VANOS. Hidden problems. Part 2

In this entry – about one quite common situation, related to the VANOS. The current problem has quite often seen both for N43/52/53 series engines with MSV/MSD Siemens management system and N42/46 series engines with ME/MEV management system. Also, M series engines are “familiar” with this problem.


Symptoms of the problem: occasionally the error message regarding the inadequacy of VANOS position (reference). Uneven running by very low loads (in idle) are possible, of course, the error messages regarding the uneven performance of VANOS will be recorded, also the engine can switch to emergency mode – the VANOS turned off.


What are suggestions from authorized BMW service centers (including dealer centers)? At first – replace the chain, it’s stretched out!


Let’s calculate (using a simple calculation) this possibility.

MSV/MSD engine management systems allow +/-10 degrees difference of reference value.

ME/MEV engine management systems allow +/-11 degrees difference of reference value.


To reach defiance of 10 degrees (the gear with a diameter of 10 cm – approximate diameter of VANOS gear), the chain length from the crankshaft to VANOS gear has to stretch for:

(PI*D)*(10/360) = 9 mm

So the whole chain has stretched for around 20 mm (2 cm)? I believe, anyone, who knows the specifics of the engines, will confirm, that such stretching of the chain is not possible without breaking.


Needless to add, that in this situation the replacing of the chain will not help. Even more – for example, for the N46 engine (exactly for this engine the experiment, described below, was performed) in case of chain stretching the defiance of reference value was in the opposite direction and not as it deviated for current engine. Yes, the owner of this vehicle has replaced the chain (and it didn’t help). Several times marking of crankshaft and camshafts were checked.

Reference values for both VANOS knots:

As we can see, the reference value of exhaust shaft is strongly deviated “to the left”, and it exceeds possible allowance. As a result, the PWM of exhaust VANOS regularly drops to 5% – the valve is completely closed, and also after some time (irregular intervals) the error message regarding reference position of VANOS position.


When the replacing of the chain don’t help, next to more popular component, which is replaced – the valve. Of course, it’s possible to check, if the valve really closes, but usually, the replacement of the valve doesn’t help.


Next element, which is usually replaced – the VANOS mechanism. I haven’t heard a justification for this replacement in the current situation. As if only – there is a problem with VANOS, so we are replacing it. Unfortunately, the mechanism will not be one to blame in this case.


Are some people replacing also camshafts? Why? Also – no justification ever heard. In fact, DME doesn’t even see, is the camshaft screwed in and in which position it’s screwed in. DME sees the position pf VANOS impulse disc.


So for the N46, an experiment was performed – the position of the impulse discs (by fixed camshaft, crankshafts and chains) was deliberately changed. Turned out, that impulse rings of both camshafts were with an upcoming phase – as if the chain was shrunk! When the rings were closed in a way, they are slightly (for 9 .. 11 degrees) “late”, positions fell in perfect – center – position! How is it possible?

Additionally, the shivering of the engine, when it was switching from idle to low load conditions, was gone. Also the PWM normalized (and it is logical).


To understand, where the problem hides this time, we have to explore, how the VANOS sensor works. To detect the position of VANOS, Holl sensor is used – this sensor “senses” changes in the magnetic field. The sensor is equipped with a magnet, which creates a magnetic field. When the salient of the impulse ring comes near the sensor, the magnetic field changes and the sensor detects these changes. everything sounds simple and logical. Where could be the problem?


VANOS shaft of the first generation was turning in a range of 30 .. 35 degrees, but, for example, for N55 series engines – already in range of 70 .. 80 degrees. As if it doesn’t require any excellent precision: +/- 5 degrees don’t change anything dramatically, I don’t know the requirements of BMW AG, the tolerance mentioned before comes from logical reasoning and also taking in account “the corridor” of allowances. It is obvious, that the error exceeds planned one for at least 2 .. 3 times.


Taking in account, that the diameter of the impulse ring is significantly smaller (60 .. 70% of VANOS gear), the error around 10 degrees, mentioned before, is created with 30 .. 40% smaller offset – around 4 .. 5 mm. If we take into account logical tolerance of other sizes (position of the sensor in the block etc.), for example, 1 mm, possible error is at least 3 .. 4 mm.

How did this unplanned offset of 3 .. 4 mm occur?


The Holl element of the sensor is equipped with comparator – electrical device (scheme), which switches over by exact changes of the magnetic field. The output signal (signal, which is received by DME) – rectangular. The sensor doesn’t give any information, how does it’s magnet “feels”. With time, especially in lowered and increased temperatures, every magnet losses it’s power – remagnetizes.


In the images below you can see, how the sensor reacts to the impulse ring, when the magnet is in working order (A) and when the magnet is “aged” (B).

IW – impulse wheel;

MF – magnetic flow;

TR – threshold;

SS – sensor’s signal.

The error, which occurs, if the sensor is aged, is colored in yellow.

Cause of the problem – changes of the MF (magnetic field) are not rectangular – the magnetic field changes “analogically”- gradually, when the salient of the impulse disc approaches and moves away. If the changes in MF are decreased (because the magnetic field of the magnet is decreased), the comparator of will switch late, because it’s threshold has stayed unchanged (according to parameters of new sensor).


Yes, this problem can be solved in several ways:

a) introduce an adaptive (variable) threshold, which “traces” real parameters of the magnet;

b) if sensor sends analog data, DME could introduce adaptive threshold, which compensates the “aging” of the sensor;

c) with an existing solution – after the signal (changes in the impulse length), translated by the sensor, detect it’s “aging” and compensate it;

d) detect the aging of the sensor and after an incorrect length of the impulse (as it’s done in case of, for example, ABS sensors, beginning with E60/90) record the appropriate error message.

Unfortunately, none of these solutions is – at least it look’s so.


Depending on, in which moment the sensor fixes the position of the camshaft (in the moment of impulse rings “appearance” or “disappearance), in case of sensor aging DME will “think”, that the camshafts are offset either “forwards” or “backward”.

For example, for N46 series engines, obviously, the position of the shaft is detected, when the impulse ring “disappears” – in case of sensor’s aging the camshaft start to “hurry”.


Summary of this entry – in case of incorrect reference position – possible culprit of its drift is the sensor!

For maximum correct result:

a) check the positions of camshafts, using special tools;

b) replace the rings of the sensor (it can be magnetized – also this obstacle generates extra error messages);

c) replace the VANOS sensor.