cparke

I recently had a worsening too rich fuel trim problem with my 1998 Audi A4 V6 30V car after using fuel cleaner additive. Using more fuel cleaner additive made the too rich condition even worse. I had a check engine light that kept coming back on and used VCDS regularly to monitor Measuring Block 32 and 55 to determine how much 'too rich' the fuel trim was and my idle air learn condition.

Finally, the ECU started to go nuts about the fuel trim (Load Calculation cross-check implausible) and I had to bring the car to a repair shop. While I ultimately found the shop to be disappointing because after 3 days of having at my car (and they can't tell me what exactly they looked at either!) all they could say is that everything appears normal and I should try getting another MAF, even though I had already changed the MAF. I chose instead to take the car away from them. I wound up paying them $192 for a 'computer diagnostic' which supposedly told them everything is normal.

The surprise, however, is the car's too rich fuel trim problem also is now gone. It has been a week now, no check engine light again and measuring block 32 values are in the same place they were last week.

My question here is pretty simple --- what could the shop's computer diagnostic have done to bring sanity to the ECU that my simple VCDS system could not? How do all engine learn values get cleared using VCDS? Is there an adaption procedure or O2 sensor aging that could have been responsible?

I really want to know because I consider my car's fuel trim now to be very sensitive to system changes, and if I have to change a sensor or something in the future and the problem re-occurs, I'd like to know what steps I could try on my own before bringing it back to a shop for a diagnostic and 'reset'.

Thanks for your thoughts.
CP

Andy@Ross-Tech

The effect is identical with VCDS and the factory scan tools. Are you sure they didn't locate and repair a vacuum leak?

cparke

Clearing faults does clear fuel trims, but what about the other learned values used to do the load calculation cross-check and to compute the new fuel trim?

I know for a fact that some learned values (like the group 55 learned idle air ctrl) don't clear with VCDS fault clearing. For that one, I've had to disconnect the battery for 1 hour. Some people on AW have suggested I leave the battery disconnected overnight. Still, I don't think even that clears everything either (some learned values must be kept in non-volatile memory).

The guy spoke of the diagnostic system they used as some $8000 computer that the dealers have (of course, I don't quite believe this rhetoric) which he had to rent from someone, yet there is no report and no log of the diagnostic results he can show me. These people consistently treated me like some kind of fool who has no business understanding how my car is diagnosed and fixed, so you can see why I took my car away from them when they also said they weren't getting anywhere and insisted I change my new MAF for another one sold by them.

I don't know exactly what they did because they won't tell me (or, they don't know themselves what they were doing). I'm sure no parts were changed, because I don't see anything new and they'd be super-foolish not to charge me for them. For vacuum leaks, all he would tell me is they did a visual check and saw nothing wrong. That is consistent with the fact that fuel cleaner additive is what causes this problem in my car. I think the new combustion properties causes the computer to learn some new values which muck up the fuel trim, and when the cleaner is gone, it can't re-learn them right, so even though I clear the fuel trim, it goes back too rich because of the bad underlying learned values used to compute it which I can't reset.

While I dare not ever use fuel cleaner additive again in this car, I really would like to know what I could try next time to mimic whatever the repair shop's diagnostic may have done to fix my ECU, in case there is another way they can get mucked up. I don't think the shop ever disconnected the battery because the clock was still correct and the trip odometer never reset.

Andy@Ross-Tech

Without getting a straight answer from the repair shop, there's no way to know what they did.