Tankanzeige hängt nach - A4 Cabrio 2.5TDI (BDG) | Posts 32+

I'm not making any assumptions, I'm just sharing my personal experiences, which is often the case.

And if the resistance value measured at the source remains the same and the same measuring device, which measures the resistance, changes the pointer position corresponding to the resistance: Then, a supposed change in resistance at the source (= defect) is simply not logically understandable.

Consequently, there must be an error somewhere in your story. Either in the presentation of the entire story or under the title "Troubleshooting AI".
I am not claiming or suggesting this, but this is what is called logic.

Best regards, Rainer

Hello Rainer,
I completely agree with you: if the display shows a standing car and a constant Ohms value according to VCDS, it can basically only be due to the KI (control unit).

My problem is that the specialist I chose couldn't reproduce the error.
And I have explained the behavior of the AI in great detail to them - but what are they supposed to do if they can't find any errors? I didn't mess up during the repair attempt (and I can also easily understand why a customer might not want the customer to mess up).
Unfortunately, the tank was also 3/4 full - so the AI also behaves normally. I can't say whether and what test resistances they connected there.

Does anyone here perhaps know a company that you believe would be capable of repairing the part?

It must actually be the AI.

Quote:

Does anyone here perhaps know a company that you believe would be capable of repairing the part?

Why not just try a similar used part?

Front-wheel drive: The cover must be on the right side, in front of the spare wheel compartment, with the lock for the fuel tank located underneath.
Quattro: The access point should be located on the left, underneath the rear seat.
hg
Herbert

A minimal update: So far, I haven't found any affordable used parts (I only found one with an unknown mileage for 255€, which is too much for just looking).

I checked again today using VCDS: the behavior is still the same: the fuel level sensor value and the fuel gauge value are constant (I mean 139 Ohms and 37L), the needle jumps to almost 3/4 when the ignition is turned on and then slowly (maybe due to the dampened speed?) drops back to about 1/2. One might therefore think that at the moment of ignition, an incorrect value is measured and displayed, which then gradually returns to the correct value.

Also, according to my understanding, both MWBs behave normally when the car is driven a bit: the values fluctuate by a few Ohms or 1-5 liters, while the diesel in the tank spills out.

What I noticed: The MWB tank level sensor correction is 136.
Does anyone know if this could have an impact, or if the value is 'normal'? The energy display is very accurate and has never needed to be adjusted.

Is the 37-liter tank half full, or 3/4?
That all looks like AI. Maybe expand on that, look for discolored components, inspect solder joints?
hg
Herbert

The value to which the needle then falls is the correct one, so 37L corresponds well to half full.

Do you understand correctly that on the A4 Cabrio, the AI only has one lens that you can simply remove and then it is only held in place by two screws?

Okay, I've just expanded the AI and managed to open it up enough to see the back of the PCB.
Apparently, the 'qualified professional' did not open the casing, at least the stickers that had to be removed were still intact.
On the circuit board itself, I, as someone with no electronics experience, could not see either a burnt-out trace (i.e., a darkened area) or a cold solder joint with the naked eye. I am, however, aware that one might only notice something like that with a magnifying glass.

Interestingly, the fuel gauge is now showing the correct level immediately after installation, which I will continue to monitor favorably. Maybe disconnecting from the power source has actually had some effect?

Does anyone perhaps know a specialist auto-electrical repair shop in the Rhein-Main area that would actually inspect the internal components when you commission them?

Well, the only thing left to do is to clarify what the "expert" actually did and whether he received payment for his services...

Are you looking for a specialist again?
Keep observing to see if the error reappears.
Does this AI still have a constant voltage?
hg
Herbert

Quote:

Interestingly, the fuel gauge is now showing the correct level immediately after installation, which I will continue to monitor favorably. Maybe disconnecting from the power source has actually had some effect?

This indicates a loose connection on the circuit board.
I avoid using connectors (e.g., for the sensor) because this would also cause the resistance measured by the KI to change.

Best regards, Rainer

I paid 50€ - I can probably write this off as an investment.
What they ultimately did remains unknown.
I cannot say whether this part maintains a constant tension.

And, of course, I will now observe it closely.
It also happens that, once the pointer has returned to the correct value, it continues to display correctly until the ignition - I believe it's longer - is turned off.
Maybe there was just a mechanical issue that has now re-aligned due to the movement - I will report.

Is this cabrio still B4 or already B5 platform?
hg
Herbert

It was manufactured in 2004, so it's a 8H model, before the facelift (meaning it still has the original, straight headlights).

So, technology from the B6.

Florian320 wrote:

What I noticed: The MWB tank level sensor reading is 136. Does anyone know if this could have an impact, or if the value is 'normal'? The energy display is very accurate and has never needed to be adjusted.

This is the value in the APK 30 for front-wheel drive models. (Quattro models have 2 channels: 29 and 30)
And it's not the standard value. That is 128. This allows for a slight adjustment of the gauge reading of the tank gauge. The range of values is from 120 to 136. In liters, that is 2.0.
That means: If he shows you 37 liters because of the stored value 136, then he shows you 36 liters with the value 128, and with the value 120 he shows you 35 liters.
If the value increases, the tank gauge moves slightly to the right, and vice versa. This then applies statically to all levels. You simply move the pointer. But only slightly. A small adjustment range, nothing more.

However, using this, you cannot individually adjust the tank capacity in liters in relation to the resistance values of the sender or the corresponding step motor deflection.