Right, a bit of progress to report on the 3B Passat cluster, the fuel gauge is fully sorted !
The problem was that the resistance characteristic of the Passat fuel sender is the opposite to that used on most earlier VW's, including the T3, T4 and Mk3 Golf.
It now is reading the correct way round, and zero with an empty tank, 1/2 with a half-full tank, and full with a full tank.
Plus, via VAGCOM/VCDS, it is actually reading the correct fuel volumes in litres, all the way from empty to full !
To get definitive T3 sender data, I spent all yesterday afternoon siphoning all the diesel out of my pickup, and then re-filling it, 2 litres at a time, all the way to 70 litres, recording the sender resistance at each point.
I made use of Excel to record the data, and to determine the best-fit curves from which I was able to derive the new values needed. Its a bit too much data to post up here, but as an example here's the Syncro vs Passat sender resistance data:
It took quite a bit of time and effort to achieve this, I've been tinkering with it on and off since last year. The most time has been taken figuring out the necessary changes to the onboard microprocessor EEPROM, which it uses for various data, eg.
Fixed configuration data such as tacho and speedo scales, petrol/diesel etc.
Variable configuration data changeable via adaptations in VAGCOM, such as auto/manual trans, fitted options, service interval reminder and so on.
Security data such as immobilizer codes and keys
Odometer data, eg. total and trip mileage
Aftermarket software is available, which using a generic ODB cable can read and write the EPROM, and modify the behaviour of the cluster. Some of the configuration data is well understood, and standardised across many versions of cluster. The software can make 'automatic' changes to this data, however most of the data is not so well understood and any changes must be made by manually using a HEX editor, on a saved dump of the EPROM, based on guesswork, experience, and trial-and-error.
After some time spent on various internet forums, I was able to establish that the data controlling the fuel gauge is held in three main tables, plus a few other bits I didn't have to be concerned with. Each table contains 8 data points, corresponding to 8 gauge needle positions, and the microprocessor controlling the stepper motors interpolates between these resulting in continuous smooth needle movement. The 8 positions are Empty, Reserve, 1/4 Full, 1/2 Full, 3/4 Full, 'Nearly Full', Full, Full (yes, twice).
The first table contains values representing the resistance read from the sender by the gauges A/D converter. I think its purpose may be to cope with the non-linearity of tank/sender. I had to modify this to accommodate the lower resistance of the T3 sender, and its different characteristic curve.
The second table contains values which map the actual fuel volume to the resistance of the sender. I had to modify this table, (a) to reverse the mapping fuel volume-> resistance, (b) modify total volume 60-70l to cope with the T3's larger tank.
The third table I didn't have to change, this determines the physical stepper motor position for each of the eight needle positions.
For testing the EPROM changes I used a spare Syncro tank sender rigged up to a cluster on the bench.
One problem I had was that at absolute 'Full' reading (minimum sender resistance), the Passat gauge would suddenly drop to zero ! After some research I established this was because the cluster has a predetermined minimum feasible sender resistance, below which it assumes a sender fault and logs a fault. I simply worked around this by adding a 10 ohm resistor in series, and adjusting the A/D converter table slightly.
This may all seem completely OTT, but I feel it was worth it, as I am completely committed to using a more modern instrument cluster, and it has to be 100% functional are else it would be just annoying. Having accurate fuel data is if course important when trying to reliably establish engine fuel consumption. Finally I have the satisfaction of having the most accurate Syncro fuel gauge its possible to have !
