2012 CAAB Limp Mode [Resolved]

[Limpy]

Hi, having a big problem with my van.

Issue:
Limp mode, van will not go over 3000rpm.
At 3000 rpm engine raddles and shakes with diesel knocking & dies down to idle or shuts off, completely losing power.
Light Blue smoke from exhaust.
No fault codes

Starts up quick & nice idle.
Drives normal around 2500 rpm. Still smoke though.

With VCDS I did a measurement & value comparison at idle with my van and a same engine working van, these are the values that stood out with the most difference:


AirMass Specified value (mg/stroke)
Mine Working Van
452.4 358.8

Injection Quantity Deviation (mg/stroke)
Mine Working Van
Cylinder 1. 0.06 0.30
Cylinder 2. -1.59 -0.10
Cylinder 3. 1.83 -0.15
Cylinder 4. -0.35 -016

Fuel High Pressure Control Deviation (hPa)
Mine Working Van
3500 0

So appearantly mine is way off.
Can this be high pressure pump timing related?

It has been done ALOT regarding this issue.
Here is the worklist:

Diagnostic with VCDS, no fault codes.
New injectors coded with VCDS
New fuel rail pressure sensor
New egr & egr cooler
New turbo & actuator
New throttle body
New intercooler
New diesel & diesel filter
New MAF & rear Lambda Sensor

Cylinder compression around 300/min each.

I have tested vacuum hoses & the actuators & sensors.

Done DPF & EGR delete & removed particle filter & blanked off EGR on the exhaust side.

It seems that the engine is stuck in limp mode with no fault codes & with blue smoke.

When taking off the oil filler cap there is a little bit of smoke coming out along with a little bit of air pressure.

The PVC oil breather seems to be working fine and looked clean without any carbon clogging.

The only thing left I can think of is the exhaust cylinder rings that are comprimised or maybe some bad valve seals or worn cylinder head but shouldn't the compression test reveal this?
And shouldn't the car run normal but only with some blue smoke?

Timing belt has been done too, can it be the high pressure pump that has been wrongly tuned in?
Does this pump need to be timed correctly along with the timing belt?

 

[mmi]

Please post engine's blockmap-file - engine ticking over:

Ignition ON

VCDS > Applications > Controller Channel Map >



Tick/fill boxes as above and hit Go.

VCDS flashes screens for a moment (when finished hit "Done, Go Back") and there will be file blockmap-01-... in folder C:\Ross-Tech\VCDS\Logs\....

 

[Limpy]

Tried this but got no file in the logs folder, do I need a login or a security code?

 

[mmi]

Tried this but got no file in the logs folder, do I need a login or a security code?

No login code is needed. Most likely the VCDS-interface cable is of older generation. It's only the newer (since 2017) interface cables HEX-V2 and HEX-NET which have the capability.

The map would have been a good starting point because it's contains engine's all available measurement blocks (=sensors mostly).

Limp mode, van will not go over 3000rpm.
At 3000 rpm engine raddles and shakes with diesel knocking & dies down to idle or shuts off, completely losing power.

I would start by monitoring what engine speed sensors (crankshaft sensor, possibly also camshaft sensor) report at that moment. Also fuel pressure would be interesting, intake air flow, EGR position, charge pressure, etc. Please have a look into thread below for measurement values to look into

204 DSG intermittent momentary power loss

Hi all, Desperately looking for some assistance! I have a 2018 204 DSG high line. Had it from 8k miles, now has 82k on the clock. I’ve done all of the servicing and maintenance on it as I’m very mechanically minded and my business is automotive electrical design. However, I have a fault that I...

www.t6forum.com

 

[Limpy]

Thank you for helping me out with great info.

I will absolutely check all that in vcds.
I have a working van exactly same engine I can measure the values up to.

I've already tried some of it as described in the post, I will come back with the result.

Thanks!

 

[Limpy]

I have uploaded adpmap and auto scan of my van.

I also uploaded different graphs, some are compared to a working van.

Picture explanation:

My van has no plates, I have nowhere to drive it but when I had temporary plates on it, it drove normal up to aprox 3500 then it cuts and died.

(In some of the tests I bypass the 2500rpm idle limit by rolling down a road in neutral while I step on the gas)

Test 1 picture 1:
My van at top, Engine rev up to a bit over 3000rpm then cuts & dies.
Working van bottom, Engine rev up to 4000rpm then back to idle.

Test 2 picture 2:
My van from idle to as much revs I can get, cuts and dies.

Test 3 picture 3:
Parked, 3 identical tests my van only.
Slowly engaging accelerator pedal up to 2500 (idle limit).
Holding it for some seconds then releasing pedal and back to idle.
I noticed the Main Injection Duration of Activation did a spike every time I released the pedal at the end on every test. Yellow Line.

Test 4 picture 4:
my van only, testing max revs 2 times.
first time it cuts down to idle then back to idle limit holding in the gas pedal.
second test it just dies when reaching a tad over 3000 rpm.

As said I have a working van, same engine available.
just tell me what I shall compare or do.

Thank you, Leon

 

[Limpy]

Adpmap or blockmap would only work when I ticked off "Adaptations.. gave me only 20 lines of info. I don`t know why.

I have more comparisons,

Picture 1 is my van at the top, 2500rpm & approximately 4000rpm vs working van at the bottom.
picture 2 is compression test my van at top.

 

[Limpy]

For me and my limited knowledge, The Air Mass Actual Value is sticking out as the one with the largest deviation.
At 4000 rpm mine is at 86,41 g/s whilst the working van is reading 44,63 g/s.

So mine has risen from 2500rpm and 40,83 g/s to 86,41 g/s at 4000 rpm whilst the working van went down from 52,27 g/s to 44.63 g/s.

 

[mmi]

Adpmap or blockmap would only work when I ticked off "Adaptations.. gave me only 20 lines of info. I don`t know why.

That's because the interface cable does not support the functionality. Anyways, the 20 lines is correct for the adaptations.

For me and my limited knowledge, The Air Mass Actual Value is sticking out as the one with the largest deviation.
At 4000 rpm mine is at 86,41 g/s whilst the working van is reading 44,63 g/s.

So mine has risen from 2500rpm and 40,83 g/s to 86,41 g/s at 4000 rpm whilst the working van went down from 52,27 g/s to 44.63 g/s.

Well, for this we don't know EGR position - as it can reduce air intake.

---

I'm sorry but the pictures are too blurry. Could you please post the actual log files (CSV-files VCDS created)

 

[Limpy]

I have done egr delete and blocked it at the exhaust side.

I worked alot with these, too bad they're to blurry. I will have to redo it tomorrow.

 

[Limpy]

1 test parked and throttle up to 2500rpm, the other two van rolling to bypass parked max rpm, it revs to over 3000 then cuts and dies.

 

[mmi]

Well, unfortunately the time resolution is quite poor with your cable. So very difficult to guess what happened between time stamps 10.09 seconds (full revs) and 11.92 seconds (fuel pressure burst). Looks like the ECU is trying to inject more fuel to compensate decreasing revs???? Remains the question why???




To get better time resolution please record only the three values below - to try to see if the ECU cuts off injection or tries to inject more fuel to keep engine running as well to see the fuel pressure behaviour.

IDE00021 Engine speed​

IDE00352 Main injection: duration of activation​

IDE00589 Fuel pressure​

 

[Limpy]

I appreciate you taking a look at this.

Here are one with the turbo option and one without, I don`t know what that does.

 

[mmi]

Thanks, much better now. The turbo option seems to improve sampling rate by approx. 60% --> see dots (data points) more densely on curves below in plot #2 (of attachment 2).





Could it be that the engine is starved of fuel??
At high revs the ECU starts to build up fuel pressure (dark blue above) (and at the end even extend injection duration) to keep engine running but then runs out of fuel and dies ????

Would the engine cut off even at standstill (at 2500 RPM) if you would run it a bit longer with all possible electric loads on (lights, heaters, heater blower, etc.)??? Perhaps you could monitor the fuel pressure at same time?

Have you checked fuel lift pump is running (Fuel system pressurisation pump) is running?
A quick test procedure here (I believe it's the same also on your engine)

[T6_measured] 01-Engine - Built-in test for Transfer fuel pump

A quick DIY way to test fuel lift pump's basic functionality. The test takes advantage of engine's fuel high pressure sensor (in fuel rail). Running the lift pump continuosly makes the pressure eventually "leak" through high pressure pump and can be detected by fuel pressure sensor. As a side...

www.t6forum.com

 

[Limpy]

I've done the stress test at 2500rpm, engine runs as normal, but I will do it again and monitor the fuel pressure.
I will check fuel lift pump tomorrow.
Thank you for the link.
It's a mystery this, Do you think it can be a bad sensor? Like the cranksensor or the camshaftsensor? I guess that would've shown on the graphs.

 

[mmi]

Do you think it can be a bad sensor? Like the cranksensor or the camshaftsensor? I guess that would've shown on the graphs.

Yes, I would expect sensor hiccups would show on the graphs. The RPM above is from crankshaft sensor. Of course would be interesting if you could make a recording having IDE00406 Camshaft RPM added into the setup.

IDE00021 Engine speed​

IDE00352 Main injection: duration of activation​

IDE00406 Camshaft RPM​

IDE00589 Fuel pressure​

 

[Limpy]

Here is the new stress test and the fuel pump test.
I hold the parked max rpm at 2500 for a while, then let the van roll forward so I could get the revs up to over 3000 then it died.
I did not have the IDE07824 Tank-internal presupply pump 1 bank 1 option.

 

[mmi]

I hold the parked max rpm at 2500 for a while, then let the van roll forward so I could get the revs up to over 3000 then it died.

Ok, Here is the plot:



Interestingly Camshaft RPM reading "stucks" at 315.8 rpm when engine dies. Not sure if that's relevant, or normal?? The reading seems to spike up at the time engine starts to die. Please note, because of still relatively slow sampling rate we really can't be sure of the exact timings between events - a kind of chicken-egg problem.
Perhaps something to try with your other van (shut off the engine by using stop/start switch, or turning ignition off and quickly back on to keep data logging alive).

---

The 2nd attachment seems to be somehow corrupted - data headers are missing, as well as 40 seconds of data from the begining. I guess the orange line is the sensor voltage - but no level change?? - no pressure change???
Did you hear the lift pump running (ignition on, engine off??

I did not have the IDE07824 Tank-internal presupply pump 1 bank 1 option.

Quite possible that the IDE-number or the name are not exactly the same - anything similar? For these kind of things the blockmap file would have been very useful.

 

[mmi]

It seems that engine CAAB has TWO electric fuel pumps! Have you checked fuse SD30 (20A)?

What tests are available in VCDS > Engine > under a) Output tests b) Basic settings? Can you make a screenshot of those?

 

[Limpy]

Indeed the camshaft reading is interesting, don`t know if that is a normal reading me either.
I tried to do the same test
with the working van, was somewhat difficult to cut the engine and still
getting vcds to read it, it has a ignition key and not push start button.
Nevertheless I have attached the csv file.

I did 2 new fuel pump tests, I can hear the pump.
I`m following the steps in your fuel pump thread but I can`t get the data to log before clicking on "Go". That`s why there is 30 to 40 seconds missing in the beginning.
I included the only tank reading option I got.

The fuse is good.

I took pictures with my phone of the options in "Output test" and Basic Settings" because the drop down menu closed on me when trying to sceenshot it.

I hope they are clear enough.
Thank you for analyzing these files.