[Gasoline] Polo 9N AZQ engine stalls when shifting gears

Hello everyone,

Our 9N with the 1.2L AZQ engine, 47 kW, 3-cylinder, is acting up again. Specifically, when upshifting, he keeps the accelerator pedal pressed down for an unusually long time during the power phase, and even increases the engine speed by about 200 RPM instead of letting it drop.

According to the VAG-Com diagnostic tool, the throttle position sensor (G79) already shows 14.5% at idle. Full throttle corresponds to 75% in this case. For the second sensor, G185, the values range from 7.1% (idle) to 37.3% (full throttle). Shouldn't it say "0-100%" there?

I also cleaned the throttle body for the sake of completeness, but there wasn't much dirt on it. Is it normal for the throttle valve to completely close approximately every 30 seconds (ignition on, engine off, assistant holding the accelerator pedal fully open) and for the throttle body unit to constantly hum as long as the accelerator pedal is pressed?

When running at full throttle (which is 75% on the G79 scale), the throttle valve is reported as 100% open via G187. Second... I have not observed Geber G 188.

Is it just the gas stove that's broken, or are there other problems we need to be aware of?

Address 01 -------------------------------------------------------
Controller: 03E 906 033 L
Component: 1.2l/4V SIMOS3 00HS4726
Coding: 00071
Shop #: WSC 31414
WVWZZZ9NZ4D027xxx VWZ7Z0C7842057
No fault code found.

[img][/img]http://home.arcor.de/rgoebig/TDI/Log.jpg{MARKER}

Have you ever performed a basic Drokla calibration/adaptation, if that applies to your vehicle?

Yes, but it didn't change anything.

Specifically, when upshifting, he keeps the accelerator pedal in the 'acceleration' position for an unusually long time, and even increases the engine speed by about 200 RPM instead of letting it decrease.

Hello Roger,
It's strange, because according to the description, I would have said that the idle switch isn't being recognized correctly, and therefore the thrust cutoff isn't engaging, or isn't engaging early enough.
However, the thrust is detected in the log file.

According to VAG-Com, the throttle position sensor G79 already shows 14.5% at idle. Full throttle here indicates 75%. For the second sensor, G 185, the values range from 7.1% (idle) to 37.3% (full throttle). Shouldn't there be 0-100% there?

Actually, yes, but here I can only give you the direct part number for the accelerator pedal sensor from my Leon:

When the accelerator pedal is not depressed, the output is approximately 0 V; at 100% accelerator pedal depression, the output is about 4.5 V.
The function is: Pedal

'Therefore, there is no offset from the sender. What the MSG calculates based on this needs to be read in the characteristic curve set. The two potentiometers at the accelerator pedal actually work in opposite directions: one shows 0% at idle, while the other shows 100%.' Measure it directly at the pedal.

The fact that the throttle valve opens 100% either indicates that the error is being calculated by the MSG (presumably a control module), or that it's normal, which I highly doubt, given that the values are very unusual.

Is it normal for the throttle valve to completely close approximately every 30 seconds, 6 times in a row (ignition on, engine off, assistant constantly gives full throttle), and for the throttle body unit to constantly hum as long as the accelerator is pressed?
If the engine is not running, a diagnostic test can be performed on the throttle valve. This could be the diagnostic procedure used to check if the throttle valve control is short-circuited, or to verify if the Engine Control Unit (ECU) is actually expecting 100% throttle opening.
'With injectors, there's also a peculiar high-frequency control signal used during the fuel cutoff process. It's so fast that the valve can't actually open, but a current flows that is detected and diagnosed.'

The constant humming sound from the throttle valve is due to the PWM signal that controls it. The frequency is very high, which causes the coil to oscillate.
You only have a problem when the throttle valve no longer makes a humming sound when activated.

Is it just the Gaspoti that's broken, or is there more coming our way?
Just measure the analog output voltages of the potentiometers and the switch for the idle detection, and then we might know more.

Best regards, Jochen.

Were the engine learning values cleared before the throttle body adaptation?

I've experienced this a few times now, where the "clean/adapt" action runs without actually doing anything for...
cat was.

Jochen_145 wrote:

Roger wrote: Specifically, when upshifting, he keeps the accelerator pedal pressed for an unusually long time during the acceleration phase, and even increases the engine speed by about 200 RPM instead of decreasing it. Hello Roger, It's strange, because according to the description, I would have said that the idle switch isn't being recognized correctly, and therefore the thrust cutoff isn't engaging, or isn't engaging early enough. In the log file, the thrust is detected.

Hello Jochen,

I would also say that, especially since the throttle cutoff works normally as always when a gear is engaged. Even when playing with the accelerator while idling, everything is as usual. I haven't found an idle speed control switch in the control unit, is it hidden in the binary values like the brake pedal switch? Where would it be located physically? In the lower part of the control unit? Then I might have to budget for about 284 euros in parts + a gasket... However, I wouldn't be surprised if, for cost reasons, it doesn't have an idle speed switch and the low-speed setting is only recognized through the dual potentiometer.

Quote:

Actually, yes, here I can only give you the direct part number for the accelerator pedal sensor from my Leon: When the accelerator pedal is not depressed, the output is approximately 0 V; at 100% accelerator pedal depression, the output is about 4.5 V. The function is: Pedal[%] = U_ist * 26.65[%/U] "Therefore, there is no offset from the sender. What the MSG calculates based on this needs to be read in the characteristic curve set. The two potentiometers at the accelerator pedal actually work in opposite directions: one shows 0% at idle, while the other shows 100%." Please measure it directly at the pedal.

"My TDI also shows 0-100%. I'll be getting comparison values for both sensors from another Polo 9N tomorrow, and they all have the same reading when running on gasoline with a manual transmission. I won't need to do any more contortions in the footwell for now. However, the connector is firmly seated, which I checked yesterday. A few months ago, the entire dashboard was removed due to climate control motors. Could something have happened then?" AFAIK, the pedal bracket and its wiring are not touched in this process.

Quote:

The fact that the throttle opens 100% indicates either that the error is being calculated by the MSG, or that it is normal, which I don't really believe, because the values are very strange.

It's difficult to assess because the electronic gas system (eGas) already keeps it at full throttle in many operating conditions, for better throttle response and reduced throttling losses. However, the good maximum speed would at least suggest that it works well under full load (whatever that may mean).

Quote:

If the engine is not running, a diagnosis can be performed on the throttle valve. This could be the diagnosis where we check if the throttle control is short-circuited, or if the MSG (Motor Steuergerät - Engine Control Unit) is truly expecting 100% opening.

I don't quite understand. In the actuator test in STG 01, the DK (likely referring to a specific component or data) is not recorded, only the AGR (likely referring to another component or data) and some peripheral information. Do you mean the default setting in a suitable diagnostic tool, or saving channel 000 in the adjustments? In the latter case, it only deletes the learned values for the throttle valve adaptation, as far as I know.

Quote:

The fact that the throttle valve is constantly humming is due to the PWM signal that controls it. The frequency is very high, which causes the coil to oscillate. Only when the throttle valve no longer makes a humming sound when activated do you have a problem

.
Understandable, but theoretically, shouldn't a full activation = 100% duty cycle result in a "rest" state?
The thing is really brutal; when cleaning the DK (spark plug) ad hoc (with the ignition on and control via an assistant), it almost bit off the tip of my finger . Nobody expects that sudden 6x closing... And that thing is really sharp.


@Schraubi-Schlumpf:

What is the correct procedure for throttle valve adaptation? I only deleted the learned values in channel 0 by saving. Are the learning values for the DK adaptation included?

Quote:

@Schraubi-Schlumpf: What is the correct procedure for throttle valve adaptation? I only deleted the learned values in channel 0 by saving. Are the learning values for the DK adaptation included?

"That should work (I don't have a VAG-COM). I've just found that it's best to clear the learned values first and then perform the adaptation. But you've probably already done it that way, if I understand you correctly."

I couldn't find an idle switch in the module, is it hidden within the binary values like the brake pedal switch? Where would it be located physically? In the DK lower section

Hmm.. I'm currently unable to help with that particular matter.
From a hardware perspective, the idle speed control switch in my Leon is integrated into the accelerator pedal sensor and is connected to the engine control unit (ECU) via a 6-pin connector. I guess that gasoline-powered cars, being mass-produced, likely have a similar accelerator pedal.

If the engine is not running, a diagnosis can be performed on the throttle valve. This could be the diagnosis where we check if the throttle control is short-circuited, or if the MSG (Motor Steuergerät - Engine Control Unit) is truly expecting 100% opening.
I don't quite understand. In the actuator test in STG 01, the DK (likely referring to a specific component or data) is not recorded, only the AGR (likely referring to another component or data) and some peripheral information. Do you mean the default setting in a suitable diagnostic tool, or saving channel 000 in the adjustments? In the latter case, it only deletes the learned values of the throttle adaptation, AFAIK.
Perhaps I expressed myself a bit too complicatedly here:
I meant the internal diagnostic functions of the control units, which measure whether there is a short circuit to ground or something similar. The result would be an error log entry.


Understandable, but theoretically, shouldn't there be 'silence' at full activation = duty cycle of 100% (?)
The thing is really brutal; when cleaning the DK (spark plug) ad hoc (with the ignition on and control via an assistant), it almost bit off the tip of my finger . Nobody expects that sudden 6x closing... And that thing is really sharp.

No, 100% PWM, meaning a constant U_bat (battery voltage), is only used in areas where no diagnostics are running, because otherwise you won't be able to detect a short circuit towards U_bat. Therefore, a 100% load request will not result in 100% PWM at the throttle, but rather less. Usually, about 95%.

You're lucky you weren't injured.
The things are closing actively!
This means that the throttle valve is not only brought to the zero position by spring force, but also actively accelerated to the zero position using a 'negative' PWM signal that pulls down to -12V (reversing the current direction).

This allows me to intentionally accelerate and decelerate the throttle. The target position is reached more quickly and maintained with greater precision, thus preventing throttle valve oscillations.
(This realization cost me quite a bit of time during a seemingly simple measurement task.)


Best regards, Jochen.

Jochen_145 wrote:

From a hardware perspective, the idle speed switch in my Leon is part of the accelerator pedal sensor and is connected to the MSG via a 6-pin connector. I guess that the gasoline cars, based on the same part principle, have a similar accelerator pedal.

I just checked the wiring diagram. None are installed... But they did include a dedicated warning LED for the accelerator pedal in the instrument cluster. There's also a clutch pedal switch in the SLP (Steuergerät Lernprogramm - Control Unit Learning Program), but it doesn't specify whether it's installed even without GRA (Gaspedalreinigung - Accelerator Pedal Cleaning). However, the second brake pedal switch explicitly states "only with GRA." So, I need to find the clutch pedal switch on the actual vehicle tonight. Is there a way to...?

Quote:

Perhaps I have expressed myself a bit complicatedly here: I meant the internal diagnostic functions of the control units, which measure whether there is a short circuit to ground or something similar. The result would be an error log entry.

Ah, I understand. Well, to be honest, I would actually prefer an error message.

Quote:

No, 100% PWM, meaning a constant U_bat (battery voltage), is only used in areas where no diagnostics are running, because otherwise you won't be able to detect a short circuit towards U_bat. Therefore, a 100% load request will not directly translate to 100% PWM at the throttle, but will be less. Usually, about 95%.

Again, aaah . That makes perfect sense.

Quote:

You were lucky to have escaped without being injured: The things are closing actively! This means that the throttle valve is not only brought to the zero position by spring force, but also actively accelerated to the zero position using a "negative" PWM signal that draws -12V (reversing the current direction).

Yes, I wrote that thing is brutal. And really fast. Almost like a small guillotine . Luckily, I was quite alert because my wife was on the pedal *ahem* .

Thank you for the explanations so far. Now I'm waiting for the benchmark results, and in the meantime, I'll take another, more detailed look at the car based on the SLP (Standard Load Profile).

"I've also found another check point. When you remove the air filter housing, which is integrated into the engine cover, there should be two hoses leading to the engine. However, I only saw the one on the top, connected to the catalytic converter. There was definitely no second hose on the battery side. I'll need to take a closer look at that tonight. If the vacuum control system was disrupted due to a detached or missing hose, that could explain the striking difference between the engine revving at idle and the sudden drop in power when releasing the throttle."
http://pimpowski.pi.funpic.de/Motorabdeckung_abbauen.htm

Roger wrote:

.... because my wife was on the gas pedal *ahem* .

No, nothing . Of course, I then tried it again myself, attaching the pedal to the bottom plate with nails . Same effect, every 30 seconds the DK closes exactly 6 times .

I just found it.

The suspicion about the clutch pedal switch was confirmed - it was defective. I removed it, and it started working normally again. That was the cheapest part that could have broken (11 €).
Oh yeah, and the second hose was also disconnected. However, nothing changed during the test drive.

Great.

It's great that it was such an affordable item.
I would have expected more common parts between the gasoline and diesel versions.

Best regards, Jochen.

Jochen_145 wrote:

Great, It's great that it was such an affordable item. I would have expected more common parts between the gasoline and diesel versions. Hi Jochen

Yes, it has already been ordered .

It's quite funny that the switch broke while it was in the "on" position. Apparently, it even tried to engage its anti-jerk damping system while the clutch was being disengaged... With the described effect.

For the sake of completeness:
The new switch has now been working as expected for > 1 week .

Roger wrote:

For the sake of completeness: The new switch has now been working as expected for > 1 week .

Hi Roger,

okay, great

By the way: I once read in a forum for the 9N model that the pedal switches are "one-way devices," meaning that their housings are likely to break if you try to disassemble them.
Can you confirm this from your experience (or, better yet, disprove it)?