Re: Crazy Gas Gauge

[Scott Fisher <sefisher@domain.elided>]

I had a similar experience with *all* the gauges on my '67 Alfa GT
Junior.  Turning the lights on would make every other gauge -- oil
pressure, gas, water and oil temp -- register about 25% higher.  At
least, most of the time; sometimes when I'd hit a bump, the gauge lights
would flicker and the gauges would read true.

The problem: the instrument panel lights are all hooked together in
series, with a single ground point on the metal of the firewall, and the
instruments themselves also share a common ground point (really -- A
grounds to B grounds to C grounds to D grounds to steel).  The
panel-light ground was loose, which meant that when I'd turn on the
panel lights, the current flowed from the panel-light circuit through
the gauge bodies and sought the good ground, through the shared gauge
ground wiring.

The solution: I stuck my head under the dash, found the ground connector
for the dash lights (easy in the '67, probably harder in a '91 Spider)
and sure enough, there was a loose screw with an eye connector at the
end.  Ten seconds with a screwdriver and the problem never came back.

So: you clearly have a problem in which the instrument panel circuit is
grounding through the gas gauge, most likely because you have an open
circuit downstream of the gas gauge light (that is, between the gas
gauge and the shared ground point).  The next piece of the puzzle is
finding out where that open circuit is.

Your temperature/humidity observation leads me to believe that the
problem may be in the socket of a panel light.  You didn't mention it,
but is there another light that's fairly dim, or out?  My guess for that
is that there's corrosion in one of the other light sockets, causing
essentially an open circuit after the gas gauge light; the current is
grounding through the gas gauge (clearly) and causing it to read high,
but it's only doing it when it's cold and humid -- classic symptoms of
corrosion in a socket causing increased resistance when it's damp; when
the resistance is low (warm/dry), it grounds through the socket
correctly, but when the resistance gets high enough (cold/damp), the
light circuit grounds through the gas gauge.

BTW, for those of you (like me) who grumble and tear out your hair about
Alfa's, er, idiosyncratic use of shared ground circuits, try this on for
size: Yes, it would be stupid to run a single circuit from one brake
light to the other brake light to the trunk light and *then* to ground
- -- if the electrical engineers could count on the material,
conductivity, and location of chassis grounds as they can in most car
companies, where there's a steel unibody that's all electrically
contiguous.  But considering that Alfa's electrical engineers
traditionally had to design circuits that might be installed in
coachbuilt automobiles, which might have aluminum bodywork (typically
insulated from the steel tube chassis by linen pads to avoid
electrolysis, a la Superleggera of Touring), it makes much more sense
for them to conserve grounding points, even if it means long runs from
the device to the known good chassis ground (or even the fusebox).  

At least, that's the little insight I had while trying to trace my
brake-light fault (which, curiously enough, ended up being due to a
corroded socket). And as an added surprise bonus, my wipers came back --
they're on the same fuse.  :-)

Best of luck,

- --Scott Fisher

------------------------------