2.0 TDI CR 170. Overboost but NO fault codes!?

[Faceliftmike]

Hi, could really do with some collective wisdom on a problem with my Tiguan. '58. 2.0 TDI CR 170. CBBB. 73k.
It starts and runs perfectly and drives very smoothly with no smoke.
While under boost conditions it pulls strongly for a few seconds then suddenly looses ALL power. If I keep the throttle to the floor it just rolls to a stop. If I lift off for a second it recovers and is back to normal unless I let the turbo boost hard which makes it repeat the process. No limp mode, no EML and, when scanned on multiple systems, no fault codes.
The vehicle has had a huge amount of diagnostic work and parts replaced and nothing makes any difference, the fault stays exactly the same.
The latest conclusion which I generally agree with is that the turbo vanes are clogged, as a physical gauge on the manifold shows that it is seriously overboosting and a new actuator, n75, and boost sensor have made no difference.
Before I go to more expense of stripping the turbo out can anyone suggest any reason why this would be happening, the ECU reacting to it, and yet it not generating any codes? The forum is full of p0234 codes and limp mode problems for clogged turbo vanes.
We have even tried replacement ECU, and rolled back the software to original.
Any suggestions would be much appreciated!

[Crasher]

An interesting one, it cannot be going into any variation of a “limp home” mode or this would be stored as a DTC so what you are experiencing is outside of the ECU’s ability to see so suggests fuel starvation BUT this should also be picked up. I imagine the fuel filter has been changed so do you know if the fuel delivery system was correctly bled after? It is not unusual to replace one fault with another which feels the same.

[Faceliftmike]

Hello Crasher,
Thanks for your reply. To avoid posting a 'novel' I did not include the whole story but, as you have suspected, so our initial suspicion was toward the fuel side of things as it has just a couple of times thrown a p0088 code which is fuel rail overpressure. We have exhaustively been through the fuel system, and can find not faults. The auto diagnostic chap that has currently been looking at it has proved that at the point it shuts down the rail is still at full pressure so then started looking at the air side. The gauge reading off the manifold proves it definitely is overboost, which we think is causing the ecu to shut down the injectors so the fuel rail momentarily goes over pressure as a symptom. This all makes sense except for the fact that the ECU is making a decision that there is a problem and to shut off fuel injection yet not registering a code. Arrhh!

[Crasher]

I am not sure his diagnostic equipment is telling the truth, in a situation like this I would want to know it was not mixing this up with

16471 (P0087) Fuel Rail System Pressure: Too Low

[Faceliftmike]

Prior to me getting the latest chap involved it had been to a diesel injection specialist who essentially gave the whole fuel system a clean bill of health. They tested the injectors and rail pressure and said that.everything was operating within parameters and suggested it must be an ECU/software fault.
Now an exchange ECU and a software rollback (it had had the ea189 emissions fix applied) have discounted that too.

[Crasher]

Well if it logged a fuel pressure code that was too high, something is wrong. I have a confidence struggle on the CR engines when it comes to deciding to fit a new high pressure pump because it is not a quick easy job and very upsetting for all involved if you are wrong, it is quite nerve wracking. Something that is also becoming clear is that bench testing these injectors is not definitive as the test fluid used has a different lubricity to The Devils Urine which affects the way the injectors internals react to the stiction the moving parts experience as they age and this masks the problem whilst under test.

[Faceliftmike]

Fitted a new Hpfp, fault stayed exactly the same. Tried another set of injectors from a fault free vehicle, made no difference!
However I am excited to say that this has been going on for months but I took it out for a drive yesterday and for the first time ever when the fault occured it went properly into limp mode with the coil light flashing- I've never been so pleased ��.
I will plug it in later this morning and hopefully find out what's going on!

[Crasher]

Where did the HPFP come from, genuine exchange, Bosch or other?

[Faceliftmike]

Initially we put a breakers pump on it but then changed it for a reconditioned and guaranteed one, both original bosch. The fault stayed exactly the same both times.
I took the car back to mechanic#1 yesterday as he is a friend and is mobile. The new fault code was for underboost so we think that mechanic#2 has set the actuator to the minimum he thought he could get away with which had caused this. By running a Delphi test cycle on the turbo, driving it with the vacuum pipe disconnected from the actuator, and inducing a boost leak by undoing the intake pipe we can definitely say that the turbo is working fine and the fault is not boost related as it still happens with no turbo boost at all. Also from each of these tests the car did generate the appropriate fault code for what we had done. It also never produced the fuel rail over pressure one even though we induced the fault many times so I am still sure that this occasional code is not directly related to the fault.
So it's Not air/turbo.
It's not fuel pump, rail sensors or injectors. (All been swopped out twice)
Not software UNLESS having the ecu software overwritten back to factory version did not actually give me a completely fresh start?
We drove it around with laptop/diagnostics connected and it looks like at the moment it faults it is the ecu cutting injection like it thinks that is what it should be doing as it doesn't generate a fault code when it does it. ( Not realtime data though) It also won't recover unless you release the throttle for a second and then recovers quickly so again it feels like an electronic 'decision'. With no fault code generated it's like it in some way 'thinks' what it is doing is correct, I am at a complete loss...

[Renars]

If there was fuel rail pressure too high code Its had some abnormality there, even if ECU closes all injectors instantly rail pressure regulator should be capable to release pressure without any error.
You swapped ECU itself or just rolled back software ?