At 04:11 PM 6/19/2012, walkingboss wrote:
>Test the voltage at the fuel sending unit. Between the sender wire and the
>ground wire. It should be 10.5 volts with the key in the run position.

10.0 +/- 0.5 is the spec, actually.

>  This
>checks the complete system except the sending unit itself. If you have 10.5
>at the wires you have verified that: 1) continuity to all components,

Yes, except maybe bad connection at the plug (which was my problem,
cured by cycling the plug a couple times).

>  2) the
>gauge is in working order,

Working, but not necessarily calibrated.  However I don't think these
gauges have any great tendency to go out of calibration.

>  3) the voltage stabilizer is in working
>order.

Yes, with the (very unlikely) proviso that the regulator may be
failing under load.  I don't know the internals of the regulator well
enough to predict likely behavior in that case.

>This only leaves the sending unit as the culprit! However if the
>voltage is higher/lower, replace the voltage stabilizer and retest.

In order for this test to be definitive you'd need to at least test
the full-scale reading with a 35 ohm resistor, because connection
problems that won't bother the voltmeter (which draws about one
microamp at ten volts if it's a digital) can still interfere with the
gauge which draws up to a quarter of an amp or so.

Also if you don't have easy access to the sender plug it may be
easier to work from the middle, i.e. the disconnect plugs behind the
dash or in the fuse/relay panel which give you access to both the
gauge side and the sender side.  At the very least you'll know
whether you have to lower the tank enough to get your hand in to the
sender connection.  If the tank's already out of course this is moot.

Yours,
David