Tear it down
5-24-00
Well, it's been a little over seven months since Water Cooler 4 was installed on my Windows box.  Since it's time for a processor upgrade and, of course, I plan to water cool and add some peltiers to it, I though it would be a good idea to strip it down and see what 7 months of chillin' does to a rig.  During this time the processor was run mostly at 633 MHz.  Yes, it would do more.  No, it wasn't completely stable at anything above 633.  If I had one crash that I could attribute to something other than just the normal Windows array of crashes, it would be one crash too many.  As well as being my play toy, this box keeps track of my life and I don't take kindly to down-time.

During the seven months, I ran Rain 1.0.  Always.  Now there are those that say that running a software cooler and peltiers is a bad mix.  I am not one of those people.  I tend to think that good insulation is all that is needed to get around the condensation issues.  However, the use of a software cooler does make for some pretty good temperature swings.  Consider that the normal internal processor idle temp for this setup was about -7°F (seven below zero) and gaming, CAD work, and a few other uses could push the temperature into the +60°F range.  That's a fairly good swing, but not really any more than a stock heat sink would produce under some situations. (+75 to +142 for example.)

Aside from just liking the cooler programs for their power saving aspects, another reason to run Rain was to see how well the packaging of my cooler held up to the humidity and the perils of condensation.  I have to say that I was impressed.  Not only did the Monokote covering do its job in keeping condensation from forming inside the peltier assisted, water cooled package, but the use of dielectric grease completely prevented any corrosion from forming on the PPGA processor's pins and the slotket's edge connector.
 
 

I can honestly say that it was very tough to have to tear this 
project apart, but before I moved on to WC7 and even colder
temps, I needed to know more about the long term effects of  running sub-zero temperatures.

The edge connector is clean.  The dielectric grease served
its purpose well.  The residue on the face of the water cooler
block is the adhesive from the duct tape that held the water block's thermo-sensor in place.

No signs of condensation or corrosion here either.  It does have a fairly nice pattern of thermal grease though!  You can see that 
both the processor and adapter were coated with polyurethane,  just in case some moisture got in.  The yellow tape holds 
another thermo-sensor against the edge of the slug.

Hmmm, aside from my strategically placed thumb print, this shot shows nothing that shouldn't be there.  No signs of corrosion.  
It would appear that the dielectric grease did its job here also.

All in all, I'm pretty happy with the way this cooler worked out.  I'd say that in this particular situation, those who predicted that this processor would die a premature death due to condensation were wrong.  It does appear that with a little care, running cold temps doesn't have to mean that condensation will kill your processor.  Now that the tear down has been completed, I have more confidence that the next cooler can go a bit colder.  Hopefully in six months or so with the next tear down, I'll be able to report the same success with WC7.