This entry is regarding one exciting letter I received from a “brother in arms” from Germany. Why am I making a public entry? I will answer as much as possible for this entry to be helpful for the repairs of other engines, too. The German colleague is reading my blog very carefully; the live data analysis is performed based on information from my blog. Obviously, there are still some nuances that were not explained clearly enough.
Given: the N55 series engine (with stock software – it is essential because, with a “chipped-up” software, you can expect all kinds of surprises), the car has a manual gearbox.
The service engineer is writing:
And the calculated average values in the rough run menu look like this:
Cyl 1: 0,013
Cyl 2: -0,158
Cyl 3: 0,040
Cyl 4: -0,179
Cyl 5: -0,725
Cyl 6: -1,468
All values are jumping +/-0,500 all the time with some real misfire between on all cylinders.
What ever you do, what ever you change, the pattern stays the same and never change.
The only things we didn’t touch from the engine top are the exhaust cam ledge incl. ex cam and what’s underneath, plus exhaust manifold and turbo.
All values are good and consistent. We did a compression test and pressure loss test. Pristine!
Cyl. 5 and 6 are always suspicious.
The efficency on cyl 5 is increased 7 times compared to cyl 1.
The efficency on cyl 6 is increased 2 times compared to cyl 5. And more than 14 times compared to cyl 1.
First objection: it is incorrect to compare the “quality” of cylinder performance like this (mutually)! Why?
In this example, the difference between the first cylinder (to which all other cylinders are compared) for the ideal is 0,013. What will happen if the short-term efficiency coincides with the ideal? Then, will the conclusion that all other cylinders are performing a million times worse be drawn? It will be a completely wrong conclusion!
How should the cylinder efficiency be evaluated?
The relative efficiency of cylinders should be evaluated against the zero/full performance of cylinders. How to do that? Read here.
A short recommendation:
a. start the engine, take off the plug of the ignition coil of the 1st cylinder for a moment;
b. evaluate the Rough run data of the 1st cylinder; these values correspond to the unworking cylinder;
c. repeat the test with the 2nd (by placement) cylinder; make sure the order (placement or firing) of the cylinders is depicted in the exact menu.
Example.
When disconnecting the ignition coil of the 1st cylinder, we see that the Rough run data of this cylinder indicates +50 units. When disconnecting the ignition coils of the 2nd cylinder, we see that the menu displays data in the placement order.
Conclusions (using data mentioned above):
1st cylinder works with average “right” efficiency. 5th cylinder is for +1.5%, but the 6th cylinder – is for around +3.0% more efficient than the average.
Is +1.5/3.0% a lot or a little?
Siemens DME tries to reach the ideal balance of mechanical efficiency. Bosch DME is not so categorical; it allows disbalance to +/-5%. Of course, these are observations “from aside” because no exact information (including formulas and coefficients) is available.
In addition, we have to understand that in the case of N55, problems are created both by injectors (yes, these are not the capricious piezo injectors, but still, they work with a high Rail pressure in DI mode) and the air intake (because the engine is equipped with Valvetronic, which gives the additional difference between cylinders). If the air amount between cylinders differs (and we CAN NOT affect this difference in any possible way), then by changing the injected fuel, we can reach:
a. a perfect fuel mixture in each cylinder, but the mechanical efficiency of these cylinders will be different;
b. a perfectly even mechanical efficiency, but then in some cylinders, the fuel mixture will be rich, in some – lean;
c. compromise – yes, the mechanical efficiency of all cylinders will not be perfect, but the fuel mixture will be pretty similar; it will reduce the CO catalytic converter load and allow it to fit in the emission norms.
So, everything is fine if we see the difference till +/-5% for the engine equipped with Bosch DME. Such difference does not give permanent vibration; the engine runs (subjectively) evenly.
Would you rather suspect a mechanical problem around cyl 5 and 6
In this case, the mechanical efficiency of cylinders No.5 and No.6 is INCREASED! If these cylinders had some mechanical problems (for example, damaged valves, significant losses due to damaged walls of the block, the reduced amount of air due to HVA of them were of the Valvetronic eccentric shaft), their efficiency would be REDUCED! The answer – is no, for sure, these cylinders don’t have any mechanical problems!
or do you think it could be DME (injection control, measurement, …) related?
No, I can not imagine a DME defect that could cause such (even if we abstract from the data of this current example) a disbalance of cylinders.
Faithful, though, there is one obstacle that seems strange to me in the given data. My experience (and also the information in the BMW TIS) says that the sum of the mechanical efficiency of all cylinders should be 0. In this example, it is not; the sum is <0 (around -2.477). Such a result is possible in one case – the flywheel is NOT ADAPTED! I have often written in my blog about how to adapt the flywheel. Obviously, this stage (which is very important) has been surpassed during the maintenance.
Second note – if the flywheel adaptation was not performed, the adaptation procedure was obviously dismissed ultimately. So the first job – is correct re-adaptation of the engine.
Imagine you have a stock manual N55 and checked everything and changed everything that makes sense.
But Mode 6 still shows you misfire on all cylinders. No repeating pattern.
Multiplicative adaption shows 0,94, DME tries to lean out the mixture, maybe due to misfire and unburned fuel.
First note regarding LTFT 0.94. Unfortunately, Bosch DEM does not divide and does not indicate Offset type LTFT. Instead, the Multiple type LTFT 0.94 (-6%) is a very correct value. Typically, the allowed corridor is 0.70 .. 1.30.
More critical – evaluate Lambda (the average fuel mixture and Intergrator. Unfortunately, the Integrator in Bosch DME (for N and B series) can be seen only via the OBD2 standard. INPA/ISTA does not show the Integrator value.
Second note: during the more extended driving session, misfire counters CAN differ from 0. Several events in each cylinder during 15 .. 30 minutes driving session – a typical result for a correctly working engine.
Increased density of misfires for these engines can be (if the fuel mixture and the cylinder performance are correct):
a. due to uneven (with potholes) road;
b. at the moments when the gears are switched (if the car is equipped with a manual gearbox);
c. less likely – due to atomization/beam problems of the injector.
In this case, we should ensure that the misfires are not registered at the moment of gear switching. Read more here.
In conclusion – regarding the replacement of the spare parts. First, if the problems (in this example – mechanical efficiency of 5th and 6th cylinders, different from 0 – even without analysis: this is or is not a problem) are with any of the cylinders – it MAKES NO SENSE to replace spare parts, which correspond to ALL cylinders! Second – replacing the spare parts without an accurate understanding of the necessity of this process creates only new problems! Typically, such testing is performed with used spare parts – the engine is “tortured” with untrusty, used spare parts, whose tolerances (difference prom ideal) are increased and unpredictable. Only damaged parts should be replaced!
