Random Phrase: Make yourself at home...


Read Me First!

This is the Spode's Abode archive. The old articles and forum have been kept here purely for historical purposes and are no longer updated.

Not all portions of this archive may work as expected.


Please visit the new site.



Jam Jar Waterblock
Written by Spode (24/Mar/01)
Page 1 of 1

Untitled Document

As a great majority of people reading this will know, I am quite obviously obsessed with jam jars. I will sit down and think of marvelous ways of doing things, but for some obscure reason, I always return to jam jars - I don't know why!!! But they are readily available. If you haven't already read my Jam Jar coolers 1-3 and reservoir - goto the article archives and do that!

If you can remember Jam Jar Cooler I, it was basically a reservoir of water that sat on the CPU. With todays CPU's, I'm not so sure I can do anything like that - so I decided to actually attempt to make a waterblock.

My main problem was to be making a clip. This was a task I really couldn't be bothered to fiddle with at the momment - instead I thought direct die cooling (much like the docs article). The celeron is easy to perform direct-die cooling on as there is just a core in the middle of a plate - with a Duron you have resistors and bridges left, right and center. I would have to try and make the water just touch the core and nothing else. We'll talk more about the attatchment later - for now, lets look at the block design.

As you can see here, the jar is going to have a just larger than core sized hole in the bottom and sit directly on the chip. I used the same jar that I used in Jam Jar cooler II - which I had ground flat on the bottom. I then made the hole by using a 3mm glass drill bit. Once I had that hole, I slowly enlarged it using larger drill bits, and eventually the grinding attatchment on the dremel.

I decided against using the lids of jam jars, as I have found them not to be completley watertight. Instead, I used a peice of plastic for the lid (in this case a blanking plate). It will be held on using epoxy putty. I have taken the budget approach for hose connectors and once again used biros :D

I drilled 2x6mm holes and jammed the biros in there - it's a tight fit. As you can see from the diagram, the idea is to have cold water put straight onto the core, and the top tube to such out the warm water.

I then sealed the whole unit using epoxy putty. I blocked the hole with bluetack and gaffa tape - and left it over night for leaks. No problems there :D




Here the lapped base of the jam jar before I drilled it.

 





Here it is after being drilled.

 

 






Here is the final block. That epoxy putty is great stuff! It's just like using modelling clay, then when you leave it (I stuck it on the radiator) it dries solid!

 

 



So, now we have the block - how should we attach it? And isn't running water on the die dangerous?

The plan was this, put a real thin layer of epoxy resin on the core to stop it seeping into the core slowly, then attatch it using epoxy putty. So I did. I fired it up and everything seemed fine. But my Duron 800@800 (1.85v) was geting into the high 50's early 60's. NO! Something was wrong. What was happening was the epoxy was acting as a thermal blanket and stopping the heat from transferring to the water. I wasn't sure if that was the only problem - I wondered if there was maybe just not enough surface area to transfer the heat. So I took a big hung of aluminium and epoxied it onto the core. Guess what, it was WORSE. It seems that epoxy is really quite pants at transferring heat.

There is one problem with this design, and thats the fact you can't use a peltier, and there is a point of no return :( For me, that point was once the epoxy putty had completley cured. Luckily it hadn't and I prized the jar off.


Here you can see (although quite blurred) the epoxy putty left behind. I slowly scraped this off with a knife, and the epoxy off the core. Everything was going well, until I slipped and took a resistor off :(

My Duron 800 was no more, and I was slightly disheartened.





Conclusion

Once again, I have tried to live up to Dr. Surlyjoe's standards and failed. I feel my idea is a good one, but perhaps using a jam jar was stoopid. Either way, to improve this design, I think what is needed is a lump of copper and some Artic Silver Epoxy. The Artic Silver Epoxy will transfer the heat instead of being a heat blanket, and then put the lump of copper straight onto the core. Remember, anything like this is pretty much permanent.

Spode



Copyright Andrew Miller
Please read our disclaimer

Search the site:

Random

ZipLinq Retractable USB Phone Charger