Lupo 3L: No Gears / Troubleshooting

Hello,

After the little Lupo took me 10,000 km across the country, it's now acting up quite a bit. Since the general problem seems to be a transmission issue, I'm posting here instead of in the electrical sub-forum.

Vehicle
Lupo 3L, model year 2003 (engine code: AYZ, gearbox code: FLE), automatic transmission, direct-shift gearbox.

Problem development
If you don't feel like reading all of this, please skip to the next heading: "Status of Error Diagnosis."

After approximately 30km of driving, the car stopped shifting gears. After waiting for almost an hour for the ADAC (German roadside assistance), the car started working again. A diagnostic scan at VW (which is, incidentally, free with VCDS) revealed nothing (the error memory was empty). I then decided to continue driving, but after another approximately 40km, the problem reappeared. In the evening, I drove the car about 70km to my garage, but I drove it in a way that minimized gear changes and managed to get there under my own power. The error memory was empty again.

I then checked the hydraulic oil level, the tightness of the gear selector, the clutch cylinder, and the hydraulic unit – everything seemed normal. I had removed the clutch cylinder, reinstalled it, and adjusted it to the same voltage as before (around 1.9V), but the transmission still wouldn't engage gears. I performed the basic adjustment (which, for some unknown reason, was shortened, but it did complete without any errors). After that, the Lupo started shifting again.

Next test drive, after approximately 60 km, the same problem occurred. This time, there was an error code stored in the system.

Code:

Adresse 02: Getriebe       Labeldatei: RKS\6N0-927-735.lbl    Steuergerät-Teilenummer: 6N0 927 735 H   Bauteil und/oder Version: DS085 Getriebe      1390                  Codierung: 00000             Betriebsnummer: WSC 00000                       VCID: 3D7B54D4376A8FB 2 Fehler gefunden: 00263 - Getriebe             27-10 - unplausibles Signal - Sporadisch 00511 - Wählhebel (E313)             49-10 - keine Kommunikation - Sporadisch

I had to fiddle with it to get the car started, then the next trip ended after about 5km - the gears wouldn't engage again. The car was towed, but the tow rope broke after a while, so the vehicle was left in an alley. A few days later, it was picked up again. It managed to move the first few kilometers on its own, then it was towed. Now it's parked in front of my door.

Error diagnosis status
The issue with the selector lever has not reappeared. However, since the last time the system froze, the following information is permanently stored and cannot be deleted:

Code:

Adresse 02: Getriebe       Labeldatei: RKS\6N0-927-735.lbl    Steuergerät-Teilenummer: 6N0 927 735 H   Bauteil und/oder Version: DS085 Getriebe      1390                  Codierung: 00000             Betriebsnummer: WSC 00000                        VCID: 3D7B54D4376A8F1 1 Fehler gefunden: 01604 - Potentiometer 1 für Gangerkennung (G239)             16-00 - Signal außerhalb der Toleranz

This potentiometer is located on the gear selector and detects the gear selection (i.e., in conventional gearboxes, it determines whether the gear lever or the transmission input shaft is in the front position (gears 1, 3, 5), the middle position (neutral), or the rear position (gears 2, 4)).
Es gibt noch einen weiteren Poti (G240) für die Gassenwahl (sprich: bei konventionellen Schaltgetriebe: Ganghebel bzw. die Getriebeeingangswelle in Mittelsstellung (Gasse 2. u. 4. Gang), links (Gasse 1. u. 2. Gang) oder rechts (5. Gang).

The G239 sensor, which is showing an error, always reads close to 0V (0.12V) in VCDS. Even when I manually shift the gear selector and engage gears, this voltage value remains the same. Subsequently, I used a standard multimeter to measure the resistance in the wiring harness, between the two outer terminals and between each outer terminal and the center terminal.
T14e/1 - T14e/3 (External taps): 853 Ohms.
T14e/1 - T14e/2 (External to Internal): 836 Ohms.
T14e/2 - T14e/3: (Internal connection to external connection 2): 56 Ohms.
The T14e connector is a connector in the wiring harness that houses all the cables leading to the gear selector (located under the hydraulic lines at the front of the engine compartment).
I was also able to measure the 5V operating voltage directly at the connector on the potentiometer (T6g).

"Oddity: These values don't change when I manually shift gears, which is precisely what the potentiometer should be doing."

Following that, I used a second potentiometer (with the same part number, according to the documentation) of type G240 for reference measurements.
T14e/4 - T14e/6 (External taps): 485 Ohms.
T14e/4 - T14e/5 (External to Internal): 437 Ohms.
T14e/5 - T14e/6 (Internal to External, pin 2): 56 Ohms.

"2. Peculiarity: These values also do not change when I manually select the channel. To rule out the wiring harness, I measured the values directly at the potentiometers again, specifically on channel G240."

3. Oddity: Why are the resistances at the outer terminals so different in what is supposedly the same potentiometer? I don't believe this is within the normal range of component variation.

Question
* Was mache ich bei der Fehlersuche falsch (Messfehler? Cognitive bias?
* What might be the next sensible step to take (e.g., ordering a potentiometer for about 70 euros and seeing if it exhibits the same strange behavior)?
* Remove the actuator and check if the mechanical linkage has been damaged in both resistors simultaneously (no work has been done on the actuator yet, so this is very unlikely)?

Bye.
peace

Hello,

I would manually adjust the potentiometers and use a multimeter to check if they are working properly.
Please also check the wiring harness to the control unit when inspecting the potentiometer that was reported as defective.

Best regards, Rainer.

Hello Rainer,

"That's exactly what I did, by manually adjusting the lane and track selection on the track selector. And at least with the G240, I measured directly at the potentiometers (excluding the wiring harness). Each of the sensors is mechanically linked to a specific longitudinal and pivoting movement. Strangely, nothing happened with either potentiometer – neither the one that supposedly was defective nor the one that was working. I'm just not understanding this right now and I think I might be overlooking something."

Unfortunately, it's currently difficult to completely remove the catalytic converters or the oxygen sensors because the car is parked on the street in a residential area and won't move on its own.

Bye.
peace

It's possible that the potentiometer isn't actually a potentiometer, but instead provides a digital output signal or functions in an inductive way, similar to the control position sensor of the VP37.

If it's a potentiometer, the slider value must change; otherwise, it's either not one or it's defective.

Hello,

Here's the translation:

"In the parts catalog, it's listed as a potentiometer. In the repair manual, it's also listed as a potentiometer. In the circuit diagram, it's directly wired and depicted as a potentiometer (not as part of a black box or a control unit). In the self-study program, it's also referred to as a potentiometer, and there's even a diagram showing the sliding contact directly."

Otherwise, the more abstract term "transmitter" is usually used when you don't want to communicate the sensor's operating principle.

Therefore, I suppose I'll have to build a circuit that actually provides 5V to the potentiometers – which would be the normal operating configuration – and see if it's then possible to obtain meaningful measurements.

Bye.
peace

Hi,
Just to be absolutely sure, did you disconnect the connector during the resistance measurement?
hg
Herbert.

Yes, the connection was disconnected. I took the measurements at that point. Additionally, I measured directly on the pins of a potentiometer. Any side effects caused by parallel resistors or internal resistances of other components, etc., have been ruled out.

I've now ordered a suitable flat connector housing, individual wires, an LM7805, a pre-resistor, etc., and I'm building a test circuit.

Bye.
peace

Hi,
It was just a cautious question because the measurement looks quite strange for a potentiometer.
Do you have a picture of that part?
hg
Herbert.

Hello,

I can teach you about photography on the weekend. Also, I just bought an old camera stand on eBay. Now I have another item to use to generate new knowledge.

Bye.
peace

Hi,
I would have searched for it the same way. Could it be a mechanical problem, i.e., does the actuator actually move the potentiometers?
Sure, here's the translation:

"Best regards," or "Sincerely,".
Herbert.

Hello,

It's not immediately obvious whether the mechanical coupling is still okay. However, I haven't read anything yet that suggests this is a common or even a known issue. Furthermore, I find it highly unlikely that this would happen on both potentiometers.

These potentiometers are essential for controlling the transmission. It's frustrating that sometimes they work for a few kilometers and then stop working again.

I guess the little car has come up with a great puzzle .

Bye.
peace

Hi,
Let's see what you can get out of that old gang mold from the online marketplace, especially those disassembled pot molds.
hg
Herbert.

Update: Unfortunately, the item was only available this weekend.

Supplement/Revision of previous findings:

"No resistance values dependent on the actuation state and therefore variable can be measured at the supposed center tap on either of the two 'potentiometers' of the replacement gear selector. Therefore, Rainer's assumption gains weight that it is not actually a conventional potentiometer - hence the quotation marks from now on."

2. The movement of the potentiometer's adjustment knob gives the impression of working against a slight magnetic damping force, possibly also against a weak spring (the component cannot be disassembled without damage to determine this definitively).

3. The resistance values of both "pots" (potentiometers) in the replacement gearbox are identical (610 ohms) and therefore different from the resistance values of the "pots" in the gearbox originally installed in the vehicle, which are also different from each other.

"Instead of the 'potentiometer' that was diagnosed as defective on the vehicle, I connected a 'potentiometer' from the replacement actuator (but only connected to the connector, not fully installed). This changes the voltages in the measurement block again. Interestingly, the voltage in the measurement block for the second 'potentiometer', G240, which is used for gas detection, is also changing again. I'm relatively sure that it was stuck around 3V last time. I can't explain that right now. Possibly, a defective potentiometer could also disable other parts of the control unit – in which case, the car is already immobilized, so running a diagnostic program wouldn't make sense."

5. I haven't built the test circuit yet to operate the sensors in an active mode (the parts are at least already here). If any new information arises, I will add it later.

Conclusion: I will likely exchange both "Potis" because the values at the external connections seem suspicious compared to the now available reference sources.

Here are some more pictures of the sender.

Bye.
peace

Small update:

Today, I wanted to replace the "Potis" on the vehicle. Because it appeared that someone had already messed with them and had glued screws with stripped heads back in, I had to remove the entire potentiometer assembly . Fortunately, all the screws came out anyway.

Upon removing the potentiometer, it was observed that its internal components were completely corroded. The sealing ring was deformed in one area, indicating that water had likely entered and caused oxidation. The housing appears to be damaged to the extent that a new sealing ring would not provide a proper seal. However, in my opinion, this corrosion does not fully explain why the potentiometer failed. The drive mechanism and shaft were functional. The potentiometer's seal primarily prevents leakage towards the gear selector housing, but not towards its internal components. The shaft leading into the potentiometer did not appear to be visibly damaged.

There's also a new picture in the image thread.

Next planned step: To remove the gear stick housing (the part where the worn potentiometer is integrated) from the eBay gear stick and install it in the vehicle's original gear stick, while also replacing the entire gasket set at the same time.

Bye.
peace

Breaking news: Today, I reinstalled the gear shifter. It now consists of the "old" upper part with the "old" valves, the "new" potentiometers, and the "new" base plate from the eBay gear shifter.

Der Zustand des Auto hat sich nun geändert von "geht gar nicht mehr" auf "fährt wieder ein bisschen". Ein bisschen, weil er zwar fuhr, aber selbst bei der kleinen Runde auf dem Hof schon teilweise Schwierigkeiten hatte, den Rückwärtsgang und den 1. Ganto be inserted.
Unfortunately, the initial calibration process failed. The control unit automatically attempted the 28 calibration steps twice, but the initial calibration process stopped with an error message "RTFM 12" after that.
Mögliche Gründe lt. Doku:
* Potentiometer values outside of tolerance -> Checked, voltage values within the specified range for all gears.
* Valves -> Well, generally it does switch, so what else could we test here?
* Selector lever -> It's possible, but I'm not sure what was wrong with the selector lever on the eBay-sourced gearbox that was just installed. How can you ensure that the selector lever is at the correct angle on the toothed shaft?
* Ball joints on the shift linkage -> Initially, nothing seemed out of the ordinary, but I didn't examine them particularly closely.

What doesn't make sense to me is that different potentiometers with different resistance values are now installed (since it wasn't possible to measure anything else conventionally). The basic setting is failing, but since switching is now possible again, something "correct" must have happened.
Generally speaking, I suspect that the failure of the potentiometer has masked the original problem with the shifting mechanism. This is a strange coincidence, but not entirely impossible. Next week, I plan to take it for a test drive outside the yard and see if it shifts beyond second gear.

If anyone has any new suggestions, please don't hesitate to share them.

Bye.
peace

Update 10/4/2012:
Although I had a charger connected during the initial setup attempts (RTFM12), it was a highly sophisticated one that follows certain voltage/current curves and, at best, delivers only 4A (but not always). Therefore, the battery was fully charged again, then the value of the clutch position sensor was adjusted closer to the center (from 1.99V to 1.89V), and then further attempts were started.

Attempt 1, 2, and 3: No errors in the memory, and even though the battery had been disconnected for a long time, it was impossible to start any basic configuration – the control unit immediately indicated "IGN. OFF." The grinding point adjustment also failed (0 attempts remaining).

4. Attempt: I disconnected the battery again for 2 minutes, then reconnected it, and lo and behold, the basic calibration started. This time, it ran through to the end without any "RTFM" errors, although the step "Pull the selector lever in the Tiptronic range towards the '-' position" was skipped. After that, the grinding point search also started.

After the unsuccessful attempts to reset the basic settings recently, the Lupo was running again, but it was shifting gears up to over 3000 RPM, sometimes it couldn't shift gears at all, etc. However, after yesterday's successful reset, it was running normally again. I only drove about 2km, though. The original problem only became noticeable after driving much longer distances. Hopefully, I'll be able to do a more thorough test drive this weekend.

Update 10/08/2012:
He has survived several in-city test drives and a longer, approximately 60km cross-country trip - however, I haven't dared to take it to work today (about 100km) because no one can pick me up if something goes wrong .

Update: October 10, 2012:
Yay, he managed to complete the 100km commute to work today .

Bye.
peace