Recommendations for coolant for a mixed metal system

[zogthegreat]

Hi everyone:

My buddy gave me his Power Mac G5 2.7 Quad computer the other day. Since current versions of Mac OS X won't run on it, I am planning to install either Debian or Gentoo Linux on it, pop in an old 3Ware 8 port raid card and 8 x 500gb laptop hard drives. I will use this for a file/minecraft/media server for my home.

On initial teardown and inspection, I discovered that the system was water cooled, (my friend did not know this after 8 years of use), and that the water cooling system had been leaking for some time. It had lost about half of it's coolant, (The recommended fluid amount was 180 ml and I only drained out 83 ml). I also found out that for some reason, Apple had decided to use an aluminium radiator and pipes with water blocks that are made out of a nickel plated base with copper fins attached. I have cleaned the system and I am ready to reassemble it, however I thought I would solicit opinions and recommendations for fluids. Mac's spec sheet calls for 66% distilled water and 33% glycol, (antifreeze).

Does anyone have any suggestions for a better coolant?

Thanks

zog

[Stewie007]

Water + Anti-corrosive agent. That should suffice. Make sure to use the appropriate level of corrosion protection as recommended on the bottle.

If you use glycol, just make sure that it is inhibited glycol. Otherwise, it will degrade and form acids that will eat your system components and blocks. If it is meant for coolant, chances are that it has an inhibitor. Though, you should verify this.

Dissimilar metals alone are not enough to present caution. Electrical contact is required for galvanic corrosion. That means if you use copper fittings on aluminum parts, then you should expect corrosion of that joint in the future at some point. Though, if the metals are isolated from each other, you shouldn't be looking at a sieve in the next few weeks. Poor system design usually brings the exhibited problems. Proper fluid choice and appropriate maintenance are key.

[zogthegreat]

Water + Anti-corrosive agent. That should suffice. Make sure to use the appropriate level of corrosion protection as recommended on the bottle.

If you use glycol, just make sure that it is inhibited glycol. Otherwise, it will degrade and form acids that will eat your system components and blocks. If it is meant for coolant, chances are that it has an inhibitor. Though, you should verify this.

Would regular automotive antifreeze, mixed 33/66 with distilled water do? Sorry for the noob question, been water cooling for some time, just never had to deal with this something like this before.

Poor system design usually brings the exhibited problems. Proper fluid choice and appropriate maintenance are key.

Yeah, I was going to go with a T-line to fill the system, but I found an old 80 mm tube reservoir in my "box". I cut the top off and cut it down to 30 mm. I then glued the top back on with some IPS Weld to get this:





Its not pretty, but I gave it a leak test it and it seems to work fine. My son just got a 3D printer, I might design a better reservoir in Google Sketchup later if I can mange the energy usage of the system. It draws 140 w at idle!

I might even do a custom radiator out of copper like I did for this project:





This was for a custom water cooling setup I did for one of my rackmount server cases. I cut down a 120 mm radiator to 120 mm x 69 mm, (177 x 69 with the fitting mounts), and then made the shroud with acrylic.

[Stewie007]

Automotive antifreeze is basically ethylene glycol with additives included. Acute toxicity is high, and gloves should be worn while handling. Spills cleaned up immediately. I personally like Propylene Glycol for that reason, but in my chiller it does tend to exhibit higher head pressure because of that. That is typical of PG in a low temp application. In your typical water cooling system, head loss shouldn't be that measurable in 33% solution.

Inhibited PG100 costs about $25 or so at a refrigeration shop.

[Rinaun]

I personally recommend Car Antifreeze (I use Mercedes Benz orange from my local car shop). I put maybe a shot glass into my loop and I've had several loops setup for clients that required zero maintenance for 4+ years, including my own loop. The temperature differentials between using glycol versus just anti-corrosion inhibitors is so trivial and antifreeze won't fail over time.

[graphic harbour]

distilled water with silver coil ........

[Stewie007]

Exactly, the difference is very small. So there is no reason to use a toxic chemical. Just use something with limited toxicity and a corrosion inhibitor.

[Rinaun]

Exactly, the difference is very small. So there is no reason to use a toxic chemical. Just use something with limited toxicity and a corrosion inhibitor.

Well for one, using only corrosion inhibitor doesnt help for other issues so you need a killcoil as well. With antifreeze it's an all-in-one solution available in almost every color. I like using the car antifreeze because it has other things inside of the composition that helps keep gunk out of my loop. I've also never had any issues with plastics regarding antifreeze. I just find using a killcoil and corrosion inhibitor to be silly when there is industrial-level solution compared to "drop a silver coil inside, some drops of this stuff and hope it's fine in the long run". Car antifreeze has proven itself to be completely fine, and you can even buy food-safe versions IIRC. I have YET to see a loop with glycol fail like I have when someone adds too much or too little PHnuke or other additives.

The only reason people should be against glycol cooling is the MINIMAL decrease in the waters ability to dissipate heat and it stains tubing (as does every single fluid with dye in it).

[Church]

silver coil never was corrosion inhibitor, it's to lessen algae growth because of silver antibacterial properties. No wonder that is used in loops where corrosion shouldn't be much of an issue (all-copper / copper,nickel,brass) by those wishing for less of misc additives worsening thermal properties of distilled water coolant and possibly leaving some residue.
Mixed metals loop for OP of course needs coolant with anti-corrosion properties. And just like others mentioned, car anti-freeze is among best choices due it's relative cheapness, wide (and local at that) availability, good enough thermal properties and due experience of many LC users that also had used antifreezes for mixed metal loops (eg. for ones with more of DIY touch, with alu rads from room heater batteries or car rads, or for Zalman LC kits with alu components and god-awful coolant included in set). Just worth to check if antifreeze is of propylene glycol base (less poisonous then ethylene glycol ones).

[Stewie007]

I think the topic has been well covered.

Though, I do not make it a habit to give advise without mentioning possibly catastrophic contingencies no matter how remote they may be. The recommendation of EG is valid. It contains anti corrosion additives, and it also will prevent algae growth. It is an all in one solution. Though, I just don't like the idea of using toxic substances if I can avoid it, so I recommended PG (which is available for about $25 bucks at URI or CCDickson). Tractor Supply has it, but it is not inhibited. You would have to add inhibitors for additional cost. Though, the price I quoted for it was $21.

I am just not in the habit of providing advice without a bit of CYA added. I'm not going to tell somebody to use Glycol without warning them of the possible break down into acidic substances which can be a problem at higher concentrations. So, maintaining the system is important. You can test for acidity with litmus paper. If anyone uses EG without maintenance, then I have no sympathy for anyone that comes up with a problem.

Just sayin'. Always CYA. Your IP Address is linked to your advice. Probably nothing to worry about, but it's a good habit to be in.

As far as anti-corrosion agents. If you are using the correct concentrations, it should prevent algae. I ran my chiller when it was 60 - 80 degrees (the perfect conditions for algae). I used Purple Ice in proper concentrations. No amount of algae showed up when I drained it. Not even residue from dead algae.

That said, who can contest a bit of extra precautions? I can't.