Has anyone ever watercooled their Ram?
 

I am thinking of water cooling my ram, Me, being the adventurous-stupid type I started poking around the inside of one of my computers to get a "feel" for the hotspots while running the burn-in on Sisoft Sandra.

My ram seemed the hotest, non-heatsinked component at this time. I don't have heat spreaders on my ram and was thinking of getting some when I read that most are just sticks of aluminum or copper. Well I gots plenty of Aluminum and am thinking of sandwiching my ram between two pieces of 1/16" aluminum and attaching it to some 3/8" or 1/2" aluminum tube by using a product called Alumite (there's all brands of this stuff, mainly being 98%Zinc/2%aluminum soldering rods).

I am wondering if anyone knows the thermal conductivity of zinc and if this is a worthwhile method. I think It would probably be best to anodize this as well as that will stop any electrical conductivity. My main concern on anodizing it would be what kind of contamination zinc will bring to my acid baths.

Anyhoo if anyone has some insights on water cooling ram, I'd like to hear(read) it.

It's been done before... I can't remember who, but someone sells ram waterblocks... it seemed pointless at the time.
if you have copper in your system, forget about using aluminum.
-Zoson

I mixed copper and Al for months without corrosion - you just need to make sure you have some anti-corrisive additive in your loop.

Or anodize it. Anodizing aluminum should stop the galvanic reaction, as it makes a uniform layer of aluminum-oxide, which does not conduct electrical charge. Anodized Aluminum and copper would be fine together in the same system.

RAM cooling has been done, the reports I've seen on it have been mixed at best, with some folks claiming slightly increased O/C's and/or better stability (especially when over volting their RAM) but many others reporting no significant change. OTOH, I've not seen any reports that it hurt anything aside from those who damaged their parts while attaching heat sinks or attempting to remove existing heat spreaders. Also if you use multiple banks of RAM, many folks report problems making everything fit, hardly a surprise considering how close together the modules usually are.

As to the mechanics of it, my comments on annodizing are...

1. Annodizing SLOW'S corrosion, it does not STOP it!
2. Any flaws in the annodizing will continue to corrode, as will any scratches or other damage that exposes raw aluminum.
3. It is difficult, verging on the impossible, to create a complete uniform annodizing coat on the inside of a small diameter tube - since the coolant flows through the tube, that is where the corrosion is most needed, but hardest to arrange.

Unless your system is already all or mostly AL based, I would think it might be better to use two peices of copper soldered onto a peice of copper tubing. This would be as easy or easier to make, and argueably more effective to the extent that ram cooling is at all useful.

Gooserider

I was thinking of doing it with 6mm OD copper piping but there is no where I can buy a 1/2" -> 1/2" + 6mm OD splitter, 6mmID tubing (flexible), and 6mm od connectors etc.. so I havent got the parts to make it. Instead I have made 2 double sided ram heatsinks out of alu which are 1/4" thick and fit quite well. they are held together by 2ba screws and have normal thermal paste betwen them and ram. Without a fan I think they will still get very hot so I am designing a mount to stick a fan onto them.
I used the copper pipe method to cool my 9700 gpu before I changed to a normal top.
How do you anodise alu?

Isn't this the reason most of us started w/c'ing?

My ram currently has no cooling devices attached, and if they did I would just be patient daniel-son when removing any heat spreader adhesives. Fit will not be a problem...only cooling two sticks of ram, 1 gig of ram should be plenty, and they are seperated into the two hi-speed dual channel sockets(blue on asus boards).

True, but using distilled water or even tap-water the anodic layer will outlast the current physical design of ram and maybe even outlive me!

Considering anodized 6061 aluminum is an 80 on the Rockwell-c hardness test scale, I would think any plastic packaging on the ram or any metal parts in any computer would have a hard time scratching it.

Difficult for whom? I find it rather easy, but then again I have been anodizing for years and I actually calculate my surface areas, test my acid baths, etch/clean/degrease/desmut my parts and adjust my rectifiers according to my calculated time/current ratio.

So then why anodize?

High Corrosion Resistance
Considering "most" peeps in the w/c hobby are using distilled water there is an extremely low occurance of anodic-layer deterioration. "we're not water cooling with ocean water here, and no, we don't need a bigger boat."

Decorative / Cosmetic Appearance
I just looks pretty dammit!

Insulator / Nonconductive
Anodic layers of Aluminum(Aluminum Oxide) are nonconductive thereby reducing if not eliminating galvanic reaction.

Stain Resistance
Not all that important but see above: Decorative / Cosmetic Appearance

Increased Durability / Scratch Resistance
In case you want to take your pc to the streets, there less chance of scratching your block if you dump...you might get road-rash, but your block should still stay pretty dammit :)

But it just won't be pretty dammit! :D

-MC

Not always... It is also a good solution for those of us into SILENT Computing, even if we have no intention or interest in overclocking.
Not saying that you're a klutz, or that your hardware wouldn't work, just reporting on the problems I've seen reported...
Depends on who you talk to - I recollect a thread here not that long ago where the annodized top on a swifty block failed in less than 6 mo. BillA attributed the failure to galvanic corrosion due to use of distilled water w/o an anti-corrosive.
I don't particularly CARE about scratches on the outside, that doesn't touch the coolant and the dry bits aren't going to corrode anyway. I do care about scratches and flaws in the areas that come in contact with the coolant, as those WILL corrode. Either way, my experience has been that annodizing is better at resisting damage than paint, but it isn't all that solid. Contact with tools damages it quite readily.
According to every article I've read on the subject of how to do anodizing. According to them, there is a problem with getting uniform electrical conductivity on the insides of hollow objects (like tubes) as the current tends to prefer the shortest paths to the electrodes. Since the depth of the anodizing is a function of current flow, if the current is irregular so is the anodized layer. This also matches my observations of annodized parts that I've examined over the years. (Note that it is also difficult to do QA on the inside of a small bore tube)

See my comment above - experience varies. Speaking of boats, I also have a fair bit of marine experience, and have observed that anodizing does NOT do all that well at extending the life of aluminum near salt water, I would say 10% max, less if the object gets much wear and tear, more if it's cleaned, polished and waxed regularly.
Perhaps, but then so does polished copper. IMHO alot of anodizing looks kind of cheezy to me, like excess make up on an woman trying to look younger than she really is.... But then I've never been big on cosmetics.
Again, mixed bag...
See above comments about cosmetics. I'm a biker, and I run a total rat bike (A Moto Guzzi) with over 100K miles on it, I always tell the Harley folks that "Soap & Chrome won't get you home, but it's nice to look pretty while waiting for a tow...."

Opinions vary...

I have no particular desire to cool my RAM, but if I were to do it, I would probably get a peice of Cu tubing approx as big in OD as the thickness of the modules and as long, plus enough extra to fit tubing on each end. I would then solder a length of RAM sized plate onto each side of the tube making a sort of 'tong' setup that I could slide over the module. Polish it up and put a layer of clearcoat on it, and I think it would look real nice and perform as well as any other product.

Gooserider

I have a project that could use something like this...
Its a server with 20 DIMMs, using 2GB DIMMs.
The things burn 25W each, getting the heat out
of a 1U chassis is going to be a big problem.

They want an upgrade path to use 4GB dimms..
(shakes head..)

why go into Alu when u can use a thin copper, remember if u seen some CPU blocks w/ a thin copper plate installed. This is wer creativity comes in.
#rotor afaik made one either Alu or Cop.
another problem u will get into is the flexing of the hose might, will move the RAMs.
I cant wait for ur positive results w/ this type of approach. :)

Now that's what I call going to xtreme. :drool:

Note that Starman97's project is a SERVER, not a personal machine. Judging by the reference to a 1U chassis, I'm assuming it's a rack mount, and therefore probably one of several machines. (or possibly driving a big drive farm)

I'm not going to ask him to divulge anything he doesn't feel comfortable with, but it sounds like a corporate application, that involves a big database and lots of users. Something like that, response time becomes a big issue, and there are lots of advantages to keeping as much of the data in RAM as possible. Another possibility that occurs is a graphics 'renderfarm' for movie special effects.

So yes, it's extreme for most apps, but not for some of the industrial strength computing that is so important to modern life even if those not directly involved in it seldom hear of it.

Assuming he can sell a WC solution (not sure it would be possible) I would think that it would be possible to make a series of the copper 'clamp' design I mentioned earlier, probably in some sort of series / parallel array which would deal with the heat issue quite nicely.

IMHO the big challenge in such an application would be to come up with a 'fool proof' method of changing out either the entire drawer or some part of it in a quick way that didn't allow for ANY leakage of coolant. The method would have to allow for very rapid service (such systems often have 'Five 9's' (99.999%) uptime requirements, which translates to about 90 MINUTES PER YEAR when the overall system can be down, this includes routine maintainance, upgrades, and so forth as well as repairs) by field service people of unknown (but must be presumed low)competence in the presence of nervous client management. Keep in mind that this kind of machine is likely to have a cost in the millions of dollar range, and be in a computer room with even more expensive hardware, and you understand why leakage is not an acceptable risk...

Gooserider

talk abt challenge

nice info on different types of xtreme hardware and uses
gosh i'd love to see things like that and type of apps it's running :drool:
:D

On the other hand, you'd have a cooling budget of at least 5 figures. ;)

Now I know where the "gooserider" name comes from! I ride a 98 Superhawk, its got 30K on it so I've got a bit of catching up to do.

Gooserider could aluminuim be plated (with copper?) instead fo anodising? would that work better or worse?

I had an idea about this. You could buy some of the copper heat spreaders for RAM. Then you could solder on a small box like the one here http://www.overclockers.com/tips1115/

Ok so the heat transfer to water wouldn't be as effiecient as it could be but is would be pretty good.

Also want pics once you implement a design please.

Buzz

I am not totally up on what can be plated on what, but I don't think it is possible to do Cu on Al.

I certainly wouldn't consider it a good solution for constant immersion since any flaw would really open you up for corrosion problems. Generally, I just feel it is best to avoid putting Al in the system at all, as it keeps life much simpler.

Gooserider

look arond at the heatspreader market and try to design something similiar. one half of a heat spreader so you can clamp two together for doudle sided RAM or clamp a "blank" one on for single sided RAM. but basically use the same design for each a thin flat tube to clamps acrosss the chips and that's all you need.

o
|
| |
| |
| |
| |
| |
| |
/
o

Or if this is just in a server get a load of good quality heat sinks for the ram and change out all the extractor/intake fans to higher cfm ones (such as deltas).

Impossible. By definition, Anodization is a uniform layer of aluminum oxide. Aluminum oxide does NOT conduct electricity thereby removing all posibilities of galvanic reaction. On top of the fact that the test model in question also used distilled water, also a non-conductor of electricity. There had to be something left out of the scenerio for that to happen. I have never owned any swifttech products, but I would venture to guess that the failed block was either annodized in someones garage or was actually powder coated. For a uniform dye of hard-coat type II anodization the layer NEEDS to be uniform to produce a uniform color. Any irregularities of the dye-job would instantly point out any flaws in the annodized layer.

who mentioned the outside, water-blocks can have inner-beuaty. :D


How did you scratch the inside with a tool and why do you need tools on the inside of a block/tube etc? You lost me on that point.

First off you need to understand something about current flow. Electrons actually flow on the surface area. The tube has surface area on the inside as well as the outside. I wouldn't consider i.d. of 3/8 or 1/2" tubing to be small bore. If the tubing were 1/8" or smaller there could be problems with to high current flow actually burning/pitting the surface ares. With 3/8" or 1/2" its not hard at all...constant current, the acid bath needs to be agitated, anodization is complete when the part no longer registers current flow through the rectifier. Electricity is similar to water, it takes the easiest path. Anodization(Aluminum oxide)=no path, therfore if the surface area was calculated correctly the part would conduct untill there was no more exposed aluminum...done. Agitation can easilly be accomplished with a simple airstone like aquarium shops sell. I use a fume hood so there is no problem of the off-gassing of hydrogen from the air agitation method.

I've never tried to annodize a boat and I have never water-cooled with salt-water.
my boat reference was to the movie "Jaws"..."I think were gonna need a bigger boat."

Polishing usually shows pride in ones possessions...so considering its a rat-bike I am reminded of a quote I once heard..."Its hard to polish a turd"

Its getting warm enough here in the northeast, should be 80F tomorow, I can fire up my annodization tanks without spending a fortune heating the place up for just a few parts...I'll snap off some pics for posting.

I am not talking about annodizing an entire boat, however I am talking about annodizing marine fittings like cleats, lights, railing trim, etc. The parts look great when new, but after only a few months have developed pits and bubble where the anodizing has flaked away. What happens is that corrosion starts in any flaw, however microscopic, and then extends under the annodizing until it forms a bubble of corroded aluminum under the annodizing. The annodizing which is more brittle than the aluminum then breaks away, exposing the surface to more corrosion and the process continues on an accellerating basis.

As to 'rat bikes' and polishing, I figure that time is a finite resource, I can spend it on polishing, or on riding. I'd rather ride. I live in MA, and haven't owned a car since 1980. I've never missed a day of work due to bad weather when work was open. The road salt in that environment is hard on a vehicle, so polishing to show quality isn't really an option. Instead I go for award winning in the other direction.

As to scratching the interior, the damage which I mentioned on the Swifty block appeared to have been caused by simply installing the barbs. That is enough to scratch the annodizing and create a gap where corrosion can start. I don't have any hands on w/ Swiftech stuff, but they are supposed to be a relatively major company with a good reputation for quality. If they say they use blue annodizing I'm inclined to believe them.

Gooserider

Hello fellow massholes...another doozy of a quote I once heard while vactioning in FLA.

I myself also live in Taxachusetts. More power to ya on the bike riding thing then. I've been here since the blizzard of '78 we do get our fair-share of snow-pounding. My friends make more money than they know what to do with...damn them...all but two of us currently own motorcycles, they are the snobbish types with all the chrome and such (my electrician friend enjoys the polish a turd phrase). I being one of the non-cycle types. Well now anyways, I hit pavement at 40MPH whan a cute fuzzy woodland creature decided he wanted to see what my front tire looked like up close and personal. That sucked. Me-0, Raccon - 1. We'll call it a tie...the racoon bit the dust.


I see what you mean by the boat parts...any part subject to saltwater/abrasion will basically get ruined. I live in western MA, so I don't see much salt water, however I've seen many an outdrive all pitted up after one year in the ocean. I think they are actually painted but the idea is the same...any flaw, be it inherrent, scratched or wear will allow the corrosive sea water to do its thing to aluminum.

I know nothing of swifttechs practices but am willing to take their word on it since they have been around for a while, just like I did with Ford and Firestone...oh wait that's not a good example after all.

I don't worry much about scratching the aluminum, I order black nylon connections for my aluminum stuff and brass for my copper, mscdirect dot com carries all that kinda stuff & o-rings and flat rubber-sheet to boot.

I just like how easy aluminum machines. I prototyped all my cooling stuff in aluminum and it seems like a shame to chuck em in the recycle pile just cuz thay're not copper, especially since I had enough prototypes to make an all aluminum setup, except radiator. Plus I like annodizing them to make them match my case color.

-mc

Since the aluminum oxide does not conduct electricity, how do you get your anodization more than a molecule thick? In a similar vein, how do you think your pigments get into and stay in your anodization layer?

Just as you can run more current through to increase your depth of anodization, galvanic currents can flow. No matter how non-conductive your aluminum oxide is, andozing produces a porous layer of it.

In exacty the same way that you get more oxygen to the raw metal to anodize deeper, in exactly the same way that your dye penetrates the surface, ions can and will cross anodization to cause corrosion. That layer of porous aluminum oxide will only slow, not stop, corrosion.

Either use anti-corrosives or don't mix your metals.

Annodization is a porous layer of aluminum oxide. It is similar to a honeycomb. This is where the dye "attaches itself." It actually collects in the microscopic pores of aluminum oxide.


Once sealed the product then becomes electrically insulated. Annodizing is a process...one that must be done strictly or poor results will definately be the outcome.

Some places to look for annodizing information:

This site is an incredible sorce of information! And now he even sells Supplies...Great site.
http://www.focuser.com/atm/anodize/anodize99.html

This site has supplies and kits and their book on anodizing is second to none. Very easy to understand.
http://www.caswellplating.com/kits/aluminum.htm

Good source for plating information.
http://www.finishing.com/

The US governments answers to epa regulations of metal finishing products and plating techniques.
http://es.epa.gov/cooperative/topics/metalfin.html


I hope this clears up any misconceptions.

-MC

MC, I agree that if you have a bunch of Al parts, it makes sense to use them, I would just use them in an all Al system. Pumps are a non issue since most of them are plastic and don't care, ditto for the res. Al rads are also easy enough to come by, they are used on alot of the newer Al block cars and on motorcycles (though it is hard to get a good used MC rad, they tend to be 'high casualty' items)

I don't consider all Al systems to be at all bad, though they aren't going to do quite as well as a Cu/brass system (all else being equal). It is only the dreaded "mixing of metals" that I think is evil and should be avoided.

Gooserider