Corroded Cpu Block (img)
 

I finally got around to cleaning my cpu block and found something quite disturbing. It seems that the top half of my dtek spiral (amd) has begun to corrode and its not just the paint but the aluminum itself. The copper bases of both the cpu block and gpu block(plexi cover) are fine. I use a distilled water/anitfreeze mixture with all plastic connectors. Is this normal? Im also wondering what my radiator which is also made of aluminum is hiding.

here is a pic..

http://img72.photobucket.com/albums/.../cpu_block.jpg

How old is it? Are you using any additive? I'd imagine it's pretty normal when you mix metals in a system. The corrosion doesn't look too deep though. Maybe you could give it a coat with something?

Those are not plastic connectors in that pic.... :shrug:
Also what I find equally troubling is the corosion outside the O-ring area.

those were the stock connectors that came on the block and are chrome. Im assuming this corrotion is common to these blocks since for some reason the poly topped block is 2x more than the aluminum one on the website (maybe they think there is a problem). Luckily they sell replacement tops so i went ahead and prdered a poly one instead.

The problem is your radiator not the block IMO. If it is all aluminum then it is NOT protected/anodized on the inside of it. That is causing a problem with the base of the block being it is copper. Clean the system very well and then run 50/50 antifreeze and water. WalMart and other places sell 50/50 premixed. That is what I have been using lately.

This is NOT common though. The Aluminum radiator is bad to mix with a copper based block. My guess is the antifreeze mix was not strong enough. :shrug: Good luck. Good post!

I was only using a 80/20 mix at most so that could be the problem

Yeah... That's about right for anodized aluminum...

If the radiator was the problem, why would the top be corroding? I'm betting it's just galvanic corrosion between the copper base and the aluminum top. With the pieces that close together, even a tiny pinhole in the top's coating would definitely lead to extensive corrosion. The corrosion in areas away from the o-rings was probably also assisted by erosion.
The radiator may also be corroded if it's aluminum, but I doubt it will be nearly as bad as the top. For galvanic corrosion to occur, there needs to be electrical contact between the parts (aside from the water). The case may be providing this, depending on how the block is mounted to the CPU, but then there is the fact that the block and radiator are separated by a substantial distance compared to the block and the top.

Particles in the water are plenty sufficent to cause the corrosion. But I agree it was probably poor anodizing on the top. Problem with this is the corrosion particles may have entered the rad and started in on it to?

Rad and block are sufficiently connected via the water to corrode - people have seen this many times (usually with a copper rad and Al blocks). It'll be real hard to find out how much the rad is corroding without tearing it apart though.

looking at this makes me cry. how can I flush out my system to clean that crap out of my possibly corroded radiator? also I have heard that the poly tops on the spir@L crack all the damn time so I will never, ever get one.

Now that you mentioned it I remeber reading about them cracking as well which was why i went with the aluminum in the first place. I really dont have the cash to go out and get a new block atm so hopefully the poly will last until christmas rolls along.

while its off might as well bust out the drill and dimple the inside of the block...at least thats what I'd do :) itll help cooling

I was thinking of doing that but im not sure how thick the copper is at the base so i dont want to accidentaly go through. Ill probably just score it up a bit with my dremel if anything. Its also been a bitch trying to get the black aluminum crude that cemented inside the copper spiral to come off.

The parts must be in electrical contact aside from the water (which is the ionic conductor), or else there is no complete path for current to flow. In many cases it may simply be the connection through the case chassis, which the block and radiator are both bolted to. In the case of the block and lid, the aluminum and copper could easily be in physical contact, and there is obviously water touching both of them, creating a very short (low resistance) path for ionic conduction.

You get particles of metal eroding off the surfaces and floating round in water cooling. Not many, but enough that given enough time you will get galvanetic corrosion.

I should have taken a picture of the inside of the block when i first opened it. All the metalic "ooz" never left the cpu block. It was all collected inside of the spiral channel of the base copper portion which most likely accelerated the galavantic corrotion. I would have discovered it sooner but I striped one of the allen screws when i was first trying to get it open a few onths ago. I eventually used a dremel to cut a line into the screw so i could use a flat head.

Due to the nature of waterblocks - they have a lot of turns and other restrictions, I can't think of a much more likely place to collect debris. ;)

I think the radiator would be worse, considering how narrow the tubes are.

I'll bet if you clean out your radiator, get new tubes, clean up the base and sand off the loose crud on the bottom of the top, run new distilled + antifreeze and run it again you wont see any more problems for a while. I'll bet it was a slow degredation of the loop that caused this.

I had a block that looked all black and cruddy once, but it had been sitting in the water for about a year without cleaning and the radiator was old and probably cruddy when I put it in there.

EDIT: and this is why I always use plastic barbs on my alu parts, with the exception of my WW because I am not sure I have any barbs that fit it.