/or the solenoid valve.

Quote:

was hast du geändert um das loch zu verringern?

The input tax field.
My post referred to a different route than my "Log House Mountain." There, you experience another slight drop in power starting at around 1500 rpm .

However, I still believe that this little problem can be solved using the pre-tax input tax credit mechanism. I will make a few more attempts in the next few days.

bei 1500 wird der solldruck erreicht - nur in dem fall übersteigt er das soll - im anderen fall erfolgt ein einbruch.
This could be due to the most recently saved learning values (see bafische's answer) or to random mechanical jams in the VTG unit, the pressure sensor, and/or the solenoid valve.

.

If this is due to the learning values, then it should be possible to temporarily fix it by resetting the learning values to their default settings by setting the main working parameter (MWP) to 0.

or am I seeing something wrong?