Must I use distilled water and where can I get it?

[MagR]

Hi

I'm building my first watercooling loop and need help with coolant. From all the research I've done the view is to use water not the proprietary coolant mixes from Phobya etc which cost a bomb and glow in the dark.

The problem I'm finding is that some say only use distilled (plus PT Nuke) whereas others say de-ionised (plus PT Nuke) is fine. I can find de-ionised no problem locally but can't find distilled. On the internet distilled is very expensive due to shipping.

Has anyone used de-ionised long term with no ill effects or will this be courting disaster? I plan to drain my loop fairly regularly for the first couple of years anyway.

Thanks in advance

Mag

[Fraz]

First time I filled a water loop, I got all obsessive about whether to use distilled or deionised water, etc etc. Nowadays I just use water from a filter-jug with a bit of pt Nuke. Works just as well. Frankly, I'd probably just use tap water if I didn't happen to have a filter-jug to hand.

If you are looking for cheapish distilled/de-ionised water, look for "battery water" in car shops. For instance, you can get 5 litres of battery water from Halfords for £4:

http://www.halfords.com/webapp/wcs/s...egoryId_255205

Definitely DO NOT USE those evil coloured pre-mixed solutions.

[kopite]

http://www.ehow.com/how_4728943_own-...led-water.html Make your own?

[peterb]

De-ionised will be better than distilled as it will remove all impurities.

Distilled water is, as the name implies made by boiling the water and then condensing the steam, so that dissolved minerals are left behind. De-ionized water is made by passing the water over an ion exchange resin which does the same job. Deionizing columns are much smaller than distilling plant.

You do need some inhibitor, mainly to stop the growth of micro-organisms that will inevitably get into the water from the air or that may be in the tubes when you fill them. Although de-ionised water should be free from dissolved gases, oxygen and carbon dioxide will dissolve into the water slowly. Carbon dioxide dissolves to form carbonic acid, and together with dissolved oxygen, forms a slightly corrosive mixture, so some corrosion inhibitor is a good idea. You could use a tiny amount of car radiator antifreeze, as that is a corrosion inhibitor. For the volume of water in a water cooling set up, you would only need a very tiny amount. Otherwise buy a product made for the job.

[Fraz]

Carbon dioxide dissolves to form carbonic acid, and together with dissolved oxygen, forms a slightly corrosive mixture, so some corrosion inhibitor is a good idea. You could use a tiny amount of car radiator antifreeze, as that is a corrosion inhibitor. For the volume of water in a water cooling set up, you would only need a very tiny amount. Otherwise buy a product made for the job.

You definitely do not want to buy one of those awful coloured pre-mix solutions that are (supposedly) "made for the job". THEY ARE PURE EVIL. If there are uncoloured, non-uv-reactive, non-gimmicky brands that are made for the job, then I'd say that was OK, although undoubtedly an unnecessary expense. De-ionised water and Pt nuke is fine. If you want your pipes to be colourful, buy coloured pipe, not coloured water.

I have no experience of adding a corrosion inhibitor except as part of the central heating system in my house! But probably an unnecessary complexity for a PC water loop in my experience. Temps are way less, most parts are either plastic or get replaced on timescales that are much shorter.

[peterb]

You definitely do not want to buy one of those awful coloured pre-mix solutions that are (supposedly) "made for the job". THEY ARE PURE EVIL. If there are uncoloured, non-uv-reactive, non-gimmicky brands that are made for the job, then I'd say that was OK, although undoubtedly an unnecessary expense. De-ionised water and Pt nuke is fine. If you want your pipes to be colourful, buy coloured pipe, not coloured water.

I have no experience of adding a corrosion inhibitor except as part of the central heating system in my house! But probably an unnecessary complexity for a PC water loop in my experience. Temps are way less, most parts are either plastic or get replaced on timescales that are much shorter.

All true. I wasn't advocating one of the gimmicky products, but one of the biocides. As Fraz says, the corrosion risk is probably very low, the risk of microbial growth is higher as the temperatures involved (with dissolved oxygen) is more conducive to that.

[Hicks12]

Halfords deionised (advertised as battery water maybe?) is the best value one to get, under testing it has been solid all around . Next step id recommend Mayhems fluid as he does a good job and its all filtered like crazy .

I would strongly advise against using tap water, you risk destroying the waterblocks and other components in the loop with all the added crap in there. For antibacterial dont use a silver coil as it does have some effects on the blocks etc, recommend just getting either ptnuke or mayhems biocide as they are both good (mayhems slightly better i think).

Thats it, you're ready to go .

[Fraz]

For antibacterial dont use a silver coil as it does have some effects on the blocks etc

Yes, this is galvanic corrosion. Most water loops are copper blocks, with some nickel-coated brass fittings. These metals all play quite well together. Silver is a lot more "cathodic" than any of the above metals, and so doesn't play quite so well, although it's just about acceptable. Aluminium and copper should be avoided at all costs, unless you want your water loop to be a 0.6 Volt battery!

[Pob255]

Are you sure it's galvanic corrosion?
http://en.wikipedia.org/wiki/Galvanic_corrosion

Galvanic corrosion is an electrochemical process in which one metal corrodes preferentially to another when both metals are in electrical contact and immersed in an electrolyte.

So unless you're wiring your blocks and rad's up to your psu then it's not galvanic corrosion.
you would get some galvanic corrosion from the water acting as both electrical conductor and electrolyte but you'd want a distinct secondary current path to get any serious galvanic corrosion.

The reasons for avoiding aluminium & copper is more down to the fact aluminium would bee used in cheap rads, so you've got a very large surface area to increase the potential of the little galvanic corrosion and the thin material of the construction.
Also being a lower density metal means it will get eaten away faster and the copper - aluminium potential is far greater than silver - copper.

[peterb]

If the silver coil is in contact with the copper of the block (ie electrical contact) then it will be galvanic corrosion. The copper and silver act as the electrodes in an electrolyte, forming a galvanic cell, and because the electrodes are touching each other, it forms a short circuit, so the electrodes will be depleted (ie corrosion) Why batteries (especially zinc carbon batteries) leak - the casing corrodes as the electrode is consumed.

You don't need an external current source.

Even if the two conductors aren't touching directly, the water will act as both a conductor and electrolyte and the silver/copper cell will self discharge, but much more slowly. (Why batteries self discharge in storage)

[Pob255]

But who puts a silver coil insider their block or rad? you stick it inside the res.
Also by not adding additional additives to the de-ionised water you're minimising the catalytic effect of dissolved materials.

This is one of those things that everyone knows and repeats but seems to lack grounding in science.

Now using fittings of different metal I can see being an issue, you'd not want silver or aluminium fittings screwed into a copper block or rad, but not a coil in a plastic res, heck if they're not contacting through the metal of the case then mixing an aluminium rad and copper block should be safe.

Current flows, the electrons of the anode material flow via the direct connection into the cathode, the electrons have to come from somewhere, namely the material of the anode, causing it to break down into ion's that then travel through the fluid to the cathode being attracted by the negative charge of the cathode, which neutralises the charge on the cathode, causing it to pull more electrons from the anode . . .
Without that direct contact the cycle cannot be started, ok, it can because we're talking fluid so it can act as both sets of connection because it is fluid, however is the resistance differentials in the dry contact and wet contact that sustain the cycle.

[scaryjim]

That said, the wiki article reckons a 0.25V difference is acceptable, and the difference between copper and silver is 0.2V, so the risk of corrosion from having a killcoil in your loop should be low (try to make sure it stays in one of the plastic tubes ).

[peterb]

But who puts a silver coil insider their block or rad? you stick it inside the res.
Also by not adding additional additives to the de-ionised water you're minimising the catalytic effect of dissolved materials.

This is one of those things that everyone knows and repeats but seems to lack grounding in science.

Now using fittings of different metal I can see being an issue, you'd not want silver or aluminium fittings screwed into a copper block or rad, but not a coil in a plastic res, heck if they're not contacting through the metal of the case then mixing an aluminium rad and copper block should be safe.

No argument with that, except that oxygen and carbon dioxide will dissolve into the water over time, forming carbonic acid and oxygen for the reaction, and there will still be a small galvanic current even if the coil ius not in direct contact with the other material. The effect probably isn't significant, but other biocides may be more effective anyway.

[Bugbait]

First time I filled a water loop, I got all obsessive about whether to use distilled or deionised water, etc etc. Nowadays I just use water from a filter-jug with a bit of pt Nuke. Works just as well. Frankly, I'd probably just use tap water if I didn't happen to have a filter-jug to hand.

Definitely DO NOT USE those evil coloured pre-mixed solutions.

If you're in a hard water area (eg. London) then absolutely do not let tap water (even if boiled) anywhere near your water loop. I used filtered water out of haste for my first build in the UK (living in London) and the tubing instantly clouded with a chalky residue inside the blocks. You know that lovely limescale you get near all your taps and wet areas? You don't want that inside your closed loop.

Distilled and biocide seems to be the best. I personally use what's let of my Feser 1 and have been for many years and haven't experienced any issues apart from staining, which doesn't really bother me. Not the recommended way for most but beats plain tap water any day.

[Fraz]

Are you sure it's galvanic corrosion?

The flow of charge can occur through the water. Obviously deionised water helps massively in preventing this, but it's never going to totally stop it. And what do you think the biocide (copper sulphate) provides when dissolved in the water? A source of ions. This is why I largely don't worry too much about deionised water and use Brita filter-water instead, as I'm only going to go and add some ions to it anyway

All in all, it's exactly the same principle as a potato battery, or a "lasagne cell", etc. As long as there are some ions in the medium connecting the two metals, then you will get a movement of charge, and galvanic corrosion will occur.

[DanceswithUnix]

The flow of charge in one direction happens through the water, for galvanic corrosion you need to complete the circuit usually by touching the two metals together.

Silver sounds amusingly exotic though, are some of the bacteria vampires?