NefMoto

I have problems with my new (still in development phase) software. ECU is 8D0907551K, and injectors are EV14 750 cc.

When accelerating hard enough (not necessarily full throttle), motor is jerking quite hard around 2300-3000+ RPMs. If I accelerate lightly, it doesn't jerk at all. It is very annoying, and I feel that it could break something after a while when it does this enough.

I can't figure out that what could be the problem? I don't understand enough the logs that I could find out the fault.

Also when running full throttle, the motor dies just above 6000 rpm, and the timing goes wild. My tuner doesn't yet have a glue that what is going on.

Here are two logs attached, other shows the jerking, and other the "6000 rpm" problem. I hope that you guys can understand these better! :)

who is your tuner?

High RPM jerking is weird. Are you going lean?
(Did not look at the logs).

it sounds like spark plug gap to me. i had the same issue after i changed my spark plugs so i had to gap them down to .025" and now the jerking is gone.

Spark plugs are new Bosch FR5DP's but I guess that I test the FR5DP0R's to make sure that those aren't the problem.

I meant that the jerking is about 2300~3000something. After about 3500 or something it doesn't jerk. But dies just after the 6000 rpm's.

Turbos are RS6's.

Looks to me like your MAF is giving odd readings... Look at your logs, locate position with highly negative timing, like row 143 in 8d0907551k_20111110_072135.csv file. It shows 0 g/s MAF flow at which time the thing retards you to -9 and shuts off boost.

I assume that you have some custom MAF set up with Hitachi conversion? Can you verify that your conversion harness is intact etc?

Could also be a faulty Throttle since it shows in logs that it is open pretty well but might be closed in actuality but I would assume that even with fully closed throttle plate you'd get some MAF activity.

Or, another possibility is that your MAF table is f.up an a rogue row has "0" in for flow at  voltage that would otherwise indicate significant flow. I list that as a possibility since a long time ago Eurodyne screwed up their Maestro defs and outputted brake boosting variable in the middle of MAF table resulting in one dead stop which cause severe bucking of the car. This would only kick in if you modified the variable so was pretty weird.

The MAF is new original RS4, not modified, and wiring harness is not touched. Also as the ECU is from RS4, the MAF table should be correct. How could I verify it?

The car run fine with other ECU (551 C) and original injectors. So the throttle cant be faulty.

I'm very sure that the problem is in the software, but as I don't know ~nothing about the interior life of the ECU, I can't do anything about it.

Well, you're gonna have to read a whole lot about how to use nefmoto flasher (to read off you ecu), use WinOLS or tunerpro (depending on map packs availability for your ecu) and then use the program to go over the maps.

Is that definitely an original ECU, no tune of any sort on it?

I would still just assume that it is a faulty MAF since bosch rs4/early s4 mafs are known to fail quite a lot. Hitachi ones are much more realiable but you cannot just swap it in due to plugs/maf housing port/ECU map differences.

Yes I can read and flash the ECU, that is not a problem :) But then making the map modifications... That needs a lot of learning. My tuner has made these now. But the Tunerpro and Winols are new to me.

The ECU is original, but the tune isn't. I can't try with the original software since I have bigger injectors (EV14 750cc @ 3bar).

The MAF is used about a 1000 km's, must be a very bad luck if it was broken from new. I have other RS4 sensors too, and I can test those also.

On what location in the map I could check that the MAF-table is ok?

MLHFM

You will have to scan all 512 entries in that table to ensure that they raise evenly and there is no abnormal odd value like "0.00" in the middle. The values and resulting curve looks logarithmic when you graph it, it increases all the time by more and more.

If I used the Tunerpro correctly (I opened the 551K-binary, and selected 551M-xdf (I didn't have the 551K-xdf)) and then double clicked the MLHFM. (Yes, I am noob with this) It opens the table which shows values from 124.6 to 1996.10 which indeed looks logarithmic.

And the table looks ok, it doesn't have zeros or strange numbers.

That won't work. You need exact XDF for tunerpro or use WinOLS for which there is 8D0907551K.OLS available on this site with ALL the maps (lucky you) defined in it.

MAF table starts with 0.00 values and goes up from there.

Ok. Could you point me a link that where is this OLS-file? I can't find it with a search, and I didn't have luck with just browsing.

here it is