A lesson in mixed metals

[Coog]

Here's a good example of what happens when you don't use antifreeze while mixing metals :bang head I was pretty lucky I discovered this before it leaked! My VGA block looks about the same. dangit! Both blocks are shot...

http://www.ctcdesign.com/misc/resize_wb.jpg

[Perseus]

Geez, that sucks hard. :bang head Well, live and learn.

[AlucardCasull]

what was the other metal? How did you find out it was all corroded?

[Coog]

Other metal was copper, it's an aluminum housing with a copper cool plate. The copper looked ok. I was swapping cases and just happened to look inside the block, once the screaming stopped I took it apart...

I was very bummed! Thats $90 worth of blocks...

[kbtat2]

That sucks!! It's wierd that it looks like hard water deposits.

[nikhsub1]

Definately corrosion... this is why i NEVER use aluminum ever. Most mfgr's are moving away from aluminum which is a good thing.

[voigts]

I have two aluminum anodized hard drive coolers, but have been using about 10% antifreeze and change the water often, and so far in about 7 months they are still as clear as glass. I might go back and replace them with something copper if I get the time to make something (and yes, I do like my hard drive coolers and intend to keep using them).

That's a bummer about the corrosion.

[Etacovda]

Other metal was copper, it's an aluminum housing with a copper cool plate. The copper looked ok. I was swapping cases and just happened to look inside the block, once the screaming stopped I took it apart...

I was very bummed! Thats $90 worth of blocks...

then perhaps you should have used anti-corrosive like the instructions told you to :P personally i run 10% AF in my systems even if they dont have mixed metal; and a bit of anti-bacterial solution.

[Perseus]

(and yes, I do like my hard drive coolers and intend to keep using them).

:)

[Omega Destroyer]

That's interesting, I thought anodized aluminum wasn't supposed to corrode, even in a mixed metal situation.

That also looks more like hard water build up to me.

[DragonTattooz]

Based on my experience with bi-metal systems and cryogenic cooling (using LN to cool superconducting coils in MRI scanners), the bi-metal has far less to do with your problem than the fact that aluminum WILL ALWAYS corrode in a 100% water environment. You MUST use some kind of corrosion inhibitor with aluminum, it's not a matter of "if", it's a matter of "when" it will corrode, and how fast.

Personally, I have friends in the semiconductor industry and I'm going to run 100% Galden (http://www.solvaysolexis.com/Galden.htm) in my system. Galden is ~$500/gal- and it's just about the best "coolant" you can buy.

Since I'm on the subject, I'll mention something else. Most of us will NEVER have to deal with this, but I'm going to mention it since we are on the subject of metal corrosion. Do not EVER run de-ionized water in your system. De-ionized water will leach the ions out of any copper, or copper alloys, in your system. Essentially, that means that it will "dissolve" the metal over time- and it doesn't take long.

[Omega Destroyer]

Great, thanks for the de-ionized. I've been running mine with deionized water and 10% antifreeze for the last couple of weeks.
Should I just add some salt or is the antifreeze enough?
I had heard elsewhere that deionized water was a good thing.

[DragonTattooz]

Great, thanks for the de-ionized. I've been running mine with deionized water and 10% antifreeze for the last couple of weeks.
Should I just add some salt or is the antifreeze enough?
I had heard elsewhere that deionized water was a good thing.

I don't know about adding salt. Interesting idea, but I would not try it on my system. Too much at stake with the electronics should a water block or fitting fail.

Rather than me telling you that DI water is NOT a good thing, I'll provide some links and you can draw your own conclusions. :) First though, keep in mind that my statement is ONLY in reference to copper and copper alloys (brass, bronze, etc). Plastic and stainless steel are fine in a DI system. But it seems that copper blocks are the way to go with water cooling, so I felt compelled to mention the DI thing.

One more thing, as you read you will see that there are ways to make a DI/copper system relatively safe, but it's a PITA. For a home computer water cooling setup, you don't want to mess with it- if for no other reason than the fact that there are better solutions.

Here's some links-
Link 1 (http://www.finishing.com/320/35.shtml)
Link 2 (http://www.koolantkoolers.com/index.php?nic=h20qualspec.html)
Link 3- This one is interesting. (http://www.frenchcreeksoftware.com/_fcsdg/00000084.htm)

If you Google "deionized water copper" you will get some science experiments where they discuss using DI water to dissolve copper. That should tell you everything you need to know.

[DragonTattooz]

Great, thanks for the de-ionized. I've been running mine with deionized water and 10% antifreeze for the last couple of weeks.
Should I just add some salt or is the antifreeze enough?
I had heard elsewhere that deionized water was a good thing.

Are you sure you are using deionized water and not distilled water? Distilled water is perfectly fine to use, as long as you use a corrosion inhibitor along with it.

[Sneaky]

mmmmmm tasty corrosion


j/w, how would that contribute to leaking? theres no major damage around where the o-ring seals... the anodizing is just gone


edit: don't feel too bad about it... it just gives you an excuse to upgrade again :p

[Coog]

Looks like hard water buildup but it's actually raw aluminum, the seat for the o-ring on that side is shot too. Very wierd how it's all on one side.. other side still looks like new. Been running that block for about 3 years so I can't complain too much :shrug:

I think I'm going to order the swifty mcw6002a, black ice pro, and a danger den dd5 mag pump... want all pieces to have fittings for 1/2 id tubing

oh yea, and I'll add 10% antifreeze :)

[martindemon]

I had a Koolance (worst kit ever IMO) and was lucky enough to have a good corrosion problem. You see, the golden paint that Koolance name "plating" just pealed off the copper of the CPU waterblock and all the other block and the radiator were in aluminium. I had put the exact quantity of Koolance liquid, exactly as shown and explained in the instructions, which I read three times before beginning. I used distilled water as written in the manual. I really did all I had to do.

Now you will perhaps ask yourself what can happen if there is corrosion. In my case, I'm almost certain it was the cause for the pressure rise that happened then. When I put the system upside down to open the reservoir and there was always that strange PFFFFFFFFFF; the reservoir was in fact spitting in my face :) Then I proceeded to rescrew it and go on. The reservoir kept leaking every 2 or 3 days, and so I kept putting it upside down and unscrewing it to remove the excess pressure. One day, I was dumb and I thought that maybe I should redo the liquid to make sure there was enough anti-corrosion liquid. So the dumb part is next. I proceeded to put the system not upside down (reservoir in on the floor of the case) on a bucket to bleed the system. I unscrewed the plug of the reservoir and then something happened. I think you know what...

SPOILER
I was attacked by my own Koolance! Something like half the liquid of the reservoir was spat out of it at sneezing speed in the bucket, only to bounce back right in my face and torso :bang head Then that's when I smelled that burnt tire odour that was emanating from the liquid (and me...) I'm not sure if it was smelling it before corrosion or not but that was awful. I changed and waited; I was not willing to try the beast so soon after having been PWNED by it so much.

During the next days, I tried several times to replace the liquid. First, I proceeded as per the instructions. Then I tried with only distilled water. No difference in the results. I then disconnected all the water circuit and put back the stock heatsink fans everywhere and sold the Koolance to the vendor who had sold it to me, at 1/3 the price. It was then I decided I would never buy a kit again.

Now, more than 2 years after that incident, I'm ready to start again with the components I will choose. In a COPPER ONLY (acetal is acceptable as it is neutral) circuit. I think brass is near enough copper to not react a lot. I will use a T-line as it seams most peoples do it that way; it must be for a reason.

[Sneaky]

Looks like hard water buildup but it's actually raw aluminum, the seat for the o-ring on that side is shot too. Very wierd how it's all on one side.. other side still looks like new. Been running that block for about 3 years so I can't complain too much :shrug:

I think I'm going to order the swifty mcw6002a, black ice pro, and a danger den dd5 mag pump... want all pieces to have fittings for 1/2 id tubing

oh yea, and I'll add 10% antifreeze :)i'd spring the extra $10 or w/e it is and get a black ice extreme and have a fan suck air through it (just add a bit more performance)

[SewerBeing]

i'd spring the extra $10 or w/e it is and get a black ice extreme and have a fan suck air through it (just add a bit more performance)

Ok the BIP will beat the BIX if the fans push less than 80CFm, if they push over 80CFM then the BIX wins. Now you should be using the BIPII or BIXII anyways. So if you like quiet the BIPII is the rad for you (the new Swiftech rad might be better than it though I would look for it).

[Coog]

I want to stick with a single 120mm fan so the BIPII is out, Danger Den has the BIP for $29 and the BIX for $45. Big difference. The fan is just a generic radio shack 120mm I've had for years that does 85 cfm, and I'll probably run it at 7v anyway.

the new Swiftech rad might be better than it though I would look for it
I looked through the swifty site and their mcr120 only comes with 3/8 barbs. Too bad, I like swiftech stuff...

[Omega Destroyer]

Yeah the water I used is definitely deionized. Really pure stuff from a Milli-Q system with 18000 Ohms of resistance.
Measuring the resistance today, I'm getting about 5000 Ohms in the stuff running through my system.
I really don't want to have to refill/bleed the system again. It's such a pain.
I read the above links. Not sure if I should go ahead and refill or just leave it since all the ions that are going to leech out probably already have.

[SewerBeing]

http://www.jab-tech.com/customer/product.php?productid=3071&cat=257&page=1 <- you can get it with 1/2" barbs.

Omega Destroyer: ions allow water to conduct electricity, thats why deionized water and distilled water ware what people use along with some additive to prevent corrosion.

[Omega Destroyer]

Yes I know they conduct water, that's why resistance is a good way of measure the purity of dionized water. The question isn't about their conductivity but the corrosive properties of deionized water compared to distilled water.
I think I'm just going to wait it out and see what happens. I'm too lazy and too strapped for time to drain it and refill it right now.

[Omega Destroyer]

Oh and by 20Kohms I meant 20MegaOhms

[DragonTattooz]

http://www.jab-tech.com/customer/product.php?productid=3071&cat=257&page=1 <- you can get it with 1/2" barbs.

Omega Destroyer: ions allow water to conduct electricity, thats why deionized water and distilled water ware what people use along with some additive to prevent corrosion.

I have extensive experience with DI water in multi-million dollar semiconductor equipment, so I'm not just talking about something that I read somewhere. The anti-conductive properties of DI water are irrelevant in a home computer system. What are you going to conduct? A12vdc signal? Come on. If the desire for anti-conductive properties is due to a concern about leaks, a short circuit is going to be the least of your worries if your case gets flooded.

The reason high resistance DI water is used in applications like the cooling loop on a 20kw generator is that a 20kw generator is putting out a whole lotta power, and the current is looking for a place to run to ground. I've seen a 20kw generator jump a 6 inch air gap looking for ground. You don't have those concerns with a home computer system. Therefore, you don't need DI water. There is NO upside to DI water in a home application- none, zip, zilch, nada.

Ask yourself this question- Is it worth the risk when there are safer, better solutions available?

OK. I've said my piece on the subject and I'm not going to say any more. If you choose to use DI water with copper, just be careful. I'd hate to see someone on a tight budget destroy their system- and not be able to replace it.

[sunrunner20]

Um, salt would only increase the problem of corrosion. I don't know of anybody has mentioned this already, but salt increases the conductivity of water. When water becomes conductive, not only does it increase the speed of corrosion, but will easily cause a system to fry if you leak some...

[David]

Ok this looks a little confused.

The tap water that comes out your ... uh... tap... is not just water. It also contains impurities which can include calcium, sodium and potassium ions, in some areas also fluorides and chlorides and so forth.

These ions, when in water, will aid corosion of metals by transferring electrons from one metal to the other. The metal atoms that lose electrons become ions and are dissolved in the water, or become attached to negative ions and form either a solution in the water, or a layer of insoluble material.

The ions also increase the conductivity of water, as they carry electrons it makes it easier for the current to flow.

So you want to use water with as few impurities as possible, and as few ions as possible, in a water cooling loop. Especially where metals are mixed.

Water has a very very very slight ionic character but this is very very small compared to water containing ions.

Dragontatooz is right in that the voltages you deal with in a home PC (3.3, 5 or 12 volts usually) are not likely going to flow through your loop. However as an anti corrosion preventative measure, you should always use deionised or distilled water in your loop.

[Omega Destroyer]

I agree with what you said DragonTatooz. It's just that the deionized water has been running through my system for a couple of weeks now, so I figure it's probably equilibrated by now anyway. It's sucked out any ions it was going to suck out.

Next time I'll be using plain old distilled water.

Conductivity was never really a concern anyway, especially considering the fact that the metal is running by metal all the time anyway, so it would become ionized and therefore be conductive.

My reasoning behind using deonized water was that it would prevent corrosion, not aid it, but I can see that was kind of stupid, because I didn't realized the whole issue you brought up. I think more people should be aware of this, since I'm sure I'm not the only one who was misinformed.

Anyway next time I fill it up, it will be with distilled and not deionized water, but like I said, it has probably equilibrated by now anyway so changing it now is kind of a waste of time.

[DragonTattooz]

Cool. I'm glad I was able to contribute something to this site.

Distilled water w/ your choice of corrosion inhibitor. Rock 'n Roll. :cool: