David, it's a 1983 diesel.

Jim

On Sat, Feb 21, 2009 at 7:51 PM, David Beierl <dbeierl@attglobal.net> wrote:
> At 08:13 PM 2/21/2009, Jim Felder wrote:
>>
>> the coolant light blinked slowly at first, but as the engine heated
>> and the temp needle rose, the blinking became more and more of a
>> constant red light until, at needle center, it was on full time. When
>> I reached my destination, I turned it off and restarted and it was
>> fine ever since.
>
> Fascinating.  I've never heard of like behavior.  What year is this beast?
>
>
>> I don't know what's in the control module, but I doubt that it is
>> really a relay...
>
> It's not.  It's a rather fancy circuit that sends a very high-impedance
> (weak, if you will) square wave out on the active level sender terminal.  If
> that square wave is tied to any constant level between ground and +12
> through a resistance of about 180K or less it will be quenched, and the
> gauge driver remains inactive and floats with the temp sender output.
>  However if the square wave continues for several seconds without being
> quenched, the gauge driver goes active and pulls the sender line down to a
> level corresponding to an overheat condition (not all the way to ground,
> which would burn up the gauge).  That triggers the flashing light in the
> gauge, which flashes for several seconds whenever it's triggered (and it's
> this behavior that gets out of hand inside the gauge and leads to constant
> flashing for no cause).
>
> At this point the old (large) and new (small) controllers part company.  The
> old one applies a steady overheat signal until the square wave can
> reestablish itself, which drives the needle into the hot zone.  The new one
> applies an intermittent signal which keeps triggering the blinker without
> greatly affecting the gauge reading.  The new one also latches on, so the
> blinking will continue until key-off regardless of what the level sender
> sees.
>
> I had a lot of enjoyment tracing out and analyzing the interior circuit of
> the old-type controller.  It was cleverly done.  I haven't seen inside the
> newer one, but I suspect they put things into a single no-doubt-proprietary
> chip that were wired out on the old one, since the new one is half the size
> and does more.
>
> --
> David Beierl - Providence RI USA -- http://pws.prserv.net/synergy/Vanagon/
> '89 Whitestar "Scamp"
>