Originally Posted by Pikey Dave
I still had the senders in the tanks & had them connected into my I-command/nmea2000 network, still useles
. Continuous "low Fuel" alarms everytime you hit a wave.
This kind of supports my findings so far. The WEMA gauges are undamped and the sender unit pictured is also undamped, it seems to be the one regularly fitted, after a fairly extensive trawl of the net. Admittedly it's a lot better than a swing arm type, but it's not right. WEMA makes an SD damped unit, but I'd bet next months wages on not many of them being fitted. Essentially, the same sender as above sits in a sealed tube with a small hole(s) top and bottom. This results in the fluid around the sender having a lag of several seconds, effectively damping out the oscillations.
Back to the OP, I don't really think he was expecting to be told to buy a new engine when all he wanted was to fix his fuel gauge. There's clearly something wrong when over 50% (30/50) of the tank volume is not read by the gauge. I'd be looking at the physical positioning of the sender, the length of the sender and finally checking that the resistance range is 0-180 ohms (I'd personally check the resistances at the gauge end rather than the sender itself.) These are reed switch types, so will never give a true linear reading and it would only take a couple of faulty reeds to make them effectively useless.
Bouncing between empty and full may just be the result of ineffective damping, but I'd also be checking all earth connections and feeds for high resistance, dead shorts and intermittent breaks.