Ladedruckeinbruch um 1500 rpm - wieso? | Posts 16+

Devilseye503 wrote:

[If this is due to the learning values, then it should be possible to temporarily disable it by deleting the learning values...by resetting the main working parameter (MWB) to its default setting.] oder sehe ich da was falsch?

Probably yes, because the default settings here have nothing to do with data deletion.

I experimented with something – see the attachment.

Considering the offset between the pre-tax CO2 emissions and the logged air-fuel ratio, it seems to me that there might already be an indication of a systematic miscalibration (at least in my car during this driving cycle).
Even if there is still a slight decrease in LDR (Loan-to-Deposit ratio), the overall result looks better with a lower offset (due to a change in the pre-tax interest rate), especially since the opacity no longer falls under the rotation speed.

I haven't been able to observe any drawbacks in other situations (e.g., excessive oversteering).

Hi Ulf,

pass auf das nicht bald eine ganze Meute Headhunter im Auftrag von Frau W bei dir vor der Tür stehen, wenn du weiterhin die Software im Alleingang verbesserst.

"A great comparison, and now it also oscillates around the target value when paired with a suitable TV."

Hi,
I've had little time lately, but this thread couldn't have come at a better moment.
My not-so-new 4b AKN used to exhibit this effect as well, although it was more noticeable between 1600 and 1800 meters when climbing in third gear, which was my test route.
I initially suspected that the boost pressure control system still had a differential component, but it seems to be much simpler than that.
It's difficult to work with logs due to the sampling rate; I ended up focusing more on LDA instead.
Since I'm not skilled in chip tuning, I managed to somewhat compensate for the problem by shortening the VTG rod. At the time, I thought I was the only one experiencing this issue...
Frage ist natürlich, wieso es irgendwann, in meinem Fall hat der Wagen gut 150tkm weg, dazu kommt. Ist das Umfeld schuld, Düsenlöcher und Kat teilweise zu, erhöhte Spaltmaße zwischen Laufrädern und Gehäuse durch Abrasion speziell beim V6 mit seinem Rost-Krümmer, oder bekommt die VTG-Mechanik Spiel? Wenn die Schaufeln so gelagert sind, daß die Gaskräfte kein nennenswertes Moment auf die Verstellung erzeugen, hätte das eine Hysteres zur Folge, mit der die Regelung Probleme bekommt. Regardless of the fact that it may be cheap but not necessarily effective to simply disable a sensitive system like the VTG (and also the AGR in most vehicles) instead of adjusting its range using a secondary controller to achieve the necessary pressure regulation. To regulate value.

christians wrote:

Mein mittlerweile nicht mehr ganz so neuer 4b AKN kannte diesen Effekt nämlich auch, allerdings eher zwischen 1600 und 1800 bergauf im dritten, meine Teststrecke halt.

Hi Christian,

In my car, the actual value of a certain parameter (indicated as "LD" in the third gear) lags significantly behind the expected value, reaching its target only beyond 2000 RPM.
Daher ist ein Einbruch auf meiner LDA nur bei langsamen LD-Anstiegen in hohen Gängen zu erkennen, und ich kann Dein Problem nicht so recht einordnen . . .

Quote:

habe ich das Problem durch Verkürzen der VTG-Stange einigermaßen austariert.

That would probably have worked for me too, but unfortunately at the cost of higher peak pressures during overshoots, which I actually didn't want to increase any further, given that my layer height was already increased due to the tuning.
My logs were created using the standard configuration to avoid any potential interference from tuning-related issues in the discussion.

As long as your "overswings" with the shorter VTG shaft remain within acceptable limits, I consider your problem to be solved.
(@Devilseye503: I would also recommend this approach as your first attempt.)

Quote:

Does the VTG (Variable Geometry Turbine) mechanism experience play? If the blades are positioned in such a way that the gas forces do not generate a significant moment on the adjustment, this would result in hysteresis, which could cause problems for the control system.

Probably, "bafische" is located exactly there with his ALH. He described that in another thread.
Unfortunately, as far as I know, it is practically impossible to distinguish between minor mechanical clamping and hysteresis in the control behavior of system A.

Quote:

Mal ganz davon ab, daß es zwar billlig aber nicht unbedingt sinnvoll ist, ein störempfindliches System wie die VTG (und auch die AGR bei den meisten Fahrzeugen) nur zu stellen statt den Verstellweg über einen untergeordneten Regler auf den für die eigentliche Druckrelgelung nötigen Wert zu regeln.

Well, in the Pmax range (where most chargers have very little headroom left, so precise operation is needed to avoid overloading and air starvation), the control system seems quite usable, as long as the mechanics are working properly.

You wouldn't be able to eliminate hysteresis caused by play in the loader, even with a rod height monitoring system.

ulf wrote:

bei meinem Wagen hinkt der LD (im 3. Gang) dem Soll so stark nach, daß es erst jenseits 2000 rpm erreicht wird.

"The same thing happened with mine. With the old settings, after a long period where the pressure was low, when it approached the target value above 1600 RPM, the VTG (Variable Turbine Geometry) would open, causing the pressure to drop initially, but then build up correctly. However, due to the test conditions, this occurred above 2000 RPM."
Since I've mostly been driving in the city and on heavily trafficked country roads lately, I haven't had a chance to fully assess the behavior of the vehicle in the higher gears.

*excavate*

Since I don't have anything more important to do on the car right now, I tried adjusting some additional parameters of the load dump control system (P, I, and D gains, characteristic curves, etc.).
For a high time resolution, I only logged data from channel MWB 11 here.
The LD drop at 770 seconds is already significantly smaller in the initial attempt, so the distance between the onset of turbulence and the rotational speed above 1400 rpm (where the motor exits the turbulent zone) should be greater than what would be achieved with just the modified pre-injection control.
However, the dwell time fluctuates more significantly up and down compared to the standard settings, which could potentially accelerate wear on the VTG mechanism due to unnecessary movements.
If there's still room for improvement (which I suspect there is), I find myself wondering again why the LDR software is configured in such a peculiar way as shown in the diagrams above, and where the corresponding benefits lie...
The remaining regulatory behavior has so far (under full load) shown no disadvantages such as increased overshoots or oscillations, see the right half of the appendix.

Devilseye503 wrote:

I recently had to deal with the exact same problem with my ARL - ... The actual value reaches approximately 1500 RPM and achieves its target pressure - it doesn't go above that - and then drops significantly... (everything is the same as with you, but more pronounced)

. I can now understand that if I enter all the LDR parameters from the (newer) ARL into my software without changing the pre-tax KF.

I recently had to deal with exactly this problem with my ARL as well - ...
The actual value is reached at approximately 1500 RPM, reaching its target pressure - it doesn't go any higher - and then drops significantly. ... (everything is the same as with you, but more pronounced)I can now understand that if I enter all the LDR parameters from the (newer) ARL into my software without changing the pre-tax KF.

Am I understanding you correctly - that all we would need to do is change the pre-tax VAT setting, and it would be perfect? Also, I now understand - the original software is terrible? Why would they do something like that? I thought VW had some idea of what they were doing?

Devilseye503 wrote:

Am I understanding you correctly - all that would need to be changed is the pre-tax KF value and it would be perfect?

Yes and no: by adjusting the pre-tax KF, you can change the average value of the LD oscillations (meaning the LD will oscillate symmetrically around the target curve, rather than being mostly above and far below), but the magnitude of the drop in the rpm range remains the same.
Smoothing out the oscillations in the load disturbance (LD) response, which means reducing the depth of the dips, can only be achieved by adjusting the proportional (P), integral (I), and derivative (D) gains of the control system. However, these parameters are not currently defined in the pre-tax configuration file.

Quote:

Furthermore, I just realized - the original software is garbage? Why would they do that? I thought VW had some idea of what they were doing?

"Is garbage" is too general: it's more of a compromise between expensive development time and usable features under common everyday conditions. Exotic driving profiles like "full throttle in high gear from 1100 rpm" are only considered marginally, and suboptimal control behavior is accepted.