Knock sensors M54B22 engine

[Bazz0847]

Hi all,

This is my first post here, after reading along for a while now. I have a serious issue with the engine of my car, but let me first introduce the car :-)

I own two cars: an E46 with M54B22 and a Z3 2.2iS with M54B22 engine. Both have the same mileage: ~130.000 km. The E46 is flawless, but the Z3 is the problem child here, engine-wise. I've bought the Z3 second-hand from a private owner. The car looked and drove fine, however the engine did make a loud knocking sound below ~1000 RPM. Furthermore, the engine made a rattling/pinking type of noise at full throttle and low RPMs. It appeared to be bearings, which meant that the engine needed replacement, which I did.

The loud knock is gone, however there still is a pinking noise when the engine is hot and I step on the throttle from a low RPM. I'm quite sure its detonation, and I assume that this detonation has destroyed the previous engine, as it makes the same noise. I want to resolve this ASAP, but I can't figure it out. Knock adaptation reads 0.0 no matter what RON fuel I refuel with, which confirms my presumption that it could be knocking, as the DME is not adapting properly to the fuel type. There is no knock-sensor fault code, and the test mechanisms are quite exhaustive on these sensors, so I assume they operate correctly. There is no fault code and I see changing voltages on DIS knock sensor inputs. The E46 I own, with the same engine, reads adaptation of 0.5 with RON95, while the Z3 reads 0.0. The values given by the E46 correspond to the values given in BMW TIS for that fuel type.

To resolve this issue, I did the following:
Flashed the latest DME firmware: no change.
Put in new (original) NGK spark plugs: no change.
VANOS rebuild: no change.
Cleaned the manifold, MAF and injectors: no change (however, throttle response improved drastically).
CCV replaced twice, no change.

Because the sensors actually DO seem to give voltage spikes (I can definitely see the voltage changing at the moment I tip-in the throttle) at the moments the rattle occurs, it appears the DME just ignores them. Could it be that the previous owner had the DME chipped/coded to "improve performance", such that the sensors are ignored? Could the DME be defective (which I doubt as the analog knock values read over OBD are correct)? Any suggestions? Could the front and rear sensors be interchanged? How to test this? I'm totally clueless here!

There seem to be more people with this issue. For instance, this youtube-video shows exactly the same symptoms as my car does: [Only for registered users... ] and I can find a bunch of posts with the same complaints. Not sure whether they also have the knock-sensor not working. The issue prevents me from using the car for weeks now :( Of course, I'm focusing on knocking, but it could be something different as well, such as the tranny mounts making noise, or whatever, but I assume knocking due to the adaptation reading zero continuously.

[Bazz0847]

Some more info: I've used DIS to find the knock sensor mean output voltage at idle and 2500 RPM. This is what I found, for sensor 1 and 2, respectively:
0.17, 0.11 at idle
0.57, 0.56 at 2500 RPM
Knock adaptation: 0.0 at both RPMs.

This is with RON95 fuel. Are these voltages normal? Will test tonight on the E46 to verify these values with that engine.

By the way, I placed this in coding because I suspect the ECU is simply neglecting the sensors. This can be due to a bad "chiptuning" done by the previous owner. Can I verify whether coding has disabled the knock adaptation? Is this even possible?