>The coolant level control module (painted #43) applies a weak square
>wave to one pin of the coolant level sender.  If that pin is shunted
>to ground or any constant level by less than about 180 kilohms it
>will swamp the square wave; and a few seconds later the module will
>trigger and apply about 35 ohms to ground to the gauge sender input,
>simulating overheat.  The original controllers did this
>continuously; but the newer square ones only do it very briefly,
>often enough to keep the blinker triggered but not enough to move
>the gauge needle to speak of.

OOPS!  That's completely backwards.  If the square wave is allowed to
run free then the module will trigger a few seconds later.  As long
as it's swamped by being shunted to a constant level (the ground pin
of the sender, via the coolant) it will NOT trigger.  D'oh!  Also,
the module may be painted 43 or 42.  On the 2.1l vans it's mounted on
the fuse/relay panel; the 1.9l have it on the cab wall forward of the
fuse/relay panel, above the star grounds.  Accessible with difficulty
by dropping the fuse/relay panel.  The old version (changed for '86
or some time in '85 I believe) is taller and will cause the gauge
needle to enter the overheat zone as well as triggering the
blinker.  The new version (cube-shaped) avoids that and will swap
directly in for the older one.  There are two p/n for the newer
version that behave the same, strictly an internal difference.

The old version automatically resets when the sender is once again
shorted; the new version may continue until the ignition key is
cycled.  Probably an improvement in practice though it makes testing
a bit more tedious.

Yrs,
d