Kniffelaufgabe: Ursachen für unrunden PD-BUK-Kaltstart? | Posts 16+

By the way, the AJM mentioned in the signature doesn't do that either.


What kind of oil do you use in the "Bärli" bus?


The Sharan, which also starts like this, had a complete new cylinder head installed with 4 new injectors about 20,000 km ago (not for that reason). It already started cold like this before. This is just to exclude any wear-related issues.

I actually meant to add this, but... Bärli is getting old *lol*.

Also, also Castrol 5W40 without the WIV additive.

EDIT: VW50501...

You too? Hmm... someone has to put in 50601 just to rule it out. I'm absolutely not a fan of this type of oil. You probably feel the same way, otherwise you wouldn't have it in your car.

ulf wrote:

... What do you think about it? Still within the normal range for a water temperature of -3°C?

"Ideally, no car meeting the EU4 emissions standard should produce smoke during a cold start. Especially not with only 25,000 kilometers on the odometer. At 125,000 kilometers, I would have considered it (barely) acceptable."

As long as the car starts and smokes like that, I'm sticking with my valve/lifter theory.
Brief blue smoke -> likely oil, probably from the turbo -> that's okay... then black smoke -> combustion with too much fuel or a lack of air.

Steffen W wrote:

This is nothing to complain about.

Well... it's quite difficult. Definitely borderline. At the very least, it's too much for the mileage.

Quote:

Maybe we should try 50601 0W30 in very cold weather sometime?

I'm just going to start my AXR now. Seven hours at -12°C (left outside in the carport) should be enough.

"I think it's a mistake to blame the oil."
Ließe sich aber 2 fach ausschließen. Ölwechsel oder Kompr. Test.
According to the lifter theory, a cylinder should show lower compression when cold if there's an issue with the lifters.
More likely, the "air vent theory" – with an attempt to cross-exchange the partial differential equation (PDE) – but please note that I have no idea about that part; it's just a thought experiment.

Steffen W wrote:

Someone has to add 50601 just to rule it out. I absolutely don't like this type of oil. You will probably think the same way, otherwise you would have it in your car.

Exactly .

Okay, I don't quite understand the oil-related theory behind the explanation that the valves aren't opening properly.
If the oil in the hydraulic lifters is too thin, they will drain more quickly when the engine is not running (the faster the draining). Consequently, on the next start, there won't be enough oil cushion in the valve actuation path (camshaft to valve), causing the valve to open later and less far, and close earlier: this is acceptable to a certain extent.
But my 5W40 oil has only been in the engine for about 3,000 km (on average, ~20 km per trip). Should it already be significantly thinner than a 5W30 oil would become over its entire service life (assuming that "higher" oil standards don't cause such problems)?

And why don't you hear the typical ticking sound caused by excessive valve clearance in PD engines anymore? What kind of noise insulation do they have that, for example, the ALH doesn't have?

teileklaus wrote:

According to the lifter theory, a cylinder should show low compression when cold.

However, this is not necessarily the case: while the air throttle may interfere with ignition at idle speed, the amount of air entering at the significantly lower starter motor speed may still be sufficient to produce seemingly normal compression values. . . ?

Hello everyone again!

Since you're all talking about the hydros, I want to mention for completeness that I had all of my hydros completely rebuilt by the original manufacturer because of a previously existing, but hard-to-locate, rattling noise coming from one of the cylinders.

Result: Absolutely no improvement regarding cold starts - and the rest as well .

"...however, in my case, the MWB13 values during a cold start do not deviate (significantly)..."

Best regards from Munich!

Hey, in reverse logic, excessively high oil pressure (which can be caused by a stuck oil pressure control valve, for example) can contribute to the valves staying open for too long. In this case, the oil pressure is greater than the return force of the valve springs.
War jetzt aber nur ein Beispiel von einem mech. defekten Motor. Ich habe damals statt des 15w40 dann 5w40 in den Golf reingeschüttet, damit bin ich mit der Möhre noch übern Winter gekommen. WaIt's a 16-valve engine.

The knocking sound at the beginning of your video is a bit strange for the first few seconds. But regarding the smoke, I don't find anything unusual; it seems to be just condensation. You need to change the name of your video, otherwise people won't be able to find it through the YouTube search.

My lathe starts up normally. Immediately after starting, the spindle speed drops slightly by about 50-100 RPM, but then it stabilizes and runs consistently.
The outdoor temperatures are really quite extreme now. This morning, it was -18°C at my place. I thought my whistle would blow, and Mother's power steering pump in her Ibiza too .

ulf wrote:

.... And why doesn't one hear the typical ticking sound caused by excessive valve clearance anymore in PD engines?

Why are winter tires significantly quieter and yet offer better grip today compared to what they were like 25 years ago?
Why don't you ask the manufacturer about it?

The issue with hydraulic components occurs to varying degrees in all cars (which is why fewer and fewer hydraulic components are being used).
Basically, it's not a big deal, and you would just have to accept it. Then there would be nothing to complain about.
Anyway, you're using an oil that is outside the manufacturer's specifications for your car. I don't want to claim here whether this only has effects during the cold season. It's conceivable (but I personally don't believe it ).

However, I wouldn't rule out the other possibilities!
But you can forget about water -> the consumption is too low.
Air -> It would sputter without any significant clouds (due to excess air).
Hanging PD needle -> poor injection pattern -> absolutely conceivable... But the error shouldn't "wander" around.

A proper compression test (both cold and hot) would provide clarity.


My AXR started up cleanly (almost) without any stuttering. After 2 seconds, everything was quiet. There was definitely a lack of oil. And that's at 186,000 km .

To further complicate things, here's a diagram:
Cold start in the garage, but now PDE 3 (red) is running normally again and reaching its usual level.
After that, I repeatedly rode up the small incline around the corner (stages 1-5) to observe how the LRR reacted as the temperature increased.
And suddenly, between phase 3 and 4 of the load cycle, the problem that the amount of PDE 3 was always so high seems to disappear .

Does anyone have any other (new) ideas on this?

I also find it remarkable how the correction factor values at idle (green line in the range of 840 rpm) change: Before phase 5, the lines for each cylinder are constantly being reordered, but only above 60°C water temperature does the picture remain largely consistent... I'm starting to believe there's a pseudo-random number generator at play within the BUK software.

Bertil wrote:

Air -> it would sputter without significant clouds (excess air).

Hmm, are you sure about that?
What do you think about the following "air-in-PDE -> smoke" theory?

Here's the translation:

"Normal operation: When the PDE (Pressure Driven Emission) solenoid valve closes, the high pressure builds up, and virtually simultaneously, the nozzle needle opens -> injection begins at the desired time."

However, if there is air in the high-pressure chamber, it will initially be compressed to a significant extent until the nozzle opening pressure is reached – and this occurs considerably later than when there is no air in the high-pressure chamber.
Consequently, the injection starts late, possibly even so late that proper combustion cannot occur -> resulting in weak or completely absent ignition, and smoke.

ulf wrote:

teileklaus wrote: According to the valve lifter theory, a cylinder should show low compression when cold. However, this is not necessarily the case: while the air throttle may cause misfires at idle speed, the amount of air entering at the significantly lower starter motor speed may still be sufficient to produce seemingly normal compression values, albeit subtly...

If you spend too much time "fiddling around," then definitely. It's possible that it won't be measurable then.

I once sent your video to my engine specialist friend.
He means: Everything is fine. Such "problems" can occur in this cold weather. There's nothing to complain about.

ulf wrote:

Does anyone have any other (new) ideas on this?

Why does everything calm down here so late?
It definitely speaks against air, water, and hydro power. That should have been completed long ago.

Quote:

... slowly, I'm starting to believe there's a Linear Congruential Random Number generator in the BUK software

.

Yes... one might say.

ulf wrote:

... I also find the behavior of the correction factors at idle (green line in the range of 840 rpm) remarkable: Before phase 5, which represents the load phase, the lines for each individual cylinder are constantly being re-mixed,...

Yes, it's me! *meld*
"...constantly being re-mixed..." actually makes me think more of ZMS, doesn't it?

Here's the translation:

"I'll repeat the test now, after another 20 hours, with similar outside temperatures. I'll let the engine idle until it reaches 60°C and see if the issue is temperature-dependent or time-dependent."

Best regards from Munich!

Hmm, it seems my edit just overlapped with your last two posts.

Bertil wrote:

I once sent your video to my engine specialist. He means -> Everything is okay. Such "malfunctions" can occur in this cold weather. There's nothing to complain about.

So, it's "state of the art" despite EU4, and therefore I can stop worrying about demanding a warranty claim... ?

Quote:

A control test would be helpful here, now after another 20 hours, with a similar outside temperature. Let the engine idle until it warms up to 60°C and see if it's temperature- or time-dependent.

It's more likely that the tank is empty than that the sputtering engine, completely unloaded at -10°C ambient temperature, would reach a water temperature of 60°C.

ulf wrote:

What do you think about the following "air-in-PDE -> smoke" theory?

Absolutely nothing...

Quote:

However, if there is air in the high-pressure chamber, it will initially be compressed to such an extent that the nozzle opening pressure is eventually reached – and this occurs significantly later than without air in the pressurized environment. Consequently, the injection starts late, possibly even so late that proper combustion cannot occur -> resulting in weak or completely absent ignition, and smoke.

However, you would see that in abnormal MMB 23 values (see Steffen W.'s post). But you didn't mention those.

Quote:

So, "state of the art" despite EU4, and therefore I can reduce the mental stress regarding warranty claims...

I tend to see it the same way you do.
But if a motor expert whom I know very well assures me that this kind of starting behavior could already occur without anything being defective, then I would also believe that.

Bertil wrote:

Quote: However, if there is air in the high-pressure chamber, it will initially be compressed to such an extent that the nozzle opening pressure is eventually reached – and this occurs significantly later than without air in the pressurized environment. Accordingly, the injection starts late, possibly even so late that proper combustion cannot occur -> weak or completely absent ignition, smoke. You would have to see that in abnormal MWB 23 values (see Steffen W.'s post).

However, you didn't mention those. Yes, in the initial post,

Quote:

It should be possible to detect air in the PDE in the MWB 23 data, which I also logged on several occasions.
Unfortunately, the MSG only starts providing real data (not equal to zero) approximately 20 seconds after the start, by which time the engine is already running almost smoothly. However, at that point, no clear outliers can be detected: in 3 logs from different days, all cylinders are within the range of -30 to zero.
I suspect I know your answer:
"But even 20 seconds after the start and later, the LRR values are still sometimes far outside the normal range!"

My answer:
"You're right. However, the MWB 23 only shows air in the area directly around the solenoid valve needle where there is strong cooling airflow. If the needle is dislodged, it's possible that residual air could still be present further down, near the pump piston and nozzle seat, delaying injection until all the air has been expelled."
If the remaining air volume becomes small enough during its expulsion that the actual brake specific impulse (BSIP) reaches a level sufficient to re-ignite the injected fuel, then the engine will run smoothly from that point onwards. However, an additional volume of air must still be pumped in per cycle to compensate for electromagnetic losses caused by the air cushions and the slower build-up of injection pressure (leading to poorer atomization!) – until all the air has been expelled from the pre-combustion chamber. This would then manifest as a slow convergence of the LRR data series of the respective pre-combustion chamber towards the pendulum value (i.e., without air), while the main combustion chamber (MWB 23) already provides completely unremarkable data.
IMO, my cold start log data basically fit seamlessly into this theory .
The question is whether a time scale of approximately 20 seconds at idle speed (about 150 injection cycles!) is actually needed to completely vent a pressure drop element (PDE) in the area between the pump plunger and the nozzle seat

Quote:

. Quote: So, "state of the art" despite EU4, and therefore I can reduce the mental stress regarding warranty claims... I tend to see it the same way you do. But if a motor expert whom I know very well assures me that this kind of start could happen even without anything being defective, then I would believe that.

Okay, from that, I can clearly determine my course of action.

Thank you in advance for your helpful contributions here }.

In any case, I will remove the valve cover at the next opportunity and take a closer look at everything (including checking and adjusting the valve clearances), as much as possible without disassembling any further parts.

Can a loose pressure equalization device (PDE) be detected by vigorously shaking the magnet valve knob (assuming the PDE isn't already visibly crooked in the cylinder head)?