At 03:09 PM 5/25/2010, neil n wrote:
>I'm curious. As coolant mix gets old over time, would the change in
>coolant chemistry affect conductivity of coolant?

Maybe.  My guess is not much.

>  i.e. would a light
>taking longer to turn off be a pre cursory warning that the coolant
>needs to be changed?

No, it doesn't work that way at all.  Blink timing is *entirely*
controlled by the internal gauge circuit -- each time the light is
triggered it should run about three seconds.  Taking longer to turn
off is a preliminary warning that your gauge will need to be opened soon.

>And, if one has added too much water, would a
>mixture that's water rich cause the light to blink constantly?

The controller will trigger if the resistance it sees is above about
180 K ohms.  Depending on how pure your water is you might possibly
exceed that, but not if your concentration is anywhere remotely near
50 per cent -- which is where it should be at all times unless you're
actually in true Arctic conditions (below -35F).  At all other times
including in the boiling desert, 50% is what VW specifies and what makes sense.

Yours,
d