N53. An interesting problem

The E series vehicle owner with the N53 series engine observed the following problem: sometimes, the RPM in idle was tended to fluctuate.

Like this:

Should the car of the premium segment behave like this? Of course, not!

The engine is in technical order; it uses a Stratified charge, no error messages in the DME error message memory. 

The car owner is technically an educated person, so he waited for the problem to appear repeatedly and, at that moment, made the video recording of all main menu frames. He approached me with an exact question: Are the changes in Lambda’s value of the first bank reasonable? He has correctly identified one of the symptoms!

Pay attention to the Lambda value/voltage of the primary probe, marked in the image, and the control probe’s indications.

Here, a video:

As we see, Lambda of the 1-st bank is changing cyclically in the 2-nd bank; instead, the fuel mixture is stable and correct.

Such a problem can be caused neither by EGR nor throttle, not by the issues of the pressure sensors and/or the pump problems. In such a case, there would be problems in both banks.

Theoretically, such a defect could be caused by the problems of the Lambda probe. But Lambda is not changing chaotically. Lambda is changing cyclically, getting correct after each second, then again – the fuel mixture gets rich. Even more – the control probe confirms lean mixture at one moment and rich at the next. And this also has a cycle of a second!

Such exact cyclicality means only one thing – DME itself is changing the fuel mixture! But why so? Could it be the process of test measurement/measurements/adaptations of some hubs? Why is the result incorrect? Even the idle is fluctuating!

There is one component, which has a direct impact on the fuel mixture in each cylinder. It’s the injector. DME performs different tests of the injectors, also in idle. Could be any of the injectors damaged?

Let’s check the results of tests of injectors/cylinders:

1. Last results of the cylinder tests, Homogeneous mode:

2. Test results of cylinder tests for short and long injector opening, Homogeneous injection mode:

3. Test results of the injector leakage, idle, Stratified charge:

4. Measurement results of the injector flowrate corrections, Stratified charge:

As we see, all data are correct! Yes, the 1-st cylinder’s injector is slightly leaking in a closed condition, but – the leaking is acceptable.

Let’s see what is happening to the injectors’ live data at the moment, when the idle is fluctuating. Pay attention to the:

a) opening of the 2-nd impulse of the injectors

Here, the video:

As we see, The opening of the injectors is changing cyclically, around once a second! From 170 .. us to 250 .. 260us and back!

b. energy changes of the opening/closing of the injectors, such as data of 1st cylinder’s nozzle:

Here, the video:

Info: in idle of both injection modes – Stratified charge and Homogenous mode – the fuel is injected by two injection impulses. Adaptation data of the first impulse includes compensation for the injector’s leaking in a closed position; data of the second impulse includes adaptation compensation for the differences of the injector delay parameters. Unfortunately, INPA live data shows only – so we observe them.

As we see, the openings of the injectors of the 2-nd impulse periodically are dramatically different; the opening energy – fluctuates!

What has happened?

The answer: DME performs the measurements of the injector delay parameters for idle, in Homogeneous mode! More about these tests (in the link below – example of Stratified charge) read here:

tests of leakage and atomization quality

In Stratified charge mode, these tests allow to identify the atomization quality of the injected fuel, but in Homogeneous mode – measurements of delay parameter of the injected fuel, ignoring the atomization quality.

The problem becomes more apparent – obviously, the parameters of some injector during tests are sharply different from planned expected!

Let’s see which cylinder is working incorrectly!

Here, the video:

Here, the mechanical efficiency of cylinder No.1 periodically becomes sharply increased! At the moment, when the fuel mixture of the 1-st bank becomes rich, the mechanical energy of cylinder No.1 increases! There is no doubt – at that moment, the injector of cylinder No.1 starts to leak the fuel. Obviously, it does not close in time!

Considering that the chemical tests of cylinders for Homogenous mode by short injector openings are correct, I assume that the problem is the following:

by very short openings (around 150 us) the injector fails to open completely (it is normal – such opening can be compared to the opening time of the injector, so-called “fly time”). For some reason (assuming that due to wear of the nozzle) is stuck in a half-open position. After a particular time (when U on the injector’s piezo elements has dropped, the element is in the released state), the nozzle, however, closes. Still, during this increased “fly” time (while the U is dropping), the nozzle has managed to inject an inadequate fuel amount. The defect is very atypical. Usually, such stuck of the injector is more unstable; the cylinder catches misfires, DME records the error messages. But in this time – only fluctuating of the idle without any error messages in the DME error message memory!

Finally – probably, you will ask: why DME has to perform such complicated injector tests? Everything is in the name of the perfect performance of the engine! By short opening, the math of injected fuel amount vs. opening time becomes strongly nonlinear. For smooth and correct operation, the injector’s behavior/parameters should be known very precisely! To understand the parameters of the injector, they should be measured. DME is doing precisely that!