Audi A4 B7/8E 1.8T BFB - Shaking/Vibration after cold start (Solution)

Greetings to everyone, and a happy new year to all!

Car: Audi A4 B7/8E quattro, 6-speed manual transmission with 1.8T engine, engine code: BFB, manufactured in June 2006, bi-fuel with Zavoli LPG system (year of installation of the LPG system unknown) - total mileage: 205,000 km.

We have the following problem:
The engine starts perfectly, but as soon as the idle speed drops from approximately 1200 to 800 rpm, it starts to shake. This is clearly transmitted to the entire drivetrain. If you keep the RPM above 2000, it's still drivable, but at idle, it feels like it's about to stall. After about 2 minutes, once the coolant temperature reaches 45 degrees or higher, it runs perfectly, whether on gasoline or gas. This only happens during cold starts. When it's slightly warm, it's completely problem-free. Even if we run the auxiliary heater (which is located in the water circuit) for a short time before starting the engine, the problem does not occur.

Since it feels similar to a misfire on only 3 cylinders, I checked the spark plugs – they were all dry, light, and a tan color. After that, I replaced all the ignition coils because one had already failed last year, and we didn't know how old they actually were. (We only bought the car used in the summer of 2016.) However, that didn't improve anything.

I then thought about clogged injectors and added a full tank of fuel with a GL01 gasoline system cleaner – unfortunately, there was no improvement.

The cold/warm compression readings are between 9-12 on all cylinders. I don't know if the cylinder head has hardened valve seats and valve guides. Last year, I had the gas system inspected: a new gas filter and a new engine mapping were installed - it's running great now.

For a long time, there were no errors stored in the control unit, but recently, these errors were found:

Sunday, December 31, 2017, 2:37:09:12617
VCDS Version: RKS 17.8.1 runs on Windows XP x86.


Address 01: Engine Control Module
Label file: DRV\06B-906-018-AVJ.lbl
Control unit part number: 8E0 909 518 AQ. Hardware: 8E0 909 018 B.
Component and/or version: 1.8L R4/5VT G 0010
Encoding: 0011711
Operating number: WSC 32468 126 86989.
VCID: 7F5AF3ED8BEE686F23B-50F4
8 error codes found:

17772 - Ignition control, cylinder 4.
P1364 - 004 - Interruption - Sporadic.
17766 - Ignition control, cylinder 2.
P1358 - 004 - Interruption - Sporadic.
17763 - Ignition control, cylinder 1.
P1355 - 004 - Interruption - Sporadic.
17769 - Ignition control, cylinder 3.
P1361 - 004 - Interruption - Sporadic.
16684 - Misfire detected.
P0300 - 008 - Sporadic.
16685 - Cylinder 1.
P0301 - 008 - Misfire detected - Intermittent.
16686 - Cylinder 2
P0302 - 008 - Misfire detected - Intermittent.
17536 - Bank 1; Mixed adaptation (multiple).
P1128 - 001 - System too lean - Intermittent - Warning light ON.

Readiness: 0000 0000





Unfortunately, my online research didn't yield any helpful results. While there are similar cases, there are no clear solutions available.


This morning, I read the following values, first with the ignition on, and then after the shaking stopped. During this time, I also recorded a CSV log and added markers to indicate the start and end of the shaking. I am trying to upload this file for analysis. Thank you in advance for your help!
'Die Sonne scheint hell, und die Vögel zwitschern fröhlich. Es ist ein wunderschöner Tag, um spazieren zu gehen oder ein Picknick im Park zu machen. Viele Menschen genießen das gute Wetter und verbringen Zeit im Freien.'


Address 01: Engine Control Module (8E0 909 518 AQ)

11:18:38 Block 003: General.
0 /min Motor speed (G28))
0.00 g/s Air mass flow meter (G70)
4.7% throttle valve angle.
0.0 °v.OT Ignition timing angle

11:18:38 Block 033: Lambda regulation.
0.0% Lambda control, Bank 1, Sensor 1.
1.540 V Voltage, Bank 1, Probe 1.



11:18:38 Block 134: Temperatures.
8.0°C Oil temperature.
4.0°C Outdoor temperature.
9.0°C Intake air temperature (G42)
6.0°C Coolant temperature (G62)

'Die Sonne scheint hell, und die Vögel zwitschern fröhlich. Es ist ein wunderschöner Tag, um spazieren zu gehen oder ein Picknick im Park zu machen. Viele Menschen genießen das gute Wetter und verbringen Zeit im Freien.'
'Die Sonne scheint hell, und die Vögel zwitschern fröhlich. Es ist ein wunderschöner Tag, um spazieren zu gehen oder ein Picknick im Park zu machen. Viele Menschen genießen das gute Wetter und verbringen Zeit im Freien.'


Address 01: Engine Control Module (8E0 909 518 AQ)

11:24:45 Block 003: General.
800 /min Motor speed (G2http://shop.dieselschrauber.org/en)
3.86 g/s Air mass flow meter (G70)
2.4% throttle angle.
9.0 ° before top dead center (BTDC) Ignition timing angle.

11:24:45 Block 033: Lambda regulation.
-6.3% Lambda regulation, Bank 1, Sensor 1.
1.500 V Voltage, Bank 1, Probe 1.



11:24:45 Block 134: Temperatures.
19.0°C Oil temperature.
4.0°C Outdoor temperature.
10.0°C Intake air temperature (G42)
45.0°C Coolant temperature (G62)

A common problem with 1.8T engines is the deterioration of the hoses and pipes associated with the crankcase ventilation system. They often crumble just by looking at them, and you can find them located behind the alternator, in the direction of travel.

Another possible culprit is the coolant temperature sensor (G62). Is the value reported by this sensor reasonable in your log? Also, take a look at what it displays when the engine is cold.

If it's the G62, installing the sensor isn't a pleasant task; it's located between the engine and the transmission, hidden near the splash guard. For installation, be sure to secure the retaining clip with a piece of string, or purchase a few extra to have on hand.

Hello,
There's also a diagnostic trouble code (DTC) list that includes misfire codes; take a look at it to see if it indicates anything.
Otherwise, I would also check for the aforementioned corrosion, which can be a bit tricky because you sometimes have to really examine things closely and not just rely on visual inspection.

Best regards,
Thilo

Hello klahaui & Turbinchen,

Thank you for your input.

1. I checked the G62 data - it's perfectly reasonable - no outliers, and the temperature increase is slower and more gradual when the ambient temperature is low.

2. I analyzed the MWB (engine control unit) data and found no recorded instances of misfires during the measurement. However, there must be occasional misfires occurring sporadically, on various cylinders. Cylinders - otherwise they wouldn't be stored in the error memory.

I also performed a Lambda aging test and the readings seem plausible, except for a missing internal resistance on sensor 1. (Lambda sensor 2 shows, for example, 0.2 kOhms, but sensor 1 shows nothing - is this possible/problematic?) However, it is definitely working and providing corrections.

'After the cold start, I sprayed the Kubel housing ventilation with starting fluid, but unfortunately, I didn't notice any significant change. Also, do you really think it's possible that this part could behave differently just because the auxiliary heater was running?' 'Above an operating temperature of 45 degrees Celsius, I no longer experience this problem.'

Another approach: I found two other people online who had the cylinder head removed in order to have new hardened valves and valve seat inserts installed. Apparently, that fixed the problem. The valves had some small cracks. They saw this as the cause of the shaking when the engine was cold.
Do you think this could also be the case for me? I thought that could be ruled out with a compression test, but given the above. Even if the system was cold, a pressure of 9 bar would likely still have been measurable.

Best regards, Alex.

I'm familiar with quite a few 1.8T engines, and I've even owned one myself, so I've heard things related to that. That doesn't mean it doesn't exist.

How does the loosely fitted oil filler cap behave when the engine is running? If it's lightly bouncing, that's okay. If it's being sucked down, you need to check the KGE valve.

Hello,

The misfire detection is clearly visible under load when the engine is running.
I would be cautious about using brake cleaner to troubleshoot issues, as it often doesn't produce the desired results. Try disconnecting the mass airflow sensor (MAF) and see if the engine still vibrates. This way, the MAF won't provide a reference value, and any vacuum leaks shouldn't have as much of an impact.
I haven't encountered any engines from that series with cracked valves, nor have I seen any that run on gas.

Regards,
Thilo.

The loose oil filler cap is rattling around and not being sucked in.

I checked the LMM (likely referring to the Lambda Measurement Module) this morning, and the engine actually didn't sound as rough as usual; it seemed to be running better! Does this mean we should now check all the rubber hoses and look for any potential sources of air leaks?
What I don't understand is the issue with the auxiliary heater. Wouldn't it constantly draw in false air through the leak, not just when the operating temperature is between 0 and 45 degrees? What's different above 45 degrees? Is it true that warm air leaks are okay, but cold air leaks are not?

I can only measure voltage drops under load once I have a new battery for the VCDS laptop - all previous measurements were taken while the vehicle was stationary and plugged into a power outlet.

Hello.

What about the hose/valve/container for the gasoline vapor pre-injection? In addition to the cold start enrichment, the vapors from the fuel tank vent are routed through an activated carbon filter and then introduced into the intake manifold.

Cracks in the plastic/rubber, damage from rodents, loose hose clamps, and other leaks can cause air leaks, leading to a lean fuel mixture and engine misfires.

A possible test would be to temporarily disconnect the hose from the intake manifold, seal the connection on the intake pipe, and close the open end of the hose. No more shaking -> problem found. Then you just need to pinpoint the exact location of the leak .


Regards, Mpire.

I thought I had finally found it – a vacuum leak – discovered in the hose connected to the valve cover breather, which feeds into the intake manifold. However, the following error messages are still appearing:

Wednesday, January 24, 2018, 1:40:02:12617
VCDS, a Windows-based VAG/VAS emulator, runs on Windows 7 x64.
VCDS Version: RKS 17.8.1.0 HEX-V2 CB: 0.4409.4
Data as of: 20171211 DS276.6



Vehicle Identification Number: WAUZZZ8E76A249705
License plate: FB-AQ9
Mileage: 205615. Repair order: 1.


Address 01: Engine Control Module
Label file: DRV\06B-906-018-AVJ.lbl
Control unit part number: 8E0 909 518 AQ. Hardware: 8E0 909 018 B.
Component and/or version: 1.8L R4/5VT G 0010
Encoding: 0011711
Operating number: WSC 32468 126 86989.
VCID: 7F5AF3ED8BEE686F23B-50F4
3 error codes found:

16684 - Misfire detected.
P0300 - 001 - Sporadic.
16685 - Cylinder 1.
P0301 - 001 - Misfire detected - Intermittent.
16686 - Cylinder 2
P0302 - 001 - Misfire detected - Intermittent.

Readiness: 0000 1001


Following that, I wrote a log file under load, including the following values:

'Der schnelle braune Fuchs springt über den faulen Hund.'

Address 01: Engine Control Module (8E0 909 518 AQ)

13:39:49
0 /min Motor speed (G2http://shop.dieselschrauber.org/en)
18.0°C Coolant temperature (G62)
0.0% Lambda regulation, Bank 1.
0.00 g/s Air mass flow meter (G70)
4.7% Throttle valve angle.
985.0 mbar Intake manifold pressure (G71)
0. Sum of skipped event counters.
Cylinder 1 misfire.
Cylinder 2 misfire.
Cylinder 3 misfire.
Misfire, cylinder 4.

Could you please take a look at the file? Does it give any further clues? Thank you!

Mpire - Thank you very much! I checked, but I didn't find anything. I'll just double-check to be sure.

Hi,

"When there are glitches in the log, the lambda control seems to fluctuate a bit; I'm not sure if that's normal to that extent for that engine. Is the fuel supply working correctly? Has anything been modified there?"

Best regards, Rainer.

The fuel supply seems to be the issue, but it doesn't cause problems when using the auxiliary heater.

Unless the auxiliary heater is controlling the fuel pump to access fuel ("priming" the pump).

Generally, these pumps don't have a very long lifespan and can cause all sorts of strange problems (often in conjunction with the fuel filter).
The pump is running constantly, so it has accumulated a lot of mileage!

Otherwise, it still sounds like there might be a leak.
Was the arch above the oil cooler also inspected? It breaks after a few years, and because it's difficult to access the area where it breaks, it's not immediately obvious.

However, a simultaneous interruption of fuel delivery to all cylinders doesn't sound like a vacuum leak to me.

I would tend to think it's related to cable connections in the area of gas systems, engine control units, as well as soldering points for the connections and control of the fuel injectors and ignition systems.

Or perhaps the gas system control unit was briefly without power.

What does the short-term/long-term trim indicate?

However, the triggering errors were not permanent.

Only the misfires remain.

Were the spark plugs removed while the engine was still cold, or later? Was the engine immediately shut off when the error occurred?

Thank you for your support!

No, the triggering errors are not permanent – I think it's more likely that it's a 'legacy' issue from the compression measurement.

The fuel system hasn't been modified. I think the auxiliary heater has a small, separate pump. Since the car runs perfectly well when the auxiliary heater is on, even under various conditions, I'm actually ruling that out as the cause.

I just replaced the spark plugs for safety and swapped the ignition coils. Result: Currently, there are no error codes stored in the engine control unit after the test drive, but the vehicle continues to shake briefly shortly after a cold start, specifically when the elevated idle speed is reduced.

There are 43 misfires on cylinder 1 and two misfires on cylinder 2, and this only occurs between 130 and 172 seconds after a cold start. Okay, so from minute 2 to the end of minute 3, and that's just idling – according to VCDS, the misfires don't increase under load!

From minute 4 onwards, nothing unusual happens anymore. It runs perfectly up to 210 km/h without any issues.

'Regarding the 'Long-Short-Term-Trim - Bank 1 Sensor 1 Idle additive 0.2% / Partial load multiplicative' values, I observed ranges from -0.6% to +2.8% during the drive. These values are also present in the new log. Please let me know if I have logged the wrong data!' Shouldn't they vary more?

Regarding the cable: Does it make sense that the problem always appears at a clearly defined time, and doesn't occur when the auxiliary heater is in use? Surely, whether it's specifically during this cold start phase or not shouldn't matter to the cable or solder joints. What temperature, approximately 50 degrees Celsius, would need to be reached near the engine for the error codes related to cylinders 1 and possibly 2 to disappear?

Conclusion: We traced the error back to the valve seat inserts, which were heavily worn. After replacing them, the problem disappeared.

Thank you again for the solution ideas.

Thank you for the feedback!