Build Log: 100,000 PPD by 1/23/2011

[artemm]

I wouldn't rule out bacteria. PT Nuke won't kill everything. There's even bacteria that lives in diesel. Do you have any silver in the loop? Is the res open or closed now? Pics would help.

Depending on what blocks you're using, you may have to take it apart as the fins can be very fine and will have lots of deposits that'll need to be physically brushed out. Boiling and vinegar alone won't do it.

[EtaCarinae]

trn I am really sorry bout your misfortune. Your system was really a major contender and I was happy to be challenged daily. I feel a little alone now.
Your water cooled system was absolutely unique at least to me. It was very courageous to build such a large and complex watercooled system. There is no other way around, you have to go through a learning curve. But through trial and error you will be back stronger and more reliable.
Keep up high your spirit and motivation. We are all with you. And now go, go, go, go

[trn]

Well the good news is I found my problem.


This radiator cap off and caused the water to all leak from the system. One of my radiators didn't have a real radiator cap in it and I never realised it and its amazing it made it this long without blowing off. The cap pictured is just a cheesy little plastic insert that temporarily plugged one of the rads. The good news is that by going to autozone and buying a real radiator cap without I can fix the system.

The bad news is I was running without water, which was causing overheating and who knows if I did any damage to my CPU's. Also I still have the brown gunk issues as you can see on the inside of the radiator cap; i'm still sick and not sure if I have the energy to work on a repair this weekend. And also I still can't explain or fix the brown gunk

[OldChap]

There are proprietary flushing solutions you can buy from an auto parts dealer but I would still try to find out if you can get this done professionally for not too many $$$'s

[trn]

And a bit more good news... I started feeling better so I made my way over to Autozone, picked up a real radiator cap, slapped it on the rad, refilled the system and fired everything up and the CPU's and pumps all still work so i'm thankful that no damage was done. I'll figure out the brown gunk later

[shoota]

well that's mostly good news trn, which i'm glad to hear. especially that you are feeling better. That brown gunk is perplexing, and i'm with the other guys in that maybe a professional cleaning is in order if the gunk gets out of hand

[trn]

Yeah yeah... i'm a lazy slob that doesn't like cleaning I also struggled to find a shop that would flush radiators that were not in cars! I had looked for this service before I ever started my build but had no luck.

For now my sludge problem doesn't seem to have any affect on temperatures (as long as I have water in my system .) My highest core temps right now are in the 50C's on my 970's at 4.2GHz which is more than good enough. The only way to really diagnose the problem is a complete system tear down which i'm also not in the mood for quite yet.

And after more inspection I have a suspect... I think I have a Rusty Taco!

Rusty Taco Photos:




Why a "Hot Water Circulator Pump" would rust this bad when circulating water is beyond me... I had assumed the green paint was a thick inpenetrable epoxy paint that would keep the cast iron parts from rusting up this bad. This just looks like awful design and engineering from Taco pumps. I have a feeling that If I tore down the pump the inside would be very rusty. I don't know for sure, I've certainly assumed many things incorrectly through this build log...

Heres what the tubing looks like today.... thick thick apple cider, yum The light color tube is dry, I pulled that tube from the system a month or so ago.


And heres a shot to try and capture what the residue looks like when dry, its faily bright orange, kinda rusty looking

[D_A]

That looks like rust to me. If it is electrolytic corrosion then it will start pitting the ferrous metals in the system pretty quick. You could try grounding the copper water blocks to a common point with the cast iron Taco pump, but you probably should still run some good radiator cleaner through that and put some serious corrosion inhibitor in the mix.

[artemm]

Very similar to an E28 Bimmer my buddy was working on today. It had a bad water pump that was neglected and the coolant that was flushed out looked like yours. Plus all of the internals were coated with brown gunk that looks exactly like your pics.

If your pump is OK, you may have used coolant that's aggravating some of the metals in your loop. If you didn't use distilled water, mixed antifreeze, or didn't use antifreeze at all, it may have caused your corrosion. Antifreeze has rust inhibitors...

Still.. your poor copper blocks Sorry, man.

[shoota]

Looks like you found the culprit. Now the question remains, how much energy do you have in order to deal with it? Fix it now or let it go and see how long the loop performs well enough? These are the tough questions in life

[trn]

Thanks for the confirmations and info; now it sounds like I need to add some antifreeze to my system to slow the rust and keep my loop from freezing on these cold cold south florida nights! I hope I can even buy aintifreeze here in S.FL

So far my temps are still unaffected by the brown residue enough to make any difference. I'm guessing because the thermal conductivity of Iron-Oxide is still good enough and its only a thin layer that is still mixed with water. I'm in no mood to try and "fix" the system right now, but putting a bandaid on the problem would be good enough for me. I'm also going to rework two of the sub-loops soon to add a 5th cruncher to those loops so each subloop will have 5 crunchers. At that time i'll replace all the tubing in those two loops and give them a fresh start

*don't fix what isn't broken:

[D_A]

Any auto parts shop will have coolant. It's the same stuff for both anti-freeze and anti-boil and they all have corrosion inhibitors. You can often get just the inhibitors without the glycol (or whatever organic substitute they use in the red stuff)

[OldChap]

Now you only have to make one decision....

Coolant What color? Blue is typical Red is available ,,,from Toyota I think,,,

[D_A]

Here green is the common one (ethylene glycol based) while a lot of the newer models use an organic based coolant (red one, but Toyota make their own) but I've never seen blue automotive coolant. Plenty of blue brake fluid, but not coolant.

[Jaco]

VW G11 coolant is blue.

[Pointhore]

You can also buy uncolored antifreeze.

[shoota]

considering his tubing is a poop brown color i don't think the color of the coolant will matter too much lol

[EtaCarinae]

trn I am very pleased you did not have too much damage to your system. One way to avoid completely the rust or coorosion issue would be to use a very low viscosity oil. Very thin synthetic oil. I am sure this exists. And oil has an excellent heat transfer capability. Your pumps may have the capability to pump such a thin oil. I think it is worth studying the issue. But you must be careful that this type of oil would be compatible with your tubing material.

[D_A]

That's something I've considered trying in a car for ages, except that oil transfers heat far slower than water (but still better than air). I still like the idea, I'm just not sure how to make it perform well enough. Yet.

[disturbed13]

dude i would rip that thing appart and check the blocks
if your tubes look like that
then your blocks must look................
i would just check em

[OldChap]

Surely oil and water do not mix??? I thought that would create an emulsion and that would be worse than you have now.

[EtaCarinae]

Surely oil and water do not mix??? I thought that would create an emulsion and that would be worse than you have now.

You must not mix oil and water. I am mentioning a completely oil based circuit. There is no water at all inside. For that the tubing must be changed and all water purged.

[EtaCarinae]

That's something I've considered trying in a car for ages, except that oil transfers heat far slower than water (but still better than air). I still like the idea, I'm just not sure how to make it perform well enough. Yet.

You are right, but the heat absorption capacity of oil is much higher than water, in that oil can withstand very high temps which would not be acceptable to water. This could be a good solution if you plan to run your system hot.
This also means that you do not need a high flow of oil. Pumps can run a lower rpm, less coolant volume, smaller diameter tubing, etc. Nevertheless it may mean larger radiator to dissipate heat. An interesting subject for a study anyhow.

[karbonkid]

You are right, but the heat absorption capacity of oil is much higher than water, in that oil can withstand very high temps which would not be acceptable to water. This could be a good solution if you plan to run your system hot.
This also means that you do not need a high flow of oil. Pumps can run a lower rpm, less coolant volume, smaller diameter tubing, etc. Nevertheless it may mean larger radiator to dissipate heat. An interesting subject for a study anyhow.

The Specific Heat Capacity of water is pretty much unbeatable (4200 J/K iirc) - that's precisely why we use it in heating and cooling systems. Oil is nowhere near as efficient.

[EtaCarinae]

The Specific Heat Capacity of water is pretty much unbeatable (4200 J/K iirc) - that's precisely why we use it in heating and cooling systems. Oil is nowhere near as efficient.

You are right karbonkid and I am wrong. Your data is correct and for oil we are at 2000. The only thing that beats water largely is Hydrogen gas. At over 14'000 that is more than three times water. But I would not advise to use it and risk to blow up crash and burn the whole system like the Hindenburg. A fireball is not the best way to crunch, but to heat and light the room and the neighbourhood it is fine, albeit for a very short moment.

What mislead me about oil is that oil has still an advantage for high temps when water boils. When water boils and becomes gaseous then the value goes down to 1850. And having steam means a lot of other problems. Oil whatsoever remains liquid and boils at much higher temperatures which allows it to be used to cool systems when water is not good anymore.
So trn my advice to use oil is ok but you must run your cpus at 150 deg (bypassing tjunction limit). Overclock to 6 or 7 Ghz and with oil youre done.

No way out I am afraid trn you are stuck with water. I see your system is up and running again. Back to the race