MembersPage/MattiasSandgren/TomasAudiSSix (2009-04-14 16:01:04)

Tomas Audi S6

Technical data

Audi S6 2,2t -96, AAN-engine, VEMS v3.3 with Motronic connector,

Box no. 1986

I have done one other Audi: MembersPage/MattiasSandgren/OskarAudiSFour

But that one doesn't have the cam Hall inverter (but works flawlessly anyway).

Here is the config, tables, MSQ and datalog of my problem:

http://www.vems.hu/files/MembersPage/MattiasSandgren/TomasAudiSSix/

Short video 3gp format (plays with VLC): http://www.vems.hu/files/MembersPage/MattiasSandgren/TomasAudiSSix/tomas_audi_s6_problem.zip

(file upload doesn't allow 3gp ending so I zipped it)

Problem description

Above 4000 rpm the engine runs very well without hesitations or other problems, feels strong and runs with a clean sound. No sign of trigger, ignition or other ECU problems.

The problem is below about 4000 rpm we have something that can be described as slight misfires or the engine simply cutting out. No real signs of problems in the datalogs from what I can tell.

The idle is slightly rougher in quality than stock and the lambda value varies just a bit too much for this type engine.

The daily driving is the "show-stopper". If the engine is given any load above light driving, like a strong acceleration it will appear to misfire.

The datalog shows trying to accelerate in 3rd gear at low throttle, the engine cuts out rhytmically, 2nd gear is then used to accelerate past 4000 rpm and then given full throttle without any hesitation from the engine.

There is no black clouds of unburnt gasoline, which is kind of strange. I have a mobile phone video of the car passing by with light acceleration in 2nd gear, I will upload it later.

The injector pulsewidth in the datalogs look normal, no fuel cut in action from what I can see. No trigger errors either, unless someone can tell me how to tell from an LCD display page (mlp02?).

What is the next logical step? Oscilloscope of the trigger signal? Gas analyzer will probably tell me something is wrong, but won't fix the problem.