Question for those of you who understand how to interpret fuel trims

[Turbo510]

For as long as I can remember, my 3.0 has had somewhat of a 'turbo lag' effect when accelerating, but of course without a turbo. In other words, during acceleration, it will pull harder if the throttle is lifted slightly, or feels stronger after several seconds of fairly constant throttle/RPM. The delayed torque seems to me the delay of the Oxygen sensor response to increase the injection time. It just feels like a mixture issue. All O2 sensors were replaced a couple of years ago, and no O2 fault codes. Fuel filter is due to be changed. What would you try?- besides junking the thing that is; hat tip to EuroxS4

[Mike Tries]

Hahaha what's wrong with the 3.0?

The Nefarious Motorsports forum has a lot of discussion about ECU tuning. Have you ever lurked around there? You'll see a lot of discussion about fuel trims, as well as the apparently 20 other ways a fuel inector can switch on and off in synchronization with the rhythm of an engine. I don't know which ECU you have, but you might learn a bit by logging things and seeing what it looks like on a graph. Logging things is always a good idea. It's also a fun topic for discussion at parties.

[Turbo510]

That's a good idea. The only time I logged anything was to plot the gm/S mass airflow vs. RPM. And thanks for responding, took over 170 'looks' to get your suggestions.

[EuroxS4]

Probably something to do with ignition timing or timing advance of the cams on that boat anchor of an engine. What kind of fuel are you using? Super or ultra? I would look at fuel trim values MVB 32. Are you sure your not comparing it to the A5? It's night and day. You can also look at MVB for phasers and log to see what they are doing. Depending on miles it may be a good idea to check injector duty cycle and possibly get them flow matched. Then again MVB 32 will tell you all that. The norm is +/-5% tops anything past means something is going on.

[Turbo510]

Thanks, I've thought about removing the injectors for service due to miles/age. My old B5 2.8 had an injector fail on me, still clicked like normal.

[EuroxS4]

Thanks, I've thought about removing the injectors for service due to miles/age. My old B5 2.8 had an injector fail on me, still clicked like normal.

Best way to check an injector is on a flow bench. Output test only tells you it's opening and closing it tells you nothing else about how its operating. I'd be curious to see if its a flow issue or fuel supply issue.

[walky_talky20]

during acceleration, it will pull harder if the throttle is lifted slightly,

Ooh. I have had this exact symptom on my 2001 1.8T for the last...forever.

I documented in a thread just here:
https://www.audizine.com/forum/showt...s)-Clogged-cat

Pertinent quote from (*oh my gosh*) 12 years ago:

After holding at full throttle for a bit, letting OFF the gas will cause the car to *accelerate* noticeably. I feel like this is significant.

Getting the throttle held just right is super tricky. But when I get it, the car *moves*. It's incredibly difficult - I have to let off enough to get the timing back, but not *so* much that I lose all the boost.

Anyway, I changed a bunch of stuff. Some in an effort to fix this issue, but most of it because it needed to changed for another reason anyway.
- Installed a High Flow Labree Cat
- Installed 2.5" exhaust
- Installed another OEM K03s turbocharger
- Changed Primary wideband O2 sensor.
- Changed Fuel Pump
- Changed plugs and coils
- Changed Exhaust valves and lifters
- Changed Timing belt
- Changed head gasket


I feel like I tracked the problem to being timing related. Both the logs and the butt-dyno say that the "acceleration" I feel when letting off the gas pedal is the ECU adding timing. More accurately it is reducing the amount of 'timing pull' - and quite a noticeable amount of it.

I do know that the engine is at least a little bit down on compression, although I have not tested how much.

My final thought as to the root cause of my timing pull is aged knock sensors that have become oversensitive. This is a known failure mode in some types of knock sensors. It is very common on late 90's Subaru 2.5 4-cam engines, for example. I went so far as to purchase a new pair of OEM knock sensors. Twice! (I actually forgot I had them and bought them again). Regretfully, I *still* have not installed them to see if this will fix the issue. Perhaps this is the year. I'm sorry this post was not helpful. I hope you have better luck than I have so far.

[Turbo510]

That's a lot of useful information, thanks. So watching the knock sensor activity during acceleration seems worthwhile. That [could] explain a lean-trim code on just one bank, which confounded me.

[walky_talky20]

I never had any mixture codes, myself. Trims always looked quite good, MAF values looked good, but Timing Pull always seemed pretty high when chipped.

I forgot to note that I'm APR 93 chipped and I can switch back to stock. The symptom was definitely still present in stock mode, but not nearly as drastic.

[Gin+]

Logging would likely tell the tail on both of these scenarios. WOT doesn't rely on the 02 sensor for feedback (open loop).

It wouldn't surprise me if APR has targeted timing higher than achievable and is just relying on the knock sensors to pull it back out. With premium fuel, stock tune, and on a stock motor I didn't see any timing pull when getting a baseline log.

[walky_talky20]

WOT is indeed in closed loop on 2001+ 1.8T. Wideband is used the entire time and chases the target AFR.

Some timing pull is expected, but mine feels pretty excessive. Certainly the APR timing map is aggressive, but I think the sensors are also oversensitive and it cannot ride the edge like it is meant to. My brother's 2000 feels *way* faster with no messing about with the pedal.

[Gin+]

WOT is indeed in closed loop on 2001+ 1.8T. Wideband is used the entire time and chases the target AFR

We'll have to disagree on that one. That said, I work in IT so I'm used to being wrong

For the sake of conversation, techtonics tuning used to sell a knock sensor spacer for the old VW motors to help dampen excess noise. IIRC, it was a vw motorsport trick as well.

[EuroxS4]

We'll have to disagree on that one. That said, I work in IT so I'm used to being wrong

For the sake of conversation, techtonics tuning used to sell a knock sensor spacer for the old VW motors to help dampen excess noise. IIRC, it was a vw motorsport trick as well.

Your wrong!!!Open loop uses a fuel map to adjust your mixture. Closed loop uses the o2 sensor always has in most if not al VAG vehicles. Spacer for the knock sensor was to maximize timing advance. In those motorsports engines some knock was acceptable. But the main function of the spacer was to maximize timing advance without compromise. But you also have to remember in motorsports most of those engines get torn down quite often so even if they had knock it wasn't catastrophic. The same can be achieved if you mess with the VE tables in the ecu without the need for spacer however at that level you better know what the fuck your doing or your going to have very bad bad and some melted pistons or destroyed engine.

[Gin+]

But is it closed loop at wide open throttle? Once you request richer than stoic the fuel trim calculations stop in the logs and I perceive it's MAF input from there even with the wideband ecus. Point being, just trying to help the OP not go down the wrong rabbit hole.

And yes, best to put on your det cans before you start numbing the knock sensors.

[walky_talky20]

Yes. Definitely still trims at WOT. The wideband works perfectly at all reasonable AFR's, so there is no reason to ignore it just because the target is no longer Stoich. It chases the Target all the way.

Example from my 2001 1.8T AWM, Wide open from 3k RPM:

[EuroxS4]

But is it closed loop at wide open throttle? Once you request richer than stoic the fuel trim calculations stop in the logs and I perceive it's MAF input from there even with the wideband ecus. Point being, just trying to help the OP not go down the wrong rabbit hole.

And yes, best to put on your det cans before you start numbing the knock sensors.

Closed loop is only present after vehicle is up to operating temp. Open loop is used on cold start till its warmed up IIRC. However in open loop on this car I believe fuel is being calculated by maf input and load calculation.

[Gin+]

Hmm, well I just looked through my old logs as well and perhaps you've convinced me my thinking was incorrect. Years of 80's cars and the all the ME7.1 (2.7t) tuning references burned it into my head I suppose.

[EuroxS4]

Hmm, well I just looked through my old logs as well and perhaps you've convinced me my thinking was incorrect. Years of 80's cars and the all the ME7.1 (2.7t) tuning references burned it into my head I suppose.

All is well no worries, that's why we are here to grow and learn and teach a few tricks as well.