Sign in with
Sign up | Sign in
Your question
Solved

Newcomer, best cooling liquid for minimal maintenance?

Last response: in Overclocking
Share
a b K Overclocking
January 18, 2013 3:13:40 AM

Hi guys

Great stickies you guys have up in the forum, been reading through them and have learned a lot. I plan on switching from an high-end air cooled rig to a custom water-cooled one sometime over the next couple of months (depending on work/vacation, budget etc :lol:  ). but first all the talk about cooling fluids got me curious.

I'll qualify this question by saying that I'm a trained physical biochemist specializing in nano-particle design but also dabble a bit in chemistry and basic engineering. I can't go into details about what I do due to legal reasons, but I'll say that it is in no way related (at least for now) to computer components or cooling, this is purely a hobby. now between stuff I've read on TH and overclockers etc, it seems like the consensus is that distilled water + copper sulfate (PTNuke) +silver (kill coil), and that the more additives you add, the more often you'll need to flush/rinse/refill your loop due to corrosion, clogs, erosion etc etc.

so in theory, if I say flush my loop with pure ethanol, fill my loop with de-ionized water (more using the process to kill any possible living thing in the water rather than getting rid of ions, I know the water will strip ions from my blocks etc), UV and/or radiation sterilize my filled loop, would I then have a maintenance free water-cooling loop?

My thinking is that since the loop is filled with clean stuff, say I go out of my way to find copper only components for the whole loop, then all i'll have is a few copper ions in my water which will never cause me any problem right?

P.S. the reason I want to do this is that I actually have a really hectic work/life schedule and cannot guarantee a semi-regular maintenance cycle. However, I am willing to even take a few days off work and spend a few extra bucks to reduce the maintenance cycle to say every 2-3 years, which will coincide with my upgrade cycle

Thanks in advance!
a b K Overclocking
January 18, 2013 3:44:39 AM

if that is what you are looking for than i have a answer, that is as well what i try to do, and yes all copper would be the way to go here. if you like have a look at my build and see what i have tried to do to make it easier to maintain my loop.


http://www.tomshardware.com/forum/274855-29-experimenta...
m
0
l
a b K Overclocking
January 18, 2013 4:42:29 AM

toolmaker_03 said:
if that is what you are looking for than i have a answer, that is as well what i try to do, and yes all copper would be the way to go here. if you like have a look at my build and see what i have tried to do to make it easier to maintain my loop.


http://www.tomshardware.com/forum/274855-29-experimenta...


Hey man, thanks for the quick reply, glad to have a comrade in arms working toward a similar goal :lol: 

as I slowly work my way through your epically long thread, I must ask, did you actually end up using the water wetter and ethylene glycol? while I agree that the small hit in cooling performance isn't a big deal, I'm honestly a bit surprised that nothing clogged up in your 4-year experiment in the past. but what do I know, you're the one who did the experiment and really I'm impressed :whistle: 

That said, I think I've seen the type of copper radiators you got in some old scientific instruments laying around in storerooms at work. they're typically used in large arrays for equipment that need to rapidly heat and/or cool something. I would say rapid automatic thermal cycling for various reactions. sometimes companies don't label these parts because the units are custom built, and obviously the exposed coils/somewhat crude designs are not issues for an industrial instrument. Just throwing in my guesses here :) 
m
0
l
Related resources
a b K Overclocking
January 18, 2013 4:50:19 AM

Yes that is what I use now to I do flush it once a year or less depending on the amount of particulates I see floating around in the res.
m
0
l
a b K Overclocking
January 18, 2013 5:21:30 AM

toolmaker_03 said:
Yes that is what I use now to I do flush it once a year or less depending on the amount of particulates I see floating around in the res.


AHH, I see that you've even added an inline filter to your current set up... man, I need to stop doing engineering under a microscope once in a while...
m
0
l

Best solution

a c 176 K Overclocking
January 18, 2013 5:43:49 AM

vmem said:
so in theory, if I say flush my loop with pure ethanol, fill my loop with de-ionized water (more using the process to kill any possible living thing in the water rather than getting rid of ions, I know the water will strip ions from my blocks etc), UV and/or radiation sterilize my filled loop, would I then have a maintenance free water-cooling loop?


If you do flush the loop using ethanol, be careful that you don't include acrylic (typically with reservoirs) or any other susceptible plastics (Acetal?) in the loop. Ethanol has a nasty habit of cracking acrylic. Dont know of any reservoirs that don't use acrylic is some aspect of its construction, so you may have to use a T-Line instead to fill the loop and bleed out air.

With the water, you dont need anything more special than supermarket distilled. Any bacteria/algae will be killed by whatever biocide you use on the loop, whether its copper sulfate, UV light or silver. Even the copper of various blocks will inhibit bacteria/algae similar to a kill-coil, though to a lesser degree.
That wont mean you will have a perfect non-maintenance loop. There will always be the potential for corrosion, simply because the radiator will have likely been soldered together using Tin and you cant get barbs made of Copper.

That being said, you wont have to maintain this thing very often and if you go to the measures you describe, I suspect corrosion wont be an issue for the useful life of the parts. Suspect maintenance would be just replacing the water every year or so, which if you set up the loop with drain valves and such can be a fairly easy process.
Share
a c 324 K Overclocking
January 18, 2013 6:10:14 PM

You can easily find biocides at your local supermarket or pet store that will inhibit growth in your system. I bought a bottle at Walmart for about $3. If you really wanted you could simply mix the entire bottle into the volume of distilled you were running.

You might wish to look into some corrosion inhibitors as well. Not all coolants that provide color are necessarily good at this, so you might look around a bit. Swiftech Hydrx comes to mind, but you could also run a diluted mixture of normal glycol antifreeze...something like 20%...maybe even less.
m
0
l
a b K Overclocking
January 19, 2013 12:28:23 AM

rubix_1011 said:
You can easily find biocides at your local supermarket or pet store that will inhibit growth in your system. I bought a bottle at Walmart for about $3. If you really wanted you could simply mix the entire bottle into the volume of distilled you were running.


I was honestly trying to avoid doing this because I thought doing this was a minor source if problems with water cooling, as most common biocides are metal-ion based (and as manof pointed out any source of copper etc in my loop will actually contribute). In the long run, these dissolved metal ions start to displace other metals in my loop and can initiate corrosion, or they may bump into some nucleating nano-particles in solution and start to crystallize, eventually forming particles that might clog up my rads or something. If it were possible I would really like to see if I can mix some non-metallic and thermal-stable biocide in the lab, but alas it would not only be inappropriate, but mysterious toxins should not be brought home :lol: 

@manofchalk Thanks for pointing out possible acetal-OH reactions... I somehow managed to forget they make res out of acrylics... Yeah, I think I'll give up on flushing the tubes with alcohol idea, and flush it with semi concentrated biocide instead, then pour in sterilized di-water and UV sterilize the whole loop overnight or something (I'll be borrowing that industrial UV sterilization lamp and some safety glasses :p )


---------------
hmm, after doing some more reading, it seems I was right that PT-nuke is partially responsible for small amounts of garbage forming in people's loops. copper sulfate, or the active ingredient in PTNuke, is known to form copper hydroxide when mixed with small amounts of sodium hydroxide or aluminum chloride, both of which are found in low quantities in regular water. the resulting copper hydroxide is pretty much not soluble in water, and has the appearance of a blue-greenish sludge (hmm, were have we seen that before? :pt1cable:  )
m
0
l
a b K Overclocking
January 19, 2013 12:51:40 AM

manofchalk said:
Why not just stick a UV light inside the rig permanently? No chemicals in the water and only a small risk of skin cancer :lol: 
http://www.frozencpu.com/products/2992/lit-23/12_Cold_C...
As long as your tubing and reservoir is clear it will work fine.


well, I've never bought that cold cathode tube specifically, but I believe all consumer UV products (tanning beds aside... eww), use long wave-length UV to minimize risk of skin cancer and damage to human eyes. This type of UV is actually not very good at killing stuff, though I guess sticking one in the case after proper sterilization can't hurt. the one I wanted to borrow from work emits 240nm UV, which is on the short side of mid-range UV, and would cause serious damage to all living things if exposed to it for any extended amount of time, specifically, any micro-organism's DNA would be completely destroyed within about 40 minutes... you actually need safety training for this type of equipment as you can easily sustain permanent eye damage without wearing the right safety goggles
m
0
l
a c 176 K Overclocking
January 19, 2013 1:07:46 AM

I think your over-thinking this. You want as non-maintenance as possible loop, and are willing go to the extreme of blasting the rig with cancer inducing, eye melting UV to do it.
Most water loops running plain distilled, a kill-coil and not ridiculous choices of metals can run without maintenance for two years or so.
http://martinsliquidlab.org/2012/01/02/distilled-water-...

I really think you could just a normal loop and still get what your looking for.

m
0
l
a b K Overclocking
January 19, 2013 1:23:32 AM

manofchalk said:

Most water loops running plain distilled, a kill-coil and not ridiculous choices of metals can run without maintenance for two years or so.
http://martinsliquidlab.org/2012/01/02/distilled-water-...

I really think you could just a normal loop and still get what your looking for.


Hey thanks, honestly the first time I heard that such a solution can last 2 years (aside from toomaker's experiment). anyway, I'll probably be popping up with a list of parts at some point to run buy you guys and we'll see, call it an experiment.

Oh, btw, I completely agree with you that I'm overthinking this mano, call it a professional habit :lol: . otherwise I think this has turned out to be a cool experiment, I've always taken pride in my sterile techniques at work and we'll see how well I can apply it to my PC building heh
m
0
l
a b K Overclocking
January 19, 2013 2:02:52 AM

Best answer selected by vmem.
m
0
l
a c 324 K Overclocking
January 19, 2013 2:44:08 AM

I've run a loop for more than 2 years without flushing without any issues...and not all biocides are based on metal sterilization...some are chemical.

vmem posted a great link...and is right on with his statements. We tend to make a bigger deal out of it than need be, but if you are running plain distilled w/ killcoil and/or biocide you can very likely go much longer than someone with a coolant, etc in their loop.
m
0
l
!