Please HELP

[ibmkg]

I noticed accumulation of some slimy stuff on walls of reservoir. It is brown-opaque gooey stuff. I did add Iodine and I did use distilled water with coolant.

What is this? Algae or Fungi? How to coup this?
Is it safe for skin contact?

How to wash this sh?

[ibmkg]

My Al block got corroded! Will post pics soon. Formed pits.
Made a Cu block that will replace it.

[=ACID RAIN=]

Yeah you need to clean out your radiator as well. Anything in that loop needs to be cleaned if you used aluminum in the loop (even anodized, the anodizing never lasts).

[madmaxx]

might be a good idea to flush the rad with some diluted CLR?

its never a good idea to mix metals, if you do : frequent fluid changes are a must

[Brians256]

It is most likely algae. Safe to touch? Probably. I would still dump the parts (not including the tubing) into a big bucket of hot water and ammonia outside where it will ventilate and not kill you with the fumes. Then, I would throw out the tubing and get fresh tubing. Don't let the metal parts sit too long in the ammonia. It'll eat it up. Scrub the parts.

If your aluminum is pitted, it is either galvanic corrosion (most likely) or acid from the biological growths. My cure for both is to use 25% to 50% antifreeze. There are other solutions, true. But it sounds like you weren't the kind of person to be changing your distilled water every 2-3 weeks, and if it was galvanic corrosion, antifreeze is one of the cheapest ways to prevent the problem.

Iodine evaporates and precipitates. It will not prevent galvanic corrosion even before it evaporates. When I mean evaporate, I mean that it diffuses through the tubing and evaporates into the air. Tygon and other clear tubing is usually not 100% water proof. It is porous like a balloon and will leak like it too (very slowly).

[ibmkg]

Picture 3 and 4 are after I scrubbed and washed it with steel wool and water.


I can make antifreeze to be 100%. Thus no need for distilled water. I am also going to replace this Al block with Cu one. Will post pics soon.

But the question is, what to add so that I never have to worry about algae or fungi forever?


What is CLR?

How do you propose to clean out rad?

Will try ammonia and see what it does.

[madmaxx]

CLR = Calcium Lime Rust remover

i wouldnt go 100% antifreeze but more frequent fluid changes for shure

[Brians256]

I wouldn't make it 100% antifreeze. That really doesn't buy you much more protection than 25-50% antifreeze, and it lowers the heat carrying capacity.

I had a system running for 3 years (maybe it was only two years, but I just didn't pay much attention) with no fluid changing. I think that the four big factors were:
- 25-50% antifreeze
- Copper and silver blocks (no aluminum)
- No light allowed to touch the fluid (no UV lights, no case windows, etc...)
- Huge (one liter) reservoir that was not completely airtight

The antifreeze that goes into cars is good at regulating the pH. It's made that way. At 17C-30C temperatures, it is going to last a lot longer than the harsh environment it sees in a car (higher temperatures accelerate chemical reactions). So, a mixture that would last a year in a car should do well over that in our PCs.

My waterblocks were not aluminum, which is highly reactive. They were copper and silver, meaning that galvanic corrosion was reduced with respect to the brass heater core. Also, I'm thinking that the silver waterblock (thank you DangerDen!) helped prevent biological growths. I don't think I thought of that before. Small amounts of silver kills bacteria, so it probably helped kill any incipient algae colony.

No light! I'm convinced that light is one of the major reasons why blingers get algae. Algae grows really fast when it gets light. If you must have light, then flush the system often.

The reservoir helped so much because most people don't realise that vinyl tubing leaks. Actually, water migrates slowly through it and evaporates on the surface. You can get 100% leak-proof tubing, but it's not what you get at Home Depot unless it's the nylon reinforced tubing. So, over time, the fluid escapes. If the reservoir is airtight, then the tubing collapses! Even if you allow pressure equalization, you need to have enough reserves to last this out over time. I figure that I lost about a cup over the system's lifetime.

[ibmkg]

Cu WB is ready. However, have some issues now:

1) Its too heavy. Weighs more than a kilogram. Will it be too much of stress on Mobo?

2) This Cu WB was nice and shiny, but after few hours, it turned brown. Eventually (today) it is somewhat dark brown. What to do about that?


Brains256:

I cannot cut light. I have a LED in reservoir and tubing is somewhat clear, nylon reinforced. Actually, I have homebrewed this rig and without little bling it looks like junk.

You said that Silver kills some germs. I can put crushed silver foil in the loop. I have heard that silver foils are used to decorate edible food in hotels. What do you think?

[Brians256]

A kilogram is very heavy. You might want to work on your block. It sounds like you can't reorient the system so that the motherboard is flat with the waterblock lying on top of it.

Copper oxidizes and turns brown. There isn't much you can do about it. I've seen some people polish it and put wax on it to slow the oxidization though.

You could put a chunk of silver in the reservoir. Don't put anything that would break down into grains and hurt your pump. I really don't know how much this would help or hurt, to be honest.

[ibmkg]

Ok now I am using a Cu block. Noticed one thing, Cu isn't performing as good as Al. May be the design works well with Al than Cu?


VGA card died somehow. Sometimes it works but usually no display. I bought a new one, 8600GT. Was wondering if I need to water cool it as the heat pipe and heat sink on it looks amazing. It's fanless, may be I just add a fan and it will work better than a Water Cooled?

[Brians256]

Cu should always work better than aluminum. So, the things that I would check are mounting and structural differences.

A heavier block could easily torque off of the CPU just a bit more. Mounting pressure can alter the CPU temp several degrees without being visibly noticiable. Finally, the block could just be settling in because of the thermal compound settling in.

Structurally, the new block could be less flat. Or, the new block could be marginally thicker at the base or the jets could be just a hair different than the older block. Check the flatness and contact patch. You've probably done this but newbies reading the thread should take note.

My bet is on the first set of issues. However, I've assumed that you have kept all other variables unchanged. Is the coolant fluid the same? Is the room temperature the same? Etc. etc..

[ibmkg]

The blocks are fairly identical. Unless, if small unnoticed changes can affect. Nevertheless, I really could not find any difference in mm scale.

For both of the blocks, I used a sand paper-glass method to flatten the blocks (the contact area).

It can be due to the break-in period of AS5 which I did not account for. At most, I am expecting 3degree change with that. I am expecting similar performance of Al and Cu at best.


It is not due to mounting pressure, I am sure about that. I can try over tightening it.

Currently, it has just distilled water. No additives. I just want to cleanse the new tubing and reservoir. I will drain it and add coolant with anti microbial.


Ohh and do advice me for 8600GT 512MB SLI card. Should I just add a fan or go water? I will be overclocking it. Keep in mind that I will homebrew the water block.

LINK:

http://global.msi.com.tw/index.php?f...30&cat2_no=136

[Brians256]

Good luck on the CPU waterblock. Tightening the mousing down more might help but remember that pure copper is softer than aluminum so you could be getting some bowing. Hmmm.. what copper alloy are you using? Some alloys aren't any better than aluminum at heat transfer, although they are much easier to mill than pure copper (too soft and the bits just get sucked into the material).

As for the video card, water should be better than air. However, I really don't know about overclocking them and how sensitive they are to temperature. Personally, I like the Accelero S1 with a 5V Nexus fan. But that is not overclocked. It might be that the S1 with a higher CFM fan could be better overall because it would cool the whole board more evenly. The S1's pack a lot of surface area to cool the GPU, and it's cheap. I've seen a lot of good press with regards to the S1 and how well it cools.

On the other hand, you might not have enough room for a pair of 8600GT's with S1's strapped onto them. Water cooling is great at moving the heat to a big radiator where you can more easily get that heat into the air. You will degrade your CPU temps a bit, though. It's probably a good tradeoff. I did that with my last watercooled system and liked it (overclocked athlon 2500 and an ati 9800).

[ibmkg]

Brians256, Thanks.


VGA:
I do not have a design in mind yet for VGA water block. I am thinking of something that is universal. It should be able to mount over most of the VGA cards.

I installed a 40mm small fan over it in the middle of HS. I see 48C idle. How do you people check VGA temperature with load?

If I am to upgrade the cooling, I will go water due to the reasons you stated.

Also, will changing the themal paste help? I do not know what stock MSI thermal compound is but I have Arctic Silver 5 (AS5) at hand.


CPU:
CPU temperatures have dropped about 5C. It is performing same as Al now. This is just the idle observation. I slightly tightened the block but could not observe any temporary difference. I cannot tighten it any further due to risk of damaging the mobo. Also, my mounting is such that if I tighten it more, the mobo will touch the back of casing.


I am using pure Cu. At least that is what the dealer said. It was forged to size. Working with Cu was indeed tougher than Al due to the softness. I can post the pic of the raw material if somehow you can identify the pureness

.

[Brians256]

I don't test GPUs but I remember a number of tools to do so. I don't know if SpeedFan can do it but there is a tool for nVidia that does it. Changing the paste probably won't help much. You need a better heatsink or just go to water. Really, though, the S1 is a very good heatsink and far cheaper than going water.

As for the copper, I have no way to determine the copper alloy. Glad to hear that the compound has had a chance to settle in and provice better temps. 5C is a big jump.

[billbartuska]

Originally Posted by Brians256

but I remember a number of tools to do so.

Overclock, Riva tuner. Temp monitor, NVTempLogger.

[ibmkg]

How often should I change fluid in my setup?

[Brians256]

The correct and short answer is: As Needed.

Unfortunately, you'll have to adjust the schedule for your particular system due to temperature, ambient biologicals and their growth speed, type and amount of various metals, composition of your coolant and other miscellaneous factors.

If it was my system, I'd take a look at 30 days and put it back together if there was no perceived change to the fluid and surfaces. You can go all technical and see if the buffers are exhausted but I don't know how to easily test it. If all was well, I'd check again in another 30 days, and so on.

Theoretically, you should be able to go 6 months to a year if you have enough coolant and thus antifreeze to buffer against corrosion. I'd still check on it since your last system showed such obvious signs of degradation.

[ibmkg]

This time I have 35% Coolant, 15 drops of iodine and 10 drops of Gentian Violent to counter any fungi. Rest is distilled water.

I am going to turn on the bling as well.

[[AfZ]PiMp J]

admittedly i didn't read the whole collection of posts, but if you want to compare computers to cars with respect to watercooling (and why not, they're about the same) there are some key points to be made:

1) aluminum will cool better than copper if you have faster heat transfer from the pipes to the radiator (more fins, more passes in the radiator, more air flowing over the radiator)

2) mixing of metals should not matter. water in a car (in this next example a high performance car) can pass through an aluminum radiator, an iron block, and aluminum heads (for the sake of this example, even a copper radiator), without fungi growing. the most likely cause of your growth is the light.

3) water wetter. it's proven to work in cars to reduce temperatures, helping with heat transfer even more than a 100 percent water "mix"

[billbartuska]

Originally Posted by [AfZ]PiMp J

1) aluminum will cool better than copper if you have faster heat transfer from the pipes to the radiator (more fins, more passes in the radiator, more air flowing over the radiator)

2) mixing of metals should not matter. water in a car (in this next example a high performance car) can pass through an aluminum radiator, an iron block, and aluminum heads (for the sake of this example, even a copper radiator), without fungi growing. the most likely cause of your growth is the light.

3) water wetter. it's proven to work in cars to reduce temperatures, helping with heat transfer even more than a 100 percent water "mix"

1.) Well, so would cement. if......

2.)Antifreeze contains antibacterial agents, but the primary problem with "mixed metals" is corrosion, and antifreeze contains anticorrosives.

3.)Please quote some computer watercooling testing with water wetter.

[Brians256]

1) aluminum beats copper for two reasons: cost and weight. If you can put 2x more fins for a given weight, you get better cooling. But, per unit volume, pure copper beats pure aluminum. There's been a bajillion real-world tests (identical design and even CNC) to prove it.

2) Cars use a lot of antifreeze and require a flush every 12 months to maintain health. Plus, even with this regime, certain designs had difficulty. This is why antifreeze has a lot of buffering agents and (if I recall correctly) a silicate chemical which coated the metal(s) to deposited into a film to prevent galvanic corrosion.

3) Water wetter seemed to help, but it did leave a film on tubing. Do a search on the forums. There was plenty of discussion on water wetter. There may have been other drawbacks but my memory fails.

[[AfZ]PiMp J]

Originally Posted by billbartuska

1.) Well, so would cement. if......

2.)Antifreeze contains antibacterial agents, but the primary problem with "mixed metals" is corrosion, and antifreeze contains anticorrosives.

3.)Please quote some computer watercooling testing with water wetter.

1) dont be a smartass

2) antifreeze may contain antibacterial agents, but this is solely a biproduct of it's chemicals, it is not on purpose. his problem was said as having a fungi type growth. his solution had antifreeze in it. my point is that WATER ALONE w/O help from any antibacterial agents in antifreeze would not cause a growth when mixing metals, therefore mixing metals was not his problem, light was. i purposefully left out corrosion, as i was not attacking that issue, only growth. i otherwise agree about the anticorrosives.

@ Brian
3) water wetter contains a lubricant that additionally helps wear and tear on water pumps, most likely the leftover film on the tubing had to do with this lubricant