• Welcome to TechPowerUp Forums, Guest! Please check out our forum guidelines for info related to our community.
  • The forums have been upgraded with support for dark mode. By default it will follow the setting on your system/browser. You may override it by scrolling to the end of the page and clicking the gears icon.

To distill or not distill what say ye?

Do you run only distilled water in your loop?

  • Yes, pure H2O or bust!

    Votes: 18 31.6%
  • Of course not, are you kidding me?

    Votes: 22 38.6%
  • The only time I water cool is when I spit on my air cooler.

    Votes: 17 29.8%

  • Total voters
    57
I will always call out companies for having bad platting. EK GTX 10 and RTX 20 series was horrible. Within a month the platting came off. CPU block for EK is still good 8 years later. Plating fully intact with distilled water.

Alphacool block GPU replacement, no problems. So for those saying im blaming a company for poor plating instead of following science Is missing the point.
Funny how after 10 years with proper coolant, I have all my plating on my EK 1080 block.

It is real simple. Don't etch it with vinegar. Don't flush it first day for no reason. Don't use distilled water. Assemble it, use proper coolant, change the coolant every 1 or 2 years per manufacturer's recommendations, and never have 'bad plating' ever again.


See how 50% of poll respondents are ignorant?


Let's play devil's advocate and say some blocks are better than others because they have thicker plating or more complete plating. Fine. I'm sure it's true. But does that mean your not ignorant for ignoring science? No. If you ignore science, you will pay the toll. How much you will pay the toll depends on all the other factors.
 
You know, we have a century of experience with radiators in automobiles and airplanes.

Probably best to use the same fluids, because you know - they’ve be proven out with literally billions of hours of data.
 
No hands-on experience, but... I'd use distilled (only, or w/ wetter/biocide) for a dedicated benching loop. Something that'll be drained and reconfigured often.

I always presumed some (pedantically-researched, prior to ordering) low/no silicate automotive coolant (50/50 or ~15-25/85-75) would be 'best' for long-term installs.
Cars/trucks sit (algae, mold, etc mitigation), are implicitly multi-metal (galvanic corrosion mitigation), and their water pumps require lubrication.
 
It's like we never went thru the EK fiasco...
 
Back when I watercooled (god I think that was on LGA1366? X58 and my 6-core westmere probably) I always mixed with what I only remember as being sold as "blue germam automotive coolant" from some webshop. I'd mix that with distilled. As you may sense, my memory is hazy, but I can say it worked well for almost 5 years and in yearly cleanings / flushes no algae was really present.

Never tried pure distilled.
 
Who watch the whole video when he said nickel and distilled water isnt a problem... It's only when zinc is added to the mix that you can have corrosion.

The ratio of brass to nickel is something to consider too. As I said some brands have fitting and radiators that aren't what they seem.

I started water-cooling in 2009. Used distilled water all this time. Only EK GPU blocks have given me problems. Swap them out and suddenly the corrosion goes away. How strange. I still use EK CPU blocks and fittings. Just those 10 and 20 series blocks were shit.
 
#1 The preview picture may reveal the hole topic.

After not bothering reading IGORSlab and a few others with their All in one = AIO = fake water coolers.

Most often you have mixed materials which need a special fluid to slow down the chemical reactions. I think in a car you also need special fluids for the cooling and the brakes.

Those aio use some water + additives like glycol or what it is called. Therefore I really dislike calling them water coolers. Fill in tap water and than we talk again after a while.
 
Voted: ofc not, are you kidding me!

Been water cooling for over 20 years now (since socket 478) and found out very early on that distilled water will destroy your loops.

Tried car coolant (~20 years ago) and haven't looked back. This was back when not a drop of liquid coolant was available on the market for PCs. I did get ridiculed at first for using it too!

I've mixed and matched metals all that I've liked and get great cooling to boot. The coolant I use is just simple automotive coolant that uses 50/50 distilled/glycol.

I've never seen parts look soooo damn clean too even after loops have been running for 3+ years. They look brand new after use. No cleaning, nothing.

Edit: Pro tip, it glows under UV lights too!

IMG_1685.JPG IMG_9197.JPG IMG_1598.jpg
 
Last edited:
Who watch the whole video when he said nickel and distilled water isnt a problem... It's only when zinc is added to the mix that you can have corrosion.

The ratio of brass to nickel is something to consider too. As I said some brands have fitting and radiators that aren't what they seem.

I started water-cooling in 2009. Used distilled water all this time. Only EK GPU blocks have given me problems. Swap them out and suddenly the corrosion goes away. How strange. I still use EK CPU blocks and fittings. Just those 10 and 20 series blocks were shit.
Any dissimilar metals in contact with water will suffer from galvanic corrosion. That’s why commercial automotive coolant has anti-corrosion additives.

Fun experiment: Fill a bucket with water. Put a strip of aluminum foil in one side of the bucket. Put a strip of zinc in the other side of the bucket. Attach a volt meter to the strips and you’ll measure about 0.5 volts. You’ve made a battery. What happens to batteries after a few years? They corrode.

The fouling layer is easiest to understand. Basically, as fluids flow over surfaces they slow down the closer you get to them (and the faster you go). Your ceiling fan needs cleaned, despite spinning around, because the edge of the blade has a fouling layer while cutting through the air, and when the air slows down so much it dumps dust. Same thing in a liquid loop
This is the boundary layer caused by laminar flow. Which is why quality cooling systems don’t use smooth surfaces. You want turbulent flow for better performance.
 
Last edited:
Funny how after 10 years with proper coolant, I have all my plating on my EK 1080 block.

It is real simple. Don't etch it with vinegar.
I try to keep in mind the 3 golden rules 1) no mixing aluminum and copper 2) no vinegar to clean nickel 3) no isopropyl alcohol to clean plexi/pom

Don't flush it first day for no reason.
What do you mean by this? When I get new radiators the first thing I do is use distilled or simply pre-boiled water to first get any large debris and particles out of the radiator and vinegar to provide for some sterilization. Then flush them finally with distilled and dry them out before using or storing them.

Don't use distilled water. Assemble it, use proper coolant, change the coolant every 1 or 2 years per manufacturer's recommendations, and never have 'bad plating' ever again.
I am a bit surprised at the poll results. I do wonder though if PC "proper coolants" are a bit more concentrated than they need to be.
 
I will always call out companies for having bad platting. EK GTX 10

The picture I posted above have 2x gtx 1070 from EK water blocks. They were running for over 4 years no issues using car coolant ;)

I still have those blocks and they still look pretty good. Here are some pictures of one that a fellow TPU member wanted me to take apart and look inside after such a long time.

Original: take note I did not touch the fins. It's just the crystals from the glycol that's dried up.

IMG_8760.JPG

Slight wipe with old t-shirt:

IMG_8763.JPG

Oh 20+year old copper block ;) This thing went through the ages - last part of 478 then I ghetto modded it for 775. See the cut out the bottom right corner. Worked great too! Car coolant the whole time.

IMG_8819.JPG
 
Looks like I'm the only one that isn't having distilled water problems. Lucky me I guess. Next flush I'll waste money and fill it with that overpriced bottled stuff. Since it's supposed to be better, I'll never have to flush again.

I don't decredit der8auer when he says not to use distilled water. He has scientific equipment saying it's bad. I'm personally just not seeing this happening. Only thing happens is slight discoloration. No gunk in the loop or platting coming off... Unless it's a bad block.
 
Last edited:
I'd go weaker as anti-freeze has a low specific heat.
Specific heat isn't really the relevant property for a cooling loop not operating near or above the boiling point; you're more interested in thermal conductivity. Ethylene glycol loses to water in either case -- the only reason it's a component of antifreeze is revealed in the product name itself: it keeps the coolant from freezing at low temperatures.

If you're concerned about corrosion and performance, add a tiny fraction of commercial corrosion inhibitor to the distilled water.
 
I ran some weak mixture of distilled water with glycol car coolant (15%?) and had a problem with some green stuff growing in the tubes and the tank of an Alphacool Eisbaer Extreme Liquid CPU Cooler 280, I connected Core 1 block to it with my tubing.

I think it might have something to do with a strong blue light they put inside to light the tank.

Anyway, as a precaution I will run an least 25% strong mixtures from now on.

I ran bleach water in it to hopefully kill anything organic in it, but is seems that there is still some residue left inside. :( I also disconnected the light.
 
The Propylene acts as an inhibitant and lube for the pump motors (same as what you get from Automotive coolants, they taste the same as well.. :laugh: accidental sip)
Neither propylene nor ethelyne glycol acts as an inhibitor in pure form; they have inhibitors added for cooling applications. I also find the claim of "lubrication" a bit dicey, as I'd hope any well-designed centrifugal pump has nothing but the impeller fins in contact with the working fluid. There may be other parts or factors I'm not considering, however.
 
Last edited:
Neither propylene nor ethelyne glycol acts as an inhibitor in pure form; they have inhibitors added for cooling applications. I also find the claim of "lubrication" a bit dicey, as I'd hope any well-designed centrifugal pump has nothing but the impeller fins in contact with the working fluid. There may be other parts or factors I'm not considering, however.
I find a high dilution of propylene glycol slightly increases the flow as to initially just pure distilled, I usually top up with distilled 1st then after bleeding air out I add the propylene glycol and I have stats where I can see water temp and flow rate and its slightly noticeable (maybe placebo or within margin of what the effin err)
 
Specific heat isn't really the relevant property for a cooling loop not operating near or above the boiling point; you're more interested in thermal conductivity. Ethylene glycol loses to water in either case -- the only reason it's a component of antifreeze is revealed in the product name itself: it keeps the coolant from freezing at low temperatures.

If you're concerned about corrosion and performance, add a tiny fraction of commercial corrosion inhibitor to the distilled water.
Do you have an example of the right kind of commercial corrosion inhibitor to use?
 
Not gonna lie, this video instantly made me want to purchase the 5L DP Ultra he had by his desk, but I'm already using the best pre-mix available on the market in AU :laugh:.

The owner of the store below does a pre-mix and sells it, has been working well for over 2 months, crystal clear.

"Contains: propanediol (anti freeze and growth inhibitor), pH balancer, benzotriazole (anti corrosive), benzalkonium chloride (anti microbial) and distilled water."
 
I ran my loop for 5 years with distilled water and no problem, didn't even change water, i was surprised watching Der8auer video too.
 
I ran bleach water in it to hopefully kill anything organic in it, but is seems that there is still some residue left inside. :( I also disconnected the light.
So this brings up another issue I've often pondered. Regardless if it's prepping a loop for the first time or trying to correct an issue of biological contamination, when needing to sterilize your loop what household agents are generally "safe" to use?

Luckily I've only had one issue with biological contamination since I started WC around 2016 and it was because of dumb mistake of using a non-sterilized container for a quick drain and refill on my NR200P build (RIP) and got a white bloom of growth for a short time before it seemed to stop that occurred in a relatively fresh batch of EK Cryofuel.
 
Funny how after 10 years with proper coolant, I have all my plating on my EK 1080 block.

It is real simple. Don't etch it with vinegar. Don't flush it first day for no reason. Don't use distilled water. Assemble it, use proper coolant, change the coolant every 1 or 2 years per manufacturer's recommendations, and never have 'bad plating' ever again.


See how 50% of poll respondents are ignorant?


Let's play devil's advocate and say some blocks are better than others because they have thicker plating or more complete plating. Fine. I'm sure it's true. But does that mean your not ignorant for ignoring science? No. If you ignore science, you will pay the toll. How much you will pay the toll depends on all the other factors.

You...need to check your anger and ignorance.

Let me ask you, how many different ways can you plate? I'll give you a second.
.
...
.
Let's cover steel. Steel+zinc = galvanized steel. Hot dipped, cold rolled, hot rolled, and electroplating are entirely different ways that create mainstream options. There absolutely are better and worse plating versions...so let me give you the insider knowledge that you don't have.

Electroplating is great. If you clean the surface badly, leave behind carbon deposits, and then electroplate it gives not one single f***. It'll encapsulate that carbon on the surface and lay down a bunch of zinc. The problem is that if you then try to form that steel you can potentially get wiping. Yes, instead of shearing or bending the zinc wipes over the surface of the carbon. It can form rifts, pocks, or if you are shearing and stretching it'll form fine wisps. Hint, if you then allow these wisps to accumulate in say a filter, which is one use of expanded metal, suddenly zinc strands get sucked down into your engine. That can't be good for it.

Hot dipped is pretty awesome. If it's hot enough the bond is slightly chemical and mechanical. IE, iron and zinc atoms in the outside layers do a bit of swapping, and you get an excellent bond. Very energy intensive, not great on the thickness consistency without rolling, and dangerous to be around. That said, it's bonded unless it was too cold...and if it fails it'll usually fail immediately.

Cold rolled is a mediocre bond, usually has some sort of cleaning and bonding flux, and prestresses the material. Great for strength, bad for formability, and it tends to be cheap. It will flake, it will disbond without critical failure, but it's cheap.

Hot rolled is cold rolled, with less grain structure changes. Not hot enough to be liquid, but the flux generally bakes off and this will provide you a dead on overall thickness. A lot of the material on the market is this, because it's generally very formable and cheaper, if not the cheapest. This is what you buy if formability matters, price is an object, but you want something at the end which will likely be meant to last.

What I haven't touched on is galvannealed (largely named for the annealing to get that grain structure unstressed), galvalum, or vacuum deposition (when cost is no object, and precision has to be king).



With all of the above said, you also have to consider the process. Let's say you're English...and an idiot. You might have worked for British Leyland, where because they hooked up the electricity backwards their galvanization process (electroplating) actually stripped the iron in their steel and formed a layer of rust that they painted over...causing the cars to rust. You could also set the voltage of your electroplating process to the moon. That'll make it plate faster, right? It couldn't possibly form new oxides inside the zinc structure, turning it from a coating of protection into a brittle surface skin that looked like swiss cheese. Most interestingly, you could plate on too low of a voltage. Yes there's some plating going on, but a zinc layer thin enough to be scraped away during regular friction events (IE usage) might as well not be there.
In short, a quick bath of vinegar, or acetic acid at insanely low concentrations, is about a damaging as the carbonic acid that the dissolution of carbon dioxide into water forms in atmospheric concentrations. Saying that would somehow matter is like claiming that a common frog on a log would have stopped the Titanic dead in the water...never mind the fact it was designed at least partially to endure impacts by icebergs. An iceberg being an order of magnitude more damaging than a log or a frog. This is why anyone with any understanding looks at the stuff that is on display with the OP's video, and laughs. He's claiming that distilled water is a problem...then showing us something that is caused not by the water, so much as what the water being an aqueous solution allows. Blame the gas company, they allowed me to fill my car with gasoline, then drive it whilst intoxicated into a building. Either that, or blame the distillery, because if I wasn't drunk I wouldn't have crashed. Never mind the actual cause of this...because they'll gladly sell you on a lawsuit to blame the refinery or distillery.
 
You...need to check your anger and ignorance.

Let me ask you, how many different ways can you plate? I'll give you a second.
.
...
.
Let's cover steel. Steel+zinc = galvanized steel. Hot dipped, cold rolled, hot rolled, and electroplating are entirely different ways that create mainstream options. There absolutely are better and worse plating versions...so let me give you the insider knowledge that you don't have.

Electroplating is great. If you clean the surface badly, leave behind carbon deposits, and then electroplate it gives not one single f***. It'll encapsulate that carbon on the surface and lay down a bunch of zinc. The problem is that if you then try to form that steel you can potentially get wiping. Yes, instead of shearing or bending the zinc wipes over the surface of the carbon. It can form rifts, pocks, or if you are shearing and stretching it'll form fine wisps. Hint, if you then allow these wisps to accumulate in say a filter, which is one use of expanded metal, suddenly zinc strands get sucked down into your engine. That can't be good for it.

Hot dipped is pretty awesome. If it's hot enough the bond is slightly chemical and mechanical. IE, iron and zinc atoms in the outside layers do a bit of swapping, and you get an excellent bond. Very energy intensive, not great on the thickness consistency without rolling, and dangerous to be around. That said, it's bonded unless it was too cold...and if it fails it'll usually fail immediately.

Cold rolled is a mediocre bond, usually has some sort of cleaning and bonding flux, and prestresses the material. Great for strength, bad for formability, and it tends to be cheap. It will flake, it will disbond without critical failure, but it's cheap.

Hot rolled is cold rolled, with less grain structure changes. Not hot enough to be liquid, but the flux generally bakes off and this will provide you a dead on overall thickness. A lot of the material on the market is this, because it's generally very formable and cheaper, if not the cheapest. This is what you buy if formability matters, price is an object, but you want something at the end which will likely be meant to last.

What I haven't touched on is galvannealed (largely named for the annealing to get that grain structure unstressed), galvalum, or vacuum deposition (when cost is no object, and precision has to be king).



With all of the above said, you also have to consider the process. Let's say you're English...and an idiot. You might have worked for British Leyland, where because they hooked up the electricity backwards their galvanization process (electroplating) actually stripped the iron in their steel and formed a layer of rust that they painted over...causing the cars to rust. You could also set the voltage of your electroplating process to the moon. That'll make it plate faster, right? It couldn't possibly form new oxides inside the zinc structure, turning it from a coating of protection into a brittle surface skin that looked like swiss cheese. Most interestingly, you could plate on too low of a voltage. Yes there's some plating going on, but a zinc layer thin enough to be scraped away during regular friction events (IE usage) might as well not be there.
In short, a quick bath of vinegar, or acetic acid at insanely low concentrations, is about a damaging as the carbonic acid that the dissolution of carbon dioxide into water forms in atmospheric concentrations. Saying that would somehow matter is like claiming that a common frog on a log would have stopped the Titanic dead in the water...never mind the fact it was designed at least partially to endure impacts by icebergs. An iceberg being an order of magnitude more damaging than a log or a frog. This is why anyone with any understanding looks at the stuff that is on display with the OP's video, and laughs. He's claiming that distilled water is a problem...then showing us something that is caused not by the water, so much as what the water being an aqueous solution allows. Blame the gas company, they allowed me to fill my car with gasoline, then drive it whilst intoxicated into a building. Either that, or blame the distillery, because if I wasn't drunk I wouldn't have crashed. Never mind the actual cause of this...because they'll gladly sell you on a lawsuit to blame the refinery or distillery.
What point are you trying to make?

When it comes to corrosion, plain distilled water is worse than coolant which has corrosion inhibitors. Every time. Well known. Proven. Anyone who wants to blame anything else is simply ignoring the fact that they are making a sub-optimal choice out of ignorance; especially if they choose to do what they do because they are dilusional in thinking what they are doing is better and some evil company is to blame for their results. It's not just ignorant, it is typically a vocal defensively combative response. My reply is measured.
 
Koolance clear 702. I've been using it in a system (not a PC) at work that uses a D5 pushing through a MO-RA3 420 radiator (copper tubes) with a series of large aluminum blocks for years (4 years?...ish) without changing the fluid. I opened a block the other day because I was sure there'd be some corrosion due to mixing the metals. Nope. Perfectly clean. After some of the industrial systems I've ran at work with Koolance 702 clear and having some gunking issues with Cryofuel clear at home, I did a mayhems cleaning on my system at home, then switched to the 702. It's now been a couple years and all is still perfect. I have a full copper CPU block, combination of EK and Koolance fittings (their QD3 quick disconnects are awesome) a MO-RA3 420, an EK radiator, and a nickel-plated copper GPU block and it's all happy. Get some proper fluid (clear) and don't worry about it.
 
What point are you trying to make?

When it comes to corrosion, plain distilled water is worse than coolant which has corrosion inhibitors. Every time. Well known. Proven. Anyone who wants to blame anything else is simply ignoring the fact that they are making a sub-optimal choice out of ignorance; especially if they choose to do what they do because they are dilusional in thinking what they are doing is better and some evil company is to blame for their results. It's not just ignorant, it is typically a vocal defensively combative response. My reply is measured.

You say that as though there is literally any truth to it...and somehow it's an absolute. The rate at which a pure metal and plastic system oxidizes is functionally zero...because literally any oxidation that is visible is water insoluable. This is why aluminum is insanely reactive...and it's not considered as dangerous as something like sodium. The oxide layer that forms prevents any interaction, cannot be dissolved, and thus is not a problem visible to the naked eye...as anecdotally confirmed here. As a hint, this is not facts...because oxidation also 100% occurs in systems with the additives discussed...because water is water and adding an alcohol to it does virtually nothing to stop chemical processes.

What we see instead is fouling that far exceeds what we should see and would be attributable to water. Read: why I stated that he's selling us the solution he's also selling us that we have. You are welcome to be ignorant, claim that water is a problem...and then use water with a pinch of fairy dust. It doesn't solve to core issue of water creating the demonstrated fouling...which at its core is not a 1:1 "because you used distilled water" as is trying to be sold to us. Heck, it isn't even a 1:1 that his solution (in the chemical sense) will solve the problem...just that he's sure that anyone could demonstrate this failure and thus you take responsibility for any fouling if you don't use something "better than" distilled water. Shameless plug of his product page, and chemistry that isn't exactly on the level.


Oh my god...did I just accuse him of lying? No. Chemistry not on the level is like using the galvanic series of electronegativity to explain things...which many people may see and absolutely thing is proven science. Cool, cool. The thing is that the corrosion that is demonstrated is not necessarily electronegativity. What it is, is reactivity. In an aqueous solutions the electronegativity matters...assuming you've only got galvanic corrosion. The simple answer is battery = dissimilar metals+conductive aqueous environment makes energy and reduction reaction....wooo. Of course, the dissimilar metals are required...so it's a no-go with infinite supplies of the same metal and the minor amount of salt that will be made with the metal combining with ions in the water.

...let me say that one more time. HE HAS USED A GALVANIC SERIES ASSUMING THAT DISSIMILAR METALS WILL BE USED...
I say that in caps because, quite simply, and anecdotally defined, distilled water in systems is not the problem. You setting up a battery is the problem. Isn't it amazing that he'll sell you the solution to having a battery as additives meant to retard, but not prevent, oxidation. Again, RETARD. To slow down. Not prevent. I get to say that with a smile on my face, because the solution being sold isn't even a solution. It's a stop-gap. Amazing work. Slow clap.



Let me put this as a parallel. The US spent one million to develop a space pen, and the soviets used pencils (Space Pen article from Scientific American). The solution to have a weak acid not form a galvanic cell is to not use dissimilar metals...not to spend huge amounts on exotic coolants. I cannot even believe that you want to argue that distilled water is the problem...because it's so face palmingly stupid that I cannot adequately express how infuriating that your callous statement to another is...and I'd like nothing better than to ask why exactly you have the ego to say this. Then again, I also know that the people who are absolutely sure of themselves are either sure to fail or sure to show their clown shoes eventually...so thanks for letting me get frustrated. It's...good to remember how trusting people sucks, and trusting der8aur should be taken with that grain of salt.
 
Do you have an example of the right kind of commercial corrosion inhibitor to use?
Not really my field, but the best inhibitor will depend on which metal(s) you have in the loop. I believe there are companies selling additives specifically for PC liquid cooling, which I assume contain a generic mix that performs well for copper, aluminum, and common ferrous alloys.

I'll also add that from an electrochemical perspectives, the arguments made by @lilhasselhoffer are relevant. Galvanic corrosion (from dissimilar metals) isn't the only mode of corrosion possible, but it is by far the fastest. If you're using the same metal throughout, a drop or two of the proper passivator for that metal in distilled water should last decades without appreciable corrosion.
 
Back
Top