B5 A4 fuel gauge malfunction

[Staffanski]

Hi, I need the help/advice from someone who knows the electric wiring of the A4 B5 (2001 model).

My problem is that the fuel gauge is malfunctioning – when the tank is full it reads ‘empty’, and stays showing empty right up to when the tank is less than ¼ full, when the gauge jumps straight to ‘full’. It’s been like this for a while now and I just zero the trip meter each time I fill up, but I’d like to fix it if I can.

So far I’ve tested:

• The fuel sender is OK – removed it from the tank and the variable resistor show the specified readings on a bench test.
• I tried a spare sender unit to rule it out, but still got the same result.
• Fuel pump works fine, and relevant fuses are good.
• Here is the wiring diagram with the relevant circuit bits highlighted:

Fuel gauge wiring.jpgFuel gauge wiring description.jpg

• There is good continuity on all the wiring between the fuel level sender to the T10 connectors in the A-pillar, and on to the instrument cluster’s blue 32-pin connector (pin #5).
• I’ve also pulled apart the instrument cluster to see if there is something obviously wrong there. It all looked OK, and while I don’t have the appropriate tester for in-situ capacitors, those associated with the fuel gauge don’t appear burnt or distended, and the on-board resistors all looked ok with no apparent dry joints.

The only anomaly I can find is there’s a lot of resistance on the earth wire that runs from pin #7 on the blue connector (2.3 Mohms), so given the whole thing is just a glorified variable resistor, I assume it may be something to do with a bad circuit at earth point #269 (“Earth connection [sender earth] – 1 -, in dash panel wiring harness”). Mind you, I did try splicing into this wire to run it to an independent earth, but that didn’t fix the gauge, and there is good continuity on the fuel sender’s earth wire run to the same #269 point via the A-pillar connector T10ao/6.

So I’m lost. Does anyone know the actual location of this #269 earth point, as the wire just disappears into the loom? I’d like to check for a loose fitting or any corrosion?
Hopefully this all makes sense to someone out there. Otherwise, next stop is a trip to the auto-sparky

[rockersteady]

Q. did you test the level sender against the fuel gauge (ie hooked up normally) while it was out of the tank?

Ground for that one should be the earth lug on drivers side of under dash just above the bonnet release lever

Another resource if you haven't checked already is VDO gauge sender troubleshooting charts, VDO likely made the sender in the tank and possible the cluster gauge too. The charts will list malfunction indications.

[rockersteady]

" My problem is that the fuel gauge is malfunctioning – when the tank is full it reads ‘empty’, and stays showing empty right up to when the tank is less than ¼ full, when the gauge jumps straight to ‘full’."


on second thoughts, sounds like the sender is upside down- the float is pegged against empty when there is any fuel in the tank, and only gets off "pegged empty' when "un"pegged and therefore reads full.
if its not that then perhaps the voltage polarity has been reversed.

[Staffanski]

Thanks rockersteady. Yep, she's certainly a conundrum. I had tried connecting my spare sender directly into the sender's pins (#2 is + & #3 is ground) of fuel pump/level sender multi-connector in to boot, but moving the sender arm (and allowing time for the 'anti-bounce' feature) with the ignition on didn't move the gauge arm. I'll have another look at the earth point you mentioned near the bonnet release. I was pretty sure I've checked it in the past when I've had the driver's kick panel out, and I can see some spanner scratches on the painted securing nut, but I also now see an adjacent earth bolt that doesn't look like it's been undone before, so I'll pull the kick panel again and check this one as well.

Those VDO gauge sender troubleshooting charts sound like a great idea. I had a web search but couldn't find anything. Would you have a link to the charts?

I can see where you're going with your 'second thoughts', but the sender can only go in one way, as it slides down into a recess in the fuel pump housing and clips into place. On the polarity thought, interestingly the sender's wiring connector has two parallel spade clips and can push together either way. I thought after the first time I'd pulled it apart that maybe I'd simply clipped it back together arse about, but unfortunately, changing it around didn't help.

It's a bit cold out tonight, so I'll have a look at the earth point tomorrow in my lunch break...one benefit of working from home in the current Sydney lockdown (stay safe everyone!).

[Tas_Audi]

Can you remove the cluster and power it externally and connect either the sender or an equivalent variable resistance directly to the cluster and see what happens. If that is successful then you have a wiring problem, if it doesn't then you have a cluster issue. I would try an equivalent variable resistance in lieu of the sender first. You can also use different fixed value resistors. Just measure the overall resistance of the sender and then pick some values from that.

I assume you have checked the wiper assembly of the sender to make sure it is in contact over the full range of travel of the float.

[Staffanski]

Hi Tas_Audi,
I had the cluster out earlier and tried what I hoped simulated a power-up, using my trusty ol' cluster diagnosis tool...salvaged from an early B6 at the wreckers:
Claudia cluster diagnosis plug.jpgFuel sender test cct.jpg

I'm not sure if this setup was enough to properly power up the gauge though, as it was inconclusive with the needle staying on empty. My multimeter showed power all along the circuit, and resistance-wise, I got the full gradual resistance sweep on the float arm between 276 Ohms at the 'empty' end, and 38 Ohms 'full'. I checked the sweep on both the original sender and my spare (parted from the same B6) and both were the in spec (the factory manual shows spec to be 40 > 280 Ohms but I don't think the couple of ohms either side should cause my issues).

I used a couple of fixed 100 ohm resistors in line back at the connector plug in the boot (pins #2 & 3 are for the sender) figuring 100 ohms should show me around 3/4 full, and 200 ohms around 1/4 full on the gauge, but alas, while showing the respective resistance readings at the connectors in the A-pillar, and on up to point #5 on the blue 32-slot cluster plug, when I turned the ignition on, I got no movement on the gauge needle.

I figure it's most likely a gauge issue if not for the anomaly of the gauge jumping to 'full' when the fuel in the tank is getting down to around 20 litres left (I'm guessing from the amount to refill). This seems to me to indicate that the mechanism of the gauge is working, it's just getting inconsistent and wrong resistance data at the wrong times.

If I can't solve it, I'll just continue to zero the counter when I re-fill. I don't want to shell out for a whole new cluster only to find that wasn't the issue anyway. If I can't locate a bad-earth, then I will see if an auto electrician can find an issue, and if not I might have to, reluctantly, look at mounting an after-market fuel gauge somewhere near the dash.

[Tas_Audi]

I don't believe believe the test circuit is correct. What is drawn would work but that is not how the cluster is configured.The test circuit assumes that Pin 5 ans Pin 7 are connected across the gauge. That is not the case.
The Gauge and sender are a series circuit which means that the one side of the gauge will be connected internally to 12v, the other side of the gauge leaves the cluster via Pin 5 to the sender and the other side of the sender goes to ground via Pin 7. By connecting to pins 5 and 7 as shown in the Test Circuit this effectively leaves one side of the gauge open circuit.

To make a test circuit you will need to apply power the the cluster itself. Check me on this, Blue connector, Pin1 and 23 connect to 12v and Pin 24 to 0V. Before connecting resistors or a sender power the cluster and check between Pin 5 and Ground, I would think you should have a voltage. If so connect a resistor between Pin 5 and Pin 7, you should see some action from the gauge, if not the cluster may be cactus.

[Staffanski]

Thanks for pointing that out Tas_Audi, makes a lot of sense.
You are correct on Blue Pin#1 - I followed the wiring diagram and it gets powered off the '15' (ignition switched) 12v source. I couldn't find the Blue 23 pin on the circuit diagram, so I'll dig a bit deeper in the manual to see what connects to it before I put 12v there. When you say "Pin 24 to 0V", does that mean to earth or a switched neutral wire? (sorry, my auto-electrical knowledge is rudimentary).

[Tas_Audi]

Thanks for pointing that out Tas_Audi, makes a lot of sense.
You are correct on Blue Pin#1 - I followed the wiring diagram and it gets powered off the '15' (ignition switched) 12v source. I couldn't find the Blue 23 pin on the circuit diagram, so I'll dig a bit deeper in the manual to see what connects to it before I put 12v there. When you say "Pin 24 to 0V", does that mean to earth or a switched neutral wire? (sorry, my auto-electrical knowledge is rudimentary).

0V is negative. Sorry I just assume everybody speaks electrical

[Gbomer]

Resurrection of thread.

We there a resolution.

My gauge was very troublesome with it now just being dead.

Was about to post to see if any one has done a full work around for this stupid situation.


Sent from my SM-G965F using Tapatalk