This time I will describe the situation – problems of persistent vibration in idle.


If any of the injectors is leaking so much, that DME is not able to maintain the fuel mixture or group misfire starts and a cylinder is switched off – it’s clear: check ../F5/F6 (fuel mixture), identify the bank which has problems. Then the work with 2 or 3 cylinders follows.

If a persistent vibration in idle/Stratified charge can be felt, check ../F5/Shift+F6/F5, if some injector doesn’t have inadequate corrections in Stratified charge.

If the engine has not completed the individual adaptations of the cylinders (if any injector has a serious problem – DME will never finish these adaptations), or by some other reason the engine in idle works in Homogeneous mode (and a persistent vibration can be felt), the menu, mentioned before, won’t help – its data will be outdated and can indicate an inappropriate results (for example – all injectors are working correctly).


If the engine in idle uses Homogeneous mode, but persistent vibration can be felt, the only menu which will give information: Rough run (../F5/F7).


Nuances, which have to be taken in account for idle of Homogeneous mode:

a) the fuel has enough time to mix with air. Accordingly – even bad atomization (the fuel is not injected as a fine mist) don’t give fundamental problems. The most important – the amount of the injected fuel;

b) DME tries to even the fuel mixture in such a way, so it is stoichiometric (Lambda 1.00) in the range of a whole bank. It means – if in any cylinder an excess amount of fuel will be injected (typical defect of the injector – it’s leaking), the fuel mixture in other cylinders will be lean.


All aspects, mentioned above, means: if any of cylinders have a leaking injector (it leaks when it is supposed to be closed – a typical defect), the disbalance of fuel mixture in cylinders appears. The fuel mixture of the damaged cylinder is rich, in other cylinders of the bank – lean.

The amount of air is even in all cylinders – it’s defined by the opening of the throttle. In case of a defect, mentioned before, the damaged cylinder works with good mechanical efficiency (because all oxygen will be burned), but other cylinders of the bank will have a reduced mechanical efficiency (because they will have not enough fuel to burn).

So – you have to look for cylinders, which has an increased (relatively) mechanical efficiency.


An example. One of BMW users sent me a following Rough run menu with a complaint of persistent vibration in idle:

How to interpret Rough run data, read here.


In this case, the patient is a car with an N53 series engine, which means:

a) sequence of cylinders: firing order (1/5/3/6/2/4);

b) increased efficiency of cylinders: negative bar value;

c) completely non-working cylinder: around +8 units.


At the first moment the video looks totally incomprehensible, isn’t it? Each cylinder – completely different mechanical efficiency!


At first, let’s “sort” the cylinders according to banks. Cylinders of the first bank (by Lambda correction) are on the left side, cylinders of the second bank – on the right side.


Note: for N43 series engines (which have 2 wideband Lambda probes), in first banks are 1-st and 4-th cylinder, in the second bank – 2-nd and 3-rd cylinders. Placement – equal (first bank – on the left, second bank – on the right).


In this case:

Looking at this video (each bank separately), we see the 1-st bank has a larger problem (on the left side of the screen) – efficiency of the cylinders of the 1-st bank is dramatically different:

a) 1-st cylinder (1st: both by firing order and by placement) has a significantly increased mechanical efficiency;

b) 5-th cylinder (menu: firing order, 2-nd by placement) has a significantly reduced mechanical efficiency;

c) 3-rd cylinder (3rd: both by firing order and by placement) has a relatively correct mechanical efficiency.

Mechanical efficiency of 1-st and 2-nd cylinder (by placement) are several times different!


At the same time, the average fuel mixture of the whole 1-st bank is close to optimal!

In case of a perfect fuel mixture in Homogeneous mode, the bar will indicate 2.00V which corresponds to Lambda 1.00. In the current situation the fuel mixture is slightly lean, but the difference from an ideal is tiny – DME has successively “pulled” the fuel mixture in a range of the bank.

So – in idle the engine works in Homogeneous mode, with a correct Lambda; Stratified charge is not used (one of the reasons – significant problems in Homogeneous mode: obviously, DME is not able to continue the tests of chemical efficiency of cylinders in this mode): ../F5/Shift+F6/F5 menu as a source of actual data of injectors is not legit.


If we exclude the possibility of the mechanical damage of the engine, we can make the following decisions, looking at the data of 1-st cylinder:

a) the mechanical efficiency of 1-st cylinder is the largest – it’s very possible, that its injector is leaking (in the moment, when it is supposed to be closed);

b) performance of the other two cylinders (of 1st bank) is significantly lower.


Note: depending on the software release, Rough run data can be with a different “sensitivity”, and also indicate different data in case of significant problems of any cylinder. For older software releases – if any cylinder has seriously increased or reduced efficiency, other (successive, by firing order) cylinders are “waving” strongly. In newer software releases, the successive (by firing order) cylinders are “waving” significantly less.


You have to take into account this waving – exactly by this reason, the solving of the problem have to be started with a cylinder, which indicates the most differences from other cylinders.


After replacing the injector of the first cylinder (even without re-adapting) after a short driving session, Rough run indicated the following data:

Here – the result close to perfect! Mechanical efficiency of all cylinders fluctuates around 0 – the bar values change polarity, so – DME has managed to even the mechanical efficiency of the cylinders. Bar deviance from 0 is around +/-0.250, which corresponds around +/-3% of the difference of mechanical efficiency from ideal.

In addition, the engine works in Stratified charge – so the engine has managed to adapt the mechanical and chemical efficiency of cylinders in Homogeneous mode: one more good news! In this video, the engine works with Lambda 2.00 (wideband probe voltage around 2.50V), the correction data of injectors in idle in Online mode are changed in ../F5/Shift+F6/F5 menu (4th .. 6th lines). This menu allows evaluating of all injectors (their performance in idle) because contains info regarding amount and atomization quality of fuel, injected by each injector.


Next step: analysis of data of ../F5/Shift+F6/F5 menu. Regarding this menu read more here.