2.0 FSI Poor idle when cold

[saabman]

I gave up on the Nox sensor - I did try a second hand unit but that didn’t make any difference so just left it.

Ok when cold but gets bad when hot sounds like something leaking as it warms up. You may not notice it when doing a smoke test as the engine would be cold. I’m not sure how you would test did that though maybe squirting carbybcleaner or Aerostart around the joints in the manifold when it’s hot and seeing if you can detect a change.

I started translating the funktionsrahmen it’s on my GitHub page I’ll grab the link shortly and post it. I just used google translate to do it so it’s a bit rough but gets the idea across. I’ve only done a few chapters but if you have a look through the index and let me know what chapeters look interesting I’ll try and get them done. Or you can have a go and add them to the repository.

The MAP sensor can fill with oil it’s on the lower front passenger side of the intake manifold - a single torx screw from memory holds it in - might be worth haveing a look at that as well. Unfortunately I don’t have ready access to the car my daughter lives about 2 hours away so don’t see her unless the car breaks down

[saabman]

Here’s the link to my translation
GitHub - saabman/MED9.5.1: Translation of the Bosch MED9.5.1 Funktionsrahmen to English

About a 1/3 of the way down the page is button that says “show code” click on that and you should see the list of chapters I’ve done so far.

Also keep in mind the throttle position is set not only by peddle position but whole range of other parameters like brake booster vacuum, temperature, etc. so some other seeming unrelated issue might be causing the throttle position to change which would alter Manifold vacuum and consequent MAF

[maro]

Thanks for the link!
Actually i have graphs somewhere proving that two sharp accelerations from idle resulted in roughly same throttle valve opening angle ~99% and same intake manifold vaccuum ~960mBar for same amount of time but with different mass airflow. One slowly building with hesitating effect and second quickly rising with space shuttle start effect.

What bugs my mind where ECU finds input data for calculating mass airflow when MAP (with integrated temperature sensor), temperature sensor after air filter and brake booster pressure sensor are disconnected.

Maybe it is just some fallback hardcoded map in ecu? But then it does not explain how the value could "predict" engine acceleration. It must be based on some other sensor's output giving real data not just some simulated vectors from Bosch test bench.

EDIT:
I found MAP covered with oil year ago. Since then i clean it on regular basis.

[saabman]

If it knows throttle position, engine rpm, injector timing and o2 sensor readings it could take a pretty good stab at calculating MAF.

Though I would guess it would have a “basic” map of throttle position x rpm = injection time

I wanted to try and extract a copy of the maps from the ECU and see what it looks like. But I never realy got to the bottom of what gear I needed to download the data ( with out paying a fortune) I also never found the template file that identifies the various maps for the MED9.5 ECU.

What happens with the short term fuel trim when the hesitation occurs

[maro]

Well i cannot make my mind about STFT.
I know they should be switching around zero, but which amplitude, frequency in which driving conditions.

This is mostly during hesitation. Engine braking, then full throttle. Is this right or not? LTFT are usually in +2 - +5



Also there is often something like offset between banks. Sometimes B1 is richer than B2 and later it is vice versa.
At least the envelope from lambdas is mostly identical. So i assume they read consistently = both are working OK.


Yesterday i performed exhaust backpressure test with cylinder compression kit screwed into O2 sensor hole in exhaust collector and needle did not even move. Gauge has resolution of 6psi per division. I red somewhere that 7psi should be pretty clogged exhaust. I know, resolution is not that good but i watched the gauge closely while other person was revving the engine and nothing.

I am planning to do regular compression cylinder test on the weekend.

Also i found helpful video of russian guy disassembling the 2.0 FSI engine. One should get access to intake ports by pulling off lower intake manifold WITHOUT pulling off the injectors.
?? ??? ??? ???????? ??????? ???????? ???????????? 2.0 FSI - YouTube

Here are the lower manifold with injectors attached to it - i guess they are fixed in lower manifold with blue o-rings as can be seen on the video.


https://i.ebayimg.com/images/g/tE8AA...UG/s-l1600.jpg
https://www.google.com/url?sa=t&rct=...DgidaO97F1Xd0Z

[saabman]

In the BLX engine the injectors go through the lower manifold so to remove the manifold means pulling the injectors out of the head. You may be able to leave them on the manifold but its the removing from the head Im not keen on.

A clear Cat should only give a about 1psi of back pressure so if your not getting any reading I would suggest that is ok.

The difference in your STFT between bank 1 and 2 seems to hint at a slight manifold leak on bank 2 - notice the variation is greater at low throttle opening and becomes less as the throttle opens up.

[maro]

Exactly. I do not want to pull injectors from head together with lower intake manifold. Otherwise i would have to renew teflon sealing on injector tip preventing gases from leaving combustion chamber. Or nightmarish scenario when injector splits in half

I understood that injector SHOULD stay in head when pulling manifold but sometimes some of them pop together with manifold because o-rings that seals injector head with fuel rail get cooked.

I did compression test and result is excellent. 180psi on 3 cylinders (one cylinder 175 and it needed 2 more cranks to reach it but i think thats ok)
Workshop manual says new engine 13-16bar =~188-232psi. So i have nearly new engine, not bad for 220k km

I confirmed on higway, hesitation is connected with air supply. Top value of air mass and time to reach it during WOT decides on acceleration sharpness.
Greatest acceleration on 6th gear which i have experienced (and is sufficient for me) is somewhere near 70g/s developed in fraction of second. According to Fuel Trim Info - Ross-Tech Wiki
i should have seen peak airflow 0.80*150(horsepower)=120g/s close to sea level. I am not even close to these values

I also logged Injection Timing (ms) and Ignition - Timing Angle (°BTDC) but i don't know how to interpret the values, yet.

[drbluetongue]

I gave up on the Nox sensor - I did try a second hand unit but that didn’t make any difference so just left it.

Ok when cold but gets bad when hot sounds like something leaking as it warms up. You may not notice it when doing a smoke test as the engine would be cold. I’m not sure how you would test did that though maybe squirting carbybcleaner or Aerostart around the joints in the manifold when it’s hot and seeing if you can detect a change.

I started translating the funktionsrahmen it’s on my GitHub page I’ll grab the link shortly and post it. I just used google translate to do it so it’s a bit rough but gets the idea across. I’ve only done a few chapters but if you have a look through the index and let me know what chapeters look interesting I’ll try and get them done. Or you can have a go and add them to the repository.

The MAP sensor can fill with oil it’s on the lower front passenger side of the intake manifold - a single torx screw from memory holds it in - might be worth haveing a look at that as well. Unfortunately I don’t have ready access to the car my daughter lives about 2 hours away so don’t see her unless the car breaks down

Hey mate,

I have spent the last year or so trying to figure out how the NOx/stratified system works on our cars (mine is the 1.6 FSI, but the systems are functionally identical).

I'm the author of the Tunerpro definition file for the 1.6 FSI on Nefmoto forums, and I'm happy to help you code out the NOx system or adjust it so you don't need to replace the sensor (and still keep the stratified mode).

To answer the questions around retarded timing on startup, this is the catalyst heating system. Easy enough to disable as well (I have to reduce noise and roughness on startup)

I've studied the FR quite extensively and can help with any questions you have to - let me know.

[drbluetongue]

Thanks for the link!
Actually i have graphs somewhere proving that two sharp accelerations from idle resulted in roughly same throttle valve opening angle ~99% and same intake manifold vaccuum ~960mBar for same amount of time but with different mass airflow. One slowly building with hesitating effect and second quickly rising with space shuttle start effect.

What bugs my mind where ECU finds input data for calculating mass airflow when MAP (with integrated temperature sensor), temperature sensor after air filter and brake booster pressure sensor are disconnected.

Maybe it is just some fallback hardcoded map in ecu? But then it does not explain how the value could "predict" engine acceleration. It must be based on some other sensor's output giving real data not just some simulated vectors from Bosch test bench.

EDIT:
I found MAP covered with oil year ago. Since then i clean it on regular basis.

The cause of the vastly different airflows I've found are due to knock - if your engine experiences any knock retard (they all do with a factory tune running so lean!) it switches maps to a different cam timing map which can account for this (map KFNWKRE). Log knock retard on VCDS and compare

[maro]

Hi drbluetongue, first of all thank you for joining and replying. Nowadays it is really rare to see someone share his knowledge. People on this forum proved to be an exception

As for high idle after cold startup, i confirmed the ECU switches to different mode (010000001). In my case, engine misfires like hell on 1st cylinder (~80 misfires) and few mis on second cyl but only in warm months. In winter it is clean high idle without misfires, strange ha? After that it stays in homogeneous mode (000000001) during entire drive.
edit: today i started car in 10°C and high idle was without single misfire, mode was 000000001 entire time. So it is definitely temperature dependent.
x?xxxxxx variant is not mentioned in VCDS label but it should be some form of double injection for rapid catalyst heating as you mentioned (maybe homogeneous split? Stratified CAT heating should be ?xxxxxxx).

Switching map in runtime perfectly supports my observation. I suspected either carbon buildup or ECU playing some game with me. Generally worse acceleration (airflow) after warm up but not always. 4 bad accelerations 2 good even from low RPMs and so on. ECU compensating for something making much more sense.

I've already logged couple of hour test drive for knocks and it knocks only during high idle startup. Even idle after warmup is perfect with 0,8l/h (PCV is OK i guess).

Which MVB are you talking about when you mention

knock retard

- ignition, timing angle [ °BTDC ]
- timing retardation, cylinder 1-4 [ °KW ]
- camshaft adjustment, intake B1 [ °KW ]
- phase position, bank 1 intake
- injection timing [ ms ]

I am very eager to "fool" NOX subsystem but preferably after solving this damned hesitation.

Thanks