BBY Error 17584 Lambda Sensor after Catalytic Converter

Hello everyone,

My Polo (BBY, 1.4 16V with 75hp, year of manufacture 2002, automatic) is currently giving me problems with some strange behavior. For weeks everything is fine, then at some point the emissions warning light comes on. There is no correlation to the current driving behavior: Sometimes it happens when simply driving with cruise control, sometimes at a traffic light, or when accelerating hard. Error is always the same (from the previous occurrence, the last one was today):

Code:

Tuesday, 03 May 2016, 19:23:47:56647 VCDS Version: RKS 15.7.4 runs on Windows 8.1 x64 http://shop.dieselschrauber.de Address 01: Engine Electronics       Label file: DRV\036-906-034-APE.lbl Part number: 036 906 034 DE Part and/or Version: MARELLI 4MV G 4442 Encoding: 00073 Operating Number: WSC 00788 VCID: 58B9D3D7B8CE2CFE66-4B4A 1 Error found: 17584 - Bank 1; Lambda correction behind the catalyst P1176 - 35-00 - Limit reached

"This means the lambda sensor is located near the catalytic converter. I tested it today according to the description in the RLF: The car was warm, and the measurement block 30 showed a value of 110 in field 2. The initial setting in display group 36, apply the brake and observe. Initially, "Test one" appeared, followed by "B1-S2 OK", exactly as it should be. The displayed voltage also fluctuated - it was in the range of 0.7...0.8V (I didn't log this, but it was definitely outside the problem range and fluctuated as it should)." I haven't further investigated the Lambda sonde heater, as the target value has been reached in measurement block 30.

Will someone understand this? Currently, I'm deleting the error after it appears and then I have peace for a certain period (2-6 weeks). The car drives perfectly normally, with no noticeable impairments. What I want to avoid is the complicated (the car is 14 years old and everything is probably securely rusted) replacement of expensive parts like the lambda sensor based on suspicion.

The software on the MSG is very old, however, the car hasn't been seen by a dealer for ages. There are several indicators online that the older SW vehicles were not as reliable as they were supposed to be, and that many have been running with a constantly illuminated check engine light for years.

According to the RLF, the aforementioned error can also be caused by outside air in the intake area. I had the car checked one and a half years ago after purchase, so I can essentially rule out leaks. At that time, I made the car operable again after a thorough (even two) cleaning of the throttle body. I also re-programmed the throttle body at that time. I have now repeated this, as the car has been repeatedly without power in the meantime (battery problems, now has a new battery). However, this had no effect. Furthermore, I suspect that problems with secondary air often lead to this error, and that this is also possible under reproducible conditions. After the previous repair, there was a year with no errors in the MSG (this is not entirely accurate, as a spark plug died in the spring, but that is something else), the above error first appeared in November (one or two weeks after the first power outage due to the dead battery).

The fuel consumption for this engine is quite high (see the Spritmonitor link in my profile), however, it's connected to a very old, 4-speed automatic transmission, and the car is used for driving in hilly terrain (Sauerland) and as a short-distance vehicle, as well as for transporting children (very short distance). Do I still consider it "within the scope"?

Best regards,

Jan

Hello Jan,

Current software version is 5084, Your car has 4442. The update was also due to problems with the lambda sensor, and it is also possible that this sensor has failed.

Possible causes I am aware of are only air leaks and probes.

Best regards, Rainer

Hi,

Updating is not quite straightforward... it would first require convincing the Friendly of the necessity. I don't believe that he will do it without conducting his own investigations/repair attempts, which would likely be quite complex depending on the skills of the personnel.

I could certainly take a few trips to inspect the condition of the Lambda sondes... What would be the most sensible way to do this, in order to actually be able to draw conclusions from it? Or does it also lack the same relevance as the test according to RLF?

Best regards,

Jan

Hello Jan,

I would also like to examine the additive and multiplicative factors of the Lambda control, is the coolant temperature in the engine control unit plausible? How many kilometers has the engine been used for?

See also:
/viewtopic.php?t=21836

Best regards, Rainer

Hi Rainer,

I will do that (tests probably on the weekend)... the linked thread and the link within it are informative in this regard. I need to familiarize myself with this gasoline-powered engine first.

Quote:

Coolant temperature in the engine control unit is plausible?

I checked this on the weekend, as the lambda sensor test requires at least 80 degrees. After about 30 km of driving (with a 2/3 of the journey pause of about 30 minutes) and a nearly ten-minute break, the MSG showed 88 degrees. This initially seems plausible. After a day of inactivity, the temperature corresponded to the outside temperature, plus or minus one or two degrees.

Quote:

How many kilometers has the engine been used for?

If no one has recorded the mileage (but it doesn't seem like they have), 126,000 km in 14 years.

Best regards,

Jan

Hello,

...it happened again yesterday... after about 70 km (35 km Hagen to Dortmund and back). On the return journey, the check engine light came on while we were driving at a leisurely 70 km/h and using cruise control going downhill. Before that, there were 20 km on the highway, sometimes even quite fast (I let the BBY have a bit of fun, he even reached 160 km/h for a short time and could easily maintain 140 km/h on inclines (A45 towards the south, the hill behind the Ruhr bridge up to the speed limit) - so, in principle, everything should be okay).

Therefore, I really want to start logging, especially regarding the overall functionality. The measurement blocks are quite scattered, and in reality, one would need to log a lot, which isn't possible all at once. MWB 32 would be the additive/multiplicative values (which refer to the sensor before the cat, not the one after). Then there was also an MWB, where both lambda values are present, and in that case, one would also need a statement about the operating condition (RPM, temperature, load?).

What type of MWB combination would be recommended for a meaningful log in such cases? And how should the test drive look like?

Best regards,

Jan

Problems under minimal load seem to indicate MSG SW.

If nothing / very little is injected, the Lambda cannot be regulated. If the MSG still attempts to do so, this could result in an error message. The mentioned MSG SW update seems to be heading in that direction.

Alternatively, small leaks in the intake path should also be clearly visible in this load range. Have you simply disabled the intake passage while idling with start pilot or similar to see if the engine's running changes (Of course, not by spraying directly into the normal intake hose, but around it)?

Greetings
Guste

Quote:

Then there was also an MWB, where both Lambda values are present, and for that, one would also need a statement about the operating condition (speed, temperature, load?).

Okay!

Best regards, Rainer

Hi,

then I'll try to make a suitable log over the weekend, if I have time.

Quote:

Problems at minimal load seem to point to MSG SW.

I am not 100% sure, as I said, whether it actually always occurred under low load. The very first time, it probably happened when going uphill on an Autobahn, which would have resulted in a much higher load.

Quote:

The mentioned MSG SW update seems to be heading in that direction.

The question is, who is responsible for in terms of vehicle valuation, without immediately issuing a repair order.

Faulty Air: I analyzed everything after purchasing it about a year and a half ago (but without the start pilot), as there were also problems at that time. However, the problem at that time was due to the dirty throttle valve. I had replaced the seal underneath it at that time, and after that, everything also worked again for about a year, before this problem started to appear. Since no one had touched it and there were no relevant events, it should still be like that. I will check it anyway.

What I definitely do: I replace the air filter. The current one I had was installed about a year and a half ago, and had about 7,000 km on it, but no one knows exactly how old it really is. Maybe that's enough to ensure adequate intake for other pressure conditions.

Otherwise, there will soon be a Lambda sonde log... let's see if anything can be derived from it (the sondes are probably as old as the car...).

Best regards,

Jan

Hello everyone,

The car helps with . I just took a long drive and even caused the check engine light to come on. This (reproducible!) happened when I accelerated hard on a long, steep hill (the hill is quite steep, so you stay at full throttle from 50 km/h to the allowed 100 km/h for a long time). Then the engine would rev up... typically around 80-90 km/h, which corresponds to 3/4 full throttle in third gear.

It was, however, a different kind of error than the well-known one:

Code:

1 Error found: 16524 - Bank 1 Probe 2 P0140 - 35-00 - No Activity

Subsequently, I repeated the above-mentioned test of the probe on a parking lot (which had worked perfectly recently) - the test failed. Although the displayed voltage was within the range of 0-1V, it only fluctuated "downward" around 0.1V. Just recently, she also managed to get into the top tier - I suspect that the test is exactly what triggers this.

In MWB 41, the heater for the sensor after the "CAT" setting is shown as "ON", with a resistance of 7-9 Ohms (while in a brief idle state, it increases from 7 to 9 Ohms, while the sensor 1 decreases from 70 to 80 Ohms, but this one is also much hotter because it is very close to the motor).

According to RLF, the above error indicates that the heating system and the sensor cable should be checked.

Here are my three different logs - they are different because I wasn't sure about the effectiveness. The full-throttle runs are recorded in the logs that start with "028" (easily visible by the increasing RPM exceeding 5000), while the log that starts with "36" corresponds to the more leisurely drive home through the mountains (with a lot of throttle).

Does anyone have any ideas?

If I have understood the setup correctly (based on a photo from the Poloforum, but it could also be a different engine), electrical measurements on the sensor after the catalytic converter are quite complicated because the connection box is somewhere hidden under the plastic trim on the underbody, so that it is not really easy to measure without a lift.

Addendum: Or could all of this be explained by the ancient software of the MSG?

Best regards,

Jan

Hello Jan,

Quote:

I just took a long drive and even managed to trigger the check engine light. This (reproducible!) happened when I was accelerating hard up a long hill (the hill is quite steep, so you stay at full throttle for a long time, from 50 km/h to the allowed 100 km/h). At this point, the engine would typically reach... approximately 80-90 km/h, which corresponds to 3/4 full throttle in third gear.

in LOG-01-028-031-032.CSV?

If so: Too bad, no idea.
If no: Is the fuel pressure and fuel supply sufficient? Not that prolonged high-speed running simply causes the lambda values/voltages to deviate, because not enough fuel can be injected.

Best regards, Rainer

Hi,

Quote:

in LOG-01-028-031-032.CSV? If so: Too bad, no idea.

Yes, also in LOG-01-028-033-037.CSV.

What surprises me in LOG-01-028-031-032.CSV: To achieve a timestamp of 490 (unfortunately, the sampling rate is so low, only 3 blocks), I was driving at full throttle and with the kickdown engaged. At 493.47, the log entry appeared just before the gear shift (the red zone at BBY starts at 5500 and the automatic transmission shifts precisely at 5500 with kickdown), and neither there nor in the preceding entries was the engine load 100%. On the contrary... it was only 70%. What does "load" mean in this context? I had previously assumed that full throttle always meant full load... however, the maximum in the entire log is only 72.6%. The minimum is 6.9, which is also a bit surprising, as I have often driven in lower gears on steeper inclines (although it is possible that the automatic transmission has compensated for this at the lower speeds).

EDIT: Just checked in the RLF... for Group 28, the "Load" field explicitly states 50-70%, while most other load fields go up to 100%. It could be related to the actual purpose of this field (knock control) - I was just using it because it has speed, load, and coolant temperature. Therefore, the limitation could also be due to the choice of the field itself, even though it seems strange if different interpretations of "Load" are used in different MWBs. "However, it might also mean that the knock-rule indicator only works when the load is in that area..."

EDIT 2: EDIT 1 is nonsense... in group 37, the load is also present and corresponds to the load from group 28 in the second log, approximately. So, the load is really only at a maximum of 72%... whatever the reason may be.

Quote:

Is the fuel pressure and fuel supply sufficient? Not that prolonged high-speed running simply causes the lambda values/voltages to deviate, because not enough fuel can be injected.

Of course, I could explain why the alarm appeared at full load, but shouldn't we also be able to see that in the values of Probe 1 (BEFORE the valve)? The issues relate to Probe 2 (BEHIND the Kat) - at least, these issues are being displayed as errors.

Even though I don't really have experience with gasoline engines, the values of sensor 1 (interpreted according to RLF) actually look reasonably okay. What do you think about the fact that the additive value of -8 in LOG-01-028-031-032.CSV is quite close to the limit of -10?

Best regards,

Jan

Quote:

Natürlich könnte ich erklären, warum die Anzeige gerade bei Vollast auftrat, aber müsste man das nicht auch anhand der Werte der Sonde 1 (VOR dem Kat) überprüfen können?

Yes, I think so too, that's why it would only have been "No" in the above case.

Hi,

Okay, so the gaze continues to focus on Probe 2 and its deployment. Are there any recent vulnerabilities that should be considered?

And: What exactly does "engine load" mean, and why doesn't it exceed 72%?

Best regards,

Jan

Hi,

There's also a TPI (2011320) issue, as Rainer already mentioned: Update the M-Stg.

Hi,

Then, I'll first check if I can find a friendly neighbor nearby who is willing to do this, without attempting any further repairs. Then there might be usable data from the lambda sensor.

I randomly spoke with someone today who drives an A2 with the exact same engine and has a very similar problem. That means a new lambda sensor is installed every two years... it sounds like the relevant workshop doesn't know about this update, or they don't want to jeopardize their good business model.

Best regards,

Jan