>On Tue, 2 Sep 2003, Vanagon Man wrote:
>
> > But there still is a wierd problem.  It seems that the pressure cap
> > allows coolant to leave the system into the overflow, but it does not
> > suck it back in.  So, if you have the overflow filled to max, after a
> > drive you look at the overflow and it is 1 1/2 " higher than high, and
> > the regular expansion tank ( the one with the fitted pressure cap) has a
> > good sized air bubble in it.  I have bled the system more than once, and
> > this same thing keeps occuring.  Tested again for exhaust gasses, none
> > present, and no coolant leaks on this vehicle.
>
>Test the pressure cap and make sure it's releasing under vaccuum.  Then
>replace the hose between the overflow and expansion tank, making sure the
>new hose fits tightly.  (I think I used a chunk of fuel hose I had lying
>around, but I don't know what size it was right now.)  If the hose has any
>pinholes in it they may break the siphon and prevent the overflow tank
>from pulling coolant back.  Check for trash in the bottom of the overflow
>tank that might be acting as a check valve, too.
>
>If none of this works, you may want to loosen the radiator bleeder with
>the engine hot and see if any air comes out.  Any air bubbles in the
>system will tend to prevent a vaccuum from forming and liquid from being
>drawn back in.  If you routinely get air at the radiator you have a leak
>somewhere.  Some people have told stories of pinhole radiator leaks that
>weren't obvious under pressure but let air in under vaccuum.

If David's suggestions don't solve the problem, then it'll be time to
consider the likelihood of compression gasses escaping the headgasket into
the waterjacket and displacing your coolant.  The dreaded internal
headgasketitis.
I hope it's something simpler, but at the list will have plenty of advice
for you depending on your choice of ways to go at the gaskets.  Just ask
'em about using JB Weld to resurface the heads.