Kniffelaufgabe: Ursachen für unrunden PD-BUK-Kaltstart? | Posts 48+

Okay, today I finally had some time to remove the valve cover.

Everything looked fine underneath; the PDEs are positioned parallel in the central zone as they should be.
All the camshafts and lifters appear to be in new condition, as far as can be seen without disassembly.

However, the PDE ball joints were probably somewhat poorly adjusted from the factory.
Regulation states:
In the UT (likely referring to a specific device or mechanism), turn the adjustment screw all the way in, and then rotate it back 180°. On my engine, the factory setting only involved rotating it back approximately 130-160°.
I've now set all the PD gains to 180 degrees... let's see how that affects the vehicle's handling.


Interestingly, the fasteners used surprised me, because the screw heads are marked with "12.9," which, to my knowledge, is a strength grade that is supposedly exclusively used in the ARL models according to official documentation.
BPX and BUK should have the same ZKS values as, for example, the ASZ -> 10.9 .
Apparently, you can't always rely on documents completely.

The action with the transparent hose is still on the ToDo list .

Quote:

I have now set all PDEs to 180°. Let's see how that affects the handling.

Okay, then it will just work a little worse now.

Rainer K. wrote:

Quote: I have now set all PDEs to 180°. Let's see how that affects the handling. Okay, then it will just perform a little worse now

. Wow... have you already gathered any clear experiences/measurements on this topic?

"It's actually quite logical, Ulf: the support ramp will connect to the PD part later on, allowing for more space between them as it did before."

teileklaus wrote:

It's actually quite logical, Ulf: the support ramp will be connected to the PD part later, when there is more space between them like before.

No, there is no play in the PDE actuation because the PDE spring presses all moving parts together with zero clearance.
The adjustment moves the working range of the pump piston within the positive displacement element (PDE) up or down.
In an engine, this would be roughly equivalent to a connecting rod with adjustable length.


EDIT:
The oil standard issue continues in https://community.dieselschrauber.org/en/viewtopic.php?p=160274#160274.

ulf wrote:

... Apparently, you can't always rely on documents completely. ...

Ulf... Put 100% on it, about 70-80%... Then it will be right .

The ET lists are sometimes completely wrong. If I were to equip my AXR with actual ET documents, it would simply not run anymore.
(without further comment)

Okay, today I logged another cold start – but this time including data from the luminosity sensor, hopefully to pinpoint the cause of that poor compression likely due to depleted hydraulic accumulators.
Compounding the issue of the 6-day runtime was my yesterday's pressure drop equipment (PDE) adjustment check, during which I incrementally rotated the motor past two full revolutions to ensure each PDE was within the specified range.
"In this process, each hydraulic lifter was completely compressed without oil pressure between the valve stem and the valve spring within the valve unit, using maximum force, which may further exacerbate issues with collapsed lifters compared to other cold starts. In the log, a lack of compression should be evident in weak knock sensor readings immediately after starting, which should then increase as the hydraulic lifters fill with oil."

The white line doesn't show anything meaningful in this regard. Okay, it's the sum curve of all cylinders, so a weak compression in one cylinder wouldn't be as noticeable in the overall LuMa sum if there are assumed to be 3 "good" cylinders, compared to looking at the data for just that single cylinder alone.
What do you think? Shouldn't a lack of air in at least one cylinder, which (along with a significantly increased LRR allocation) would certainly lead to bluish-white smoke, be noticeable as at least an increasing trend in the total LuMa values?
Alternatively, can a poor compression ratio be directly identified immediately after starting the engine by examining the lambda curve?

I can't imagine that an LMM (Linear Motor Module) would react so quickly; the air column alone introduces a significant amount of inertia into the signal.

Otherwise, you would have to be able to detect every intake stroke in the air mass. Not only that, but the amount of air entering would also need to be discernible if you want to diagnose it like that.

While I don't have any experience in this area, my logic tells me a clear "no"; it's not possible to determine that in this way.

Hello everyone!

I completely agree with what Benni said.

To further illustrate this with the following example:
Let's assume that cylinder 3 experiences approximately 20% compression loss or draws 20% less air mass in the first 7 seconds. However, this results in "only" a 5% reduction in the total measured air mass across all 4 cylinders, which will likely be masked by the EGR control system anyway.

It remains peculiar that the LLR (left-linkage rod) of cylinder 3 continues to misbehave, as it slips out of alignment again even after adjustment, especially when a short burst of gas is applied. This is noteworthy because it's happening in contrast to the LLR of cylinder 4.

Ulf, even though you might not like it due to the risk of something getting stuck: In my opinion, the only thing that will help is swapping a seemingly normal cylinder with number 3 (e.g., PDE 1 <-> 3). Otherwise, if I may be so bold, you won't really figure out what's causing this phenomenon.

Best regards from Munich!

Benni wrote:

I can't imagine that an LMM would react so quickly; the air column alone will introduce significant inertia into the signal. Otherwise, one would have to recognize every intake stroke in the air mass.

According to Bosch, the HFM (Hot-Film Mass Air Flow sensors) of PD-TDIs are fast enough to even capture back pressures from the oscillating air columns before the closing intake valves in their signal voltage.

Quote:

Let's assume that cylinder 3 experiences approximately 20% compression loss in the first 7 seconds, or takes in 20% less air mass. However, this results in "only" a 5% reduction in the total measured air mass (LMM) across all 4 cylinders, which would anyway be masked by the EGR control system.

The EGR is actually inactive for the first ~5 seconds (at least) after engine start. During this time, you can observe the air mass without the influence of the EGR - unless the ECU activates the EGR significantly earlier -> I need to log this again with certainty using a diagnostic tool (MWB3).

Your assumed 5% reduction in total luminous flux across all cylinders would translate to a variation of approximately 20 mg/H, which should be visible in the graph as roughly 1/10 of the distance between 300 and 500 - appearing as an overlaid, slow fluctuation on the (typical) luminous flux noise pattern.
However, if even a compression loss of 10% or less in a cylinder is sufficient to produce the aforementioned symptoms, it would indeed be difficult to detect this in the diagram .

ulf wrote:

According to Bosch, the HFM sensors for PD-TDIs are fast enough to even capture back pressure waves from the oscillating air columns before the closing intake valves in their signal voltage.

And you also mean that the data coming from the diagnostic interface are real data from the mass airflow sensor (MAF)? ... So, it seems like Goettmann already "pulled your tooth" on that...

goettmann wrote:

.... You won't be able to record these values using a standard diagnostic interface. To do so, you need real-time data that hasn't been manipulated or altered through the diagnostic interface. ...

Bertil wrote:

ulf wrote: According to Bosch, the HFM sensors of the PD-TDIs are fast enough to even capture back pressure waves from the oscillating air columns before the closing intake valves in their signal voltage. And you also mean that the data coming from the diagnostic interface are real data from the mass airflow sensor (MAF)? ..... So, it seems Goettmann already pulled your tooth...

That's clear. I just wanted to respond to the idea of a "completely useless" MAF.

Quote:

I can't imagine that an LMM would react so quickly; the air column alone will introduce significant inertia into the signal.

Observing the fluctuations in the LuMa line, one can see which "d-Luma/dt" rates the entire EDC15 system plus the diagnostic tool is still capable of representing.

ulf wrote:

... According to Bosch, the HFM sensors of the PD-TDIs are fast enough to even capture back pressure waves from the oscillating air columns before the closing intake valves in their signal voltage. ...

Ähm... Where does it say that?
I actually know the interpretation you're describing quite differently.
The back pressure, which inevitably distorts the mass flow meter's measurement result, is compensated for using both mechanical means (e.g., a grid placed before the mass flow meter, flow straighteners) and electronic means (e.g., a low-pass filter applied to the mass flow meter signal internally before it is sent to the motor control unit).
It's completely new to me that Bosch is supposedly using mass airflow sensors (MAF) to analyze pulsations in the airflow.
Alternatively, my idea is to use an oscilloscope to directly measure the voltage waveform at the LMM output, in order to minimize aliasing distortion. However, for the reasons mentioned above, I doubt whether this will actually make a difference.

After carefully reviewing your diagram again, it's clear that the air mass is actually higher than average immediately after takeoff. This suggests either that the compression is working correctly or that the measurement method is unreliable (so there's something for everyone to consider *wink*).
Did you administer the two gas bursts shown there slowly or abruptly? Why does the air mass hardly change ?

Long story short (as usual): In this case, it's more likely to be a problem with the ZMS (Zylinderkopf Multiventil System) based on your previous description, rather than a fixed issue with cylinder 3's PDE (Pneumatic Direct Injection) system.
If we could simply exchange PDEs and many, many questions would suddenly become self-explanatory in the logical thought process.

Best regards from Munich.

DieselBär30x wrote:

Did you apply the two gas boosts shown there gradually or abruptly?

I played with the pedal a little bit (approximately once per second). The second time, I pressed it down a bit further, but still not all the way to the floor.

Quote:

Why is the air mass changing so little

Because hardly any lift drive has been built up, resulting in a lack of load?

Quote:

If we were to swap the PDEs, many questions would suddenly become clear through logical reasoning.

It might be, but as long as the practical problems remain small, I'm not willing to take on the effort and risk (like using a threaded collar!).

DieselBär30x wrote:

ulf wrote: ... According to Bosch, the HFM sensors of the PD-TDIs are fast enough to even capture back pressure waves from the oscillating air columns before the closing intake valves in their signal voltage. ... Um... Where does that say?? ...

Please provide the text you would like me to translate from German to English. I will then give you only the translation, without any explanations.
...not because I want to annoy you, but because it personally interests me .

DieselBär30x wrote:

DieselBär30x wrote: ulf wrote: ... According to Bosch, the HFM sensors of the PD-TDIs are fast enough to even capture back pressure pulses from the oscillating air columns before the closing intake valves in their signal voltage. ... Um... Where does that say?? ... Please provide the text you would like me to translate from German to English. I will then give you only the translation, without any explanations. ...not because I want to annoy you, but because I am personally interested

. I obtained this information from documents whose further distribution/publication would likely be problematic.
There might be some sources that are not critical, but I can't think of any at the moment.

Regarding the response time: For newer Bosch mass airflow sensors, the time it takes for their output signal to reach 63% of a change in flow is less than 10 milliseconds.