OK, I decided I wanted a fuel gauge on my 1989 K100/2.  Last time I had the tank off I fitted the harness for fuel and temp gauges.  Then I was lucky to get a used fuel gauge off Ebay cheaply.  It just needed the wires re-soldering.  Before I go any further, I am sure the wires are connected correctly (searched for and found a clear picture to check).

The gauge is apparently not working on the bike.  On the bench, when I connect a 9v battery across the terminals, positive to green/black, negative to brown, the gauge does not read.  (The following will show it's actually trying to go backwards).  If I then reverse the polarity it swings to full, then if I quickly return it to the original polarity it swings back to zero.  Is this normal?

My fuel tank is roughly 2/3 full.  I don't know if I'm doing this correctly, but measuring the ohms between the yellow/brown wires at the gauge end, with the gauge unplugged and ignition on, gives me infinite resistance (OL).  With the fuel gauge fitted and ignition on, I have 12.2 volts at the green/black wire and just over 7 volts at the yellow.  

Do you think the problem lies with the sender, or with the gauge itself?  Thank you for any input.

From your description I would say the gauge is faulty. When the power is removed the needle should return to the zero position.

a picture of the wiring at the gauge  would help howard

is this possible ?

Thanks for the replies.  Two things I should have mentioned.  

First, the gauge will return to the zero position on its own, but applying power as described above drives it back faster.

Second, a picture won't really help as I've lengthened the wires and they are not colour co-ordinated.  With the gauge on the bike, green/black is connected nearest the rider, brown central (screw terminal), yellow furthest away.  Grey/black goes to the bulb.  It is wired the same as this one: http://www.ebay.co.uk/itm/1989-BMW-K100LT-K100-LT-S468-gas-tank-fuel-level-gauge-motometer-/351258226543?hash=item51c89f2f6f  The wires were attached when I got the gauge, badly.

The gauge also works when either the green/black wire OR the grey/black bulb wire is attached to the battery positive, yellow to negative.  Very confusing.

EDIT:  Does this make any sense?  Perhaps the working part of the gauge is earthed through the yellow wire, and the position of the gauge depends on the resistance through that wire from the sender giving various voltages.  If that wire is open circuit, i.e. not effectively connected, the gauge will not work.  A 9v battery gives a reading of almost full on the gauge.

Aha!   From: http://www.aa1car.com/library/fuel_gauge_diagnosis.htm

ANALOG FUEL GAUGES........

The amount of current that flows through the fuel gauge is controlled by the ground circuit provided by the sending unit in the fuel tank. As the fuel level inside the tank goes up and down, the hinged arm that's attached to the float rotates a rheostat. This changes the amount of resistance in the ground circuit which allows more or less current to flow through the gauge.


A shorted sending unit or a short in the wiring between the sending unit and gauge would reduce circuit resistance causing the fuel gauge to read full. And with nothing to slow the amps, the circuit would probably overload and blow a fuse. An open in the sending unit or wiring, on the other hand, would prevent the needle from moving at all and the gauge would read empty.



Looks like a job for winter.  This is very useful, thanks RobMack: http://hrsbstaff.ednet.ns.ca/bvogel/K100/download/electrical/fuel%20sensors/fuel%20level%20sensors.htm
Obviously, first I need to check the wiring further.

UPDATE:  The yellow wire from the sender is definitely open circuit.  So as it's Bank Holiday weekend, a good time to stay off the roads, I've stripped the sender out of the tank expecting a clean up of the corrosion would fix it.  No chance, the sender electrical windings are badly messed up.  A secondhand unit will soon be on its way.  So my cheap and easy fuel gauge project turned out not-so-cheap and easy.