Skip to content
Shop
Explore

INDEPENDENT BMW ENGINE EXPERTISE

VANOS problems. Part 3

N43 / N53 · NOXEM · DIAGNOSTICS

Select a technical image to view it in detail.

Qualitative Indicators of VANOS Performance

This part of the article is a fairly scientific treatise, so feel free to skip it if you don’t want to go deeper into the theme. If you want the justification for my conclusions from a side view, read on. I don’t know the manufacturer’s own calculations, so these conclusions rest on observation. The practical part of the article follows this one.

Reference positions of the camshafts

First, a short review of camshaft reference positions, compiled from a significant amount of engine diagnostics data (around two thousand engines).

BMW AG sets the camshafts correctly at the factory, and I don’t think anyone doubts that. By the DME’s own readings, the reference position differs from the ideal by more than 4 to 5 degrees only in sporadic cases, even on quite heavily used engines.

Originally (a long time ago), the allowed deviation of the reference positions (the “green” zone) was ±10 degrees. The Siemens DME was very quick to log a reference position error, already at ±11 degrees from the ideal. The Bosch DME was slightly more patient and allowed an additional 3 to 4 degrees as a reserve.

When the DME software was updated, these limits were widened. I don’t have exact data, but practical observation suggests an additional 3 to 5 degrees was granted. On engines with considerable mileage, camshaft reference position errors start to appear because the chain stretches. Evidently the manufacturer received a significant number of complaints.

A second possible reason for the larger allowances is that dealer technicians could not “correctly” install the camshafts after an engine repair. I put “correctly” in quotes because the repair was done by the book: the chain was tensioned and the BMW camshaft locking tool was used. Yet the repair was still unsuccessful. How can that be? Here I describe one of the reasons. The manufacturer cannot prevent the free movement, which logically increases as the hub wears, nor compensate for it automatically. So the solution was to increase the technological tolerances.

Accuracy of maintaining camshaft positions

To keep this discussion technically correct, we must distinguish between three terms: measuring accuracy, positioning accuracy and the accuracy of the measurement display. For this reason my conclusions do not claim absolute accuracy, but they do allow the actual situation to be evaluated.

  • Measuring accuracy of the shaft position is the highest. Leaving aside mechanical problems such as free movement of the hubs, it is defined by the quality of the sensor, the resolution of the MCU (internal timers and interrupts, timer frequency, code quality) and the accuracy of the mathematics (the type of data registers and memory cells used). The hardware and software developers define all of this, and we can only rely on there being no gross mistakes.
  • Positioning accuracy is lower, because on top of the measuring error come other problems. These include free movement of the VANOS mechanism, short-term inaccuracies from oil pressure changes, vibration, changes in chain tension and uneven engine running.
  • Display accuracy is the lowest. In addition to all the errors above, rounding errors appear at this stage, both within a single measurement and across the displayed series of measurements.

Why such a long introduction? This time we are dealing with data and math. In the rest of the article I will state what I believe the correct accuracy of camshaft position control should be, and explain how I arrived at those conclusions.

What the diagnostic software actually shows

INPA

In the VANOS menu for MSV70/80/MSD80, INPA refreshes data about 3 times per second. The situation is very similar with Bosch ME/MEV. The display accuracy is:

  • reference positions: ±0.1 degree
  • required and actual camshaft positions: ±0.4 degree

From observing the data, it appears not to be averaged. INPA displays the current measurement, taken at the moment it is sent over CAN to the diagnostic equipment. At idle we therefore see only every 7th to 10th camshaft cycle. That is every 28th to 40th position measurement, since the toothed ring has 4 teeth.

ISTA

ISTA behaves differently. In the call-up functions, it gives no information about camshaft reference positions. Required and actual positions are displayed with two digits after the decimal point, implying a resolution of ±0.01 degree. The refresh rate runs from 3 results per second (with only 1 to 2 parameters selected) down to 1 per second (with 5 to 10 parameters). The measurements are rounded between display cycles, and this rounding is where the apparently high accuracy and resolution comes from.

Because of the rounding of both the measurements and the displayed data, it is hard to judge the true data resolution in Expert mode. One thing is clear, though: the software developers did not economize on computing resources, and all parameters are displayed with a resolution of tenths or even hundredths of a degree.

So in both INPA and ISTA we see only every 7th to 10th camshaft cycle at idle, and the data become even sparser as RPM increases. ISTA additionally rounds the data between outputs. In Expert mode the data density is higher, but you still will not see every measurement (cycle).

Consequences for short-term faults

The most important point is this: if the camshaft “shivers” for a very brief moment (tenths of a second), then

  • in INPA we will not see the problem at all, or, if we are lucky, we will see it very clearly;
  • in ISTA call-up functions we will see a significantly reduced “amount” of the problem;
  • in ISTA Expert mode logs we will see the most correct picture.

For these reasons I suggest the following:

  • When using INPA, observe the data several times during transition processes and take the worst readings into account.
  • When using ISTA, use Expert mode if it is available. If it is not, unfortunately, you will not be able to see short-term problems with the camshaft position. The good news is that a VANOS with problems in transition processes typically also has problems maintaining the average (long-term) position.

What values are correct?

This brings us to the essence of the entry.

INPA. A correctly working VANOS hub under stationary conditions holds the camshaft position within ±1 displayed “step”, which means within ±0.4 degrees. Allowing for hub wear, technological tolerances and so on, I consider ±2 steps, or ±0.8 degrees, acceptable. During transition processes (when the required position changes quickly), the short-term error is of course larger. Even so, the delta between the displayed required and actual values should not exceed a few (2 to 3) degrees.

ISTA, call-up functions. Because of data rounding, you will not manage to see short-term problems. You will not even see the required camshaft positions, so in this mode we can only evaluate “swimming” or slow jitter of the actual position. For a correctly working VANOS, the swimming of the (rounded) values should be no more than ±0.5 to 1.0 degree.

ISTA, Expert mode. Expert mode lets you evaluate the difference between the instantaneous value and the required value. Correct readings, analogous to INPA, are no more than ±1 degree. In this mode we also see how the VANOS hub reacts during transition processes. Within a tenth of a second, the actual position should match the required position to within a few degrees.

BMW’s own opinion on the allowed error of the actual position can be seen only in the ISTA VANOS test menu. For Bx8 engines, for example, the allowed short-term deviation is as follows:

The range is quite wide, but even a brief moment outside it is enough for the test to be declared failed.

BEYOND THE FAULT CODE

A clearer picture.
A better next step.

Get in touch
Enlarged technical image
Open original image