Loop fail

[Powell]

Hi guys, 

 

Ive built a new pc recently and have experienced  a complication.

 

Parts list:

• thermaltake p7 case

• Gigabyte aorus motherboard

• gigabyte aorus gpu 

• thermaltake psu 1050 plat

• team force ram

• thermaltake pacific radiators (allum)

• thermaltake pacific fittings 

• thermalktake resivour D5 pump combo kits

• thermaltake petg pipe

 

I think thats all that is relevant, as far as i can tell the only cross metal contamination is the copper fittings and alloy rads. The copper fittings are painted though with no metal exposed.

 

The gpu and cpu both seem to have mostly plastic composition of the waterblock.  Ive tried researching what the heat transfer plates are made of and if they are exposed to the coolant but have not made any significant findings. 

( if someone knows i would appreciate the info, im guess either nickel or some other material similar) which would add to the cross contamination.

 

So i got the pc built, ive taken my time doing work in small doses checking everything twice and re making anything that does not look 100%

 

On startup no issues, installed windows and away she goes. spent some time installing nividia platform/ updated drivers. Few other bits temp monitors system run down and rgb software (anti rainbow software) ha.

 

Prior i did leak test th system for a few hours with the psu jumped and no power anywhere else, just with distilled water.

 

Three days gaming no issues and excellent performance.

 

Get home from  night shift pc will turn on but turns straight back off with a waring stating a leak has been detected, instant panic check every fitting check blocks shine a torch around mobo/ gpu. No issues dry as a whistle. So i google and find out the mobo has a very well know issue to throw false codes. Suggestions to either unplug the rgb/leak detect plug or unscrew then re install leak detect screw. Well i turn the pc off before it shuts down to tired to try the fix so i go to bed thinking it ca wait. Also while this was going on the pc would only boot if the water loop was unplugged from psu, so all fans/rgb/pumps run off two headers. 

 

Get up later in the day an elbow has blown off water everywhere on the floor and desk nothing on the pc except small dots of spray that have  dried up, very small but also many of them.

 

Fix loop, pc wont boot. I cant remember if  it wouldnt boot of the morning but i think it would. So this leads me to think something has been damaged by the water. Mobo or psu. 

 

So my partners brother who works in IT and pc repair is coming next week to test components and help me out in general. 

 

What im looking for here is an explanation on the pressure build up, i have three theorys in mind so far.

 

• 1 pumps on to high speed. Both were on setting 5 which is 160 litres an hour in my case. 

 

• 2 air getting in some how loose fitting ect

 

• 3 mixed metals. Allum copper and possible nickle but all surfaces seem painted or coated.

 

Ive since turned the pump do to setting 2 which is roughly 70-80 lph.

Replaced coolant with the stuff i originally bought but was waiting to make sure i was happy with all tube placemnet before using, this stuff has corrosion inhibitors  and biocides.

Re bent 2 tubes that were maybe a fraction off dead staight onto fitting ao they are now 100%.

 

Ive only been able to run it for 20min since then and seems to have less pressure but is also cold and after a short run time.

 

What did i do wrong guys? Well i know what i did wrong but did it cause the build up of pressure?

 

 

Photos of the loop, blocks pipes the lot.

Of it before and after coolant. Of new glow speed, of fitting that blew. And of the loop in test mode with psu jumped.

[Tristerin]

Im not saying I know what happened but:

 

Galvanic Corrosion (mixing metals in a loop in this case) can be beaten by science if you are using a proper corrosion inhibitor.  That also includes the items not coming in physical contact (I.E. copper on aluminum).  I have 3 mixed metal loops, and use only chrome plated fittings on my aluminum rads to stop corrosion from occurring where touching, and I read you used painted?  I also use radiator fluid 50/50 premix to ensure the system itself is properly corrosive inhibited.

 

EDIT - I see that its in fact two loops, not parallel pumps

[Powell]

Hi, yes the fittings seem painted as i have scuffed a spare quite roughly and now see copper. It may be they are coated but being black i assume paint. Maybe anodized or something?

 

I have full pre mix thermaltake cryo fuel in the loop now that supposedly uses corrosion inhibitors biocides.  

 

# edit,  the pumps were running really fast to, like stirring the distill up to a state of milky air looking water and you could hear the water running through all the rads and tubes.

Now its backed off it is almost silent and looks alot more stable. 

[Tristerin]

1 minute ago, Powell said:

Hi, yes the fittings seem painted as i have scuffed a spare quite roughly and now see copper. It may be they are coated but being black i assume paint. Maybe anodized or something?

 

I have full pre mix thermaltake cryo fuel in the loop now that supposedly uses corrosion inhibitors biocides.  

Which would be fine, in terms of galvanic corrosion occurring, if there was no bare metal to bare metal contact.  Anodized is not a "barrier" of protection I would trust, as I have no mixed metals loops experience with it (I buy all my stuff individually from china and make my own loops) - Chrome is a physical barrier and why I choose it when using non aluminum fittings with aluminum radiators. 

[Powell]

4 minutes ago, Tristerin said:

Which would be fine, in terms of galvanic corrosion occurring, if there was no bare metal to bare metal contact.  Anodized is not a "barrier" of protection I would trust, as I have no mixed metals loops experience with it (I buy all my stuff individually from china and make my own loops) - Chrome is a physical barrier and why I choose it when using non aluminum fittings with aluminum radiators. 

This is very upsetting haha, i understand what your saying. One of the main design factors was the rgb fittings so it would be a shame to lose them. Would a option like thread tape be viable?

 

Also would this have caused enough pressure in 48 hours to blow off the fitting? 

I did edit the previous post to if you didnt notice. 

 

Thanks for your help also

[Tristerin]

6 minutes ago, Powell said:

This is very upsetting haha, i understand what your saying. One of the main design factors was the rgb fittings so it would be a shame to lose them. Would a option like thread tape be viable?

 

Also would this have caused enough pressure in 48 hours to blow off the fitting? 

I did edit the previous post to if you didnt notice. 

 

Thanks for your help also

 

The only expertise I have is building my own homebrew mixed metals loops (cut one open that had been running 24/7 for over 6 months, rads and aluminum and copper blocks 1.5 weeks ago and posted all about it here on LTT and how to stop galvanic corrosion and proof of mixing metals and how to avoid GC hehe) so my advice would be:

 

I would not use thread tape - that's only meant to act as a filler, like thermal paste just in case there is a spot the water can leak through - its not a barrier between metals

If you have the money, see what it would take for a local shop to chrome dip the fittings.  But I wouldn't choose that option either because Ive ran a chrome dip machine, its not a very thick barrier at all that can still be scuffed compared to something completely encased in chrome.

 

It may not have been to much pressure - what if its just a loose fitting?  However Galvanic corrosion CAN occur immediately and depending on the environment really start to tear down the aluminum (aluminum is where the GC should occur first) so if your radiators are aluminum I would check where the fittings go and (specifically the one that blew out) and see if there is damage in the threads.

[Powell]

 I assume pressure based on the way it blew off and when i drained the other intact loop it also had a fair amount of pressure built up. The first time i ran the loop up today also prior to turning pump speed down, i opened res to top up coolant and found massive pressure to. Since backing off the pump it has decreased significantly. 

 

I will look at getting the fittings chromed, or maybe copper radiators if i have to. Dont mind spending some more money for thr right result.

 

You have given me alot to think about and trial/ monitor results.

 

Thanks for taking the time to help! 

 

[SolarNova]

D5 pumps running at 1 or 2, are actually rather slow. I run my D5 at its max settings (5), and have done since i built it.

You want to aim for about 1 to 1.5 GPM in flow. Thats roughly 230 to 340 LPH. 70-80 LPH isnt near enough.

 

Sounds like you just got a loose fitting tube tbh. You have coolant temp read outs so u can check to ensure the coolant isnt heating up to high causing pressure to build.

 

When building a new loop u need to regularly pop the open the res to help bleed the system, You may need to do this regularly for a week or so before all the micro bubbles and air pockets that are going to work through the system themselves do so completely.

[For Science!]

Opaque fluid and aluminium? 

[Powell]

On 11/3/2019 at 1:21 AM, SolarNova said:

D5 pumps running at 1 or 2, are actually rather slow. I run my D5 at its max settings (5), and have done since i built it.

You want to aim for about 1 to 1.5 GPM in flow. Thats roughly 230 to 340 LPH. 70-80 LPH isnt near enough.

 

Sounds like you just got a loose fitting tube tbh. You have coolant temp read outs so u can check to ensure the coolant isnt heating up to high causing pressure to build.

 

When building a new loop u need to regularly pop the open the res to help bleed the system, You may need to do this regularly for a week or so before all the micro bubbles and air pockets that are going to work through the system themselves do so completely.

Hey, thanks i had opened the caps a few times to let the pressure out. Maybe not enough? 

The fitting was not loose any more then it should be. But the pressure was very high, i can say the one thing that really seems out of place was the pressure. 

[Powell]

On 11/3/2019 at 5:32 AM, For Science! said:

Opaque fluid and aluminium? 

Yes, ek cryo fuel. I had distilled water to begin, replaced with the cryo fuel as it has all the necessary additives and looks quite nice. Says compatible with aluminium on the bottle also. 

[For Science!]

2 minutes ago, Powell said:

Yes, ek cryo fuel. I had distilled water to begin, replaced with the cryo fuel as it has all the necessary additives and looks quite nice. Says compatible with aluminium on the bottle also. 

This is not the case unforuntalty, as far as I know, all opaque coolants have bad reactions with aluminium parts. As a minimum this is probably precipitating and clogging your loop to some degree and probably contributing to pressure buildup.

 

https://www.ekwb.com/shop/ek-cryofuel-solid-cloud-white-premix-1000-ml

Quote

Not suitable for aluminum-based liquid cooling loops!

 

[Powell]

14 minutes ago, For Science! said:

This is not the case unforuntalty, as far as I know, all opaque coolants have bad reactions with aluminium parts. As a minimum this is probably precipitating and clogging your loop to some degree and probably contributing to pressure buildup.

 

https://www.ekwb.com/shop/ek-cryofuel-solid-cloud-white-premix-1000-ml

 

Strange, says on bottle its good for alloy. Also the pressure was with distilled water since putting the opaque coolant in ive experienced very little pressure build up. I have only ran it for a short while though.

[Gegger]

I'm just going to pop in and say your build looks sick.

 

I was never here

[LOLZR]

I had a tube pop off a fitting one time, and it was because I didn't put the O-ring that was in the package onto the tube between the base piece and the collar piece - could this be why?

[Powell]

On 11/5/2019 at 12:55 AM, LOLZR said:

I had a tube pop off a fitting one time, and it was because I didn't put the O-ring that was in the package onto the tube between the base piece and the collar piece - could this be why?

O ring was on, Im thinking it was because i did not bleed the loop enough now. No pressure thus far with the new coolant.

[Powell]

On 11/4/2019 at 6:18 AM, For Science! said:

This is not the case unforuntalty, as far as I know, all opaque coolants have bad reactions with aluminium parts. As a minimum this is probably precipitating and clogging your loop to some degree and probably contributing to pressure buildup.

 

https://www.ekwb.com/shop/ek-cryofuel-solid-cloud-white-premix-1000-ml

 

Hi ! You have enlightened me actually. I spent more hours researching many forums to Find out the dead set truth about all this. It seems for $508 Australian i can get 6 copper replacement rads EK brand and 8 lengths of petg tube i need to re do the pipes. Ive experienced no more pressure but im the paranoid type and would rather everything be copper and not have to worry. After what seems like losing a $1000+ mobo to a watwr leak, cant confirm yet though, i don't want to take any chances. 

Thanks again for your help, excuse my ignorance haha. 

[Powell]

I really have to ask, are you guys actually confident that now im replacing my tt alloy rads with ek copper ones. Not mixing metals or running ek opaque coolant in alloy, that i will no longer see the pressure build up?? 

 

This just seems so strange to me but ive blown the loop twice now so im hesitant as to what i shoulf do.. besides run it in test mode for 12 hours and see if it goes pop.

[For Science!]

8 minutes ago, Powell said:

I really have to ask, are you guys actually confident that now im replacing my tt alloy rads with ek copper ones. Not mixing metals or running ek opaque coolant in alloy, that i will no longer see the pressure build up?? 

 

This just seems so strange to me but ive blown the loop twice now so im hesitant as to what i shoulf do.. besides run it in test mode for 12 hours and see if it goes pop.

I dont like opaque fluids generally but in a all copper/brass/nickel system you should be fine, nothing should build enough to blow a pop.

 

Having said that running for a couple of hours for testing is good practice generally 

[Powell]

3 hours ago, For Science! said:

I dont like opaque fluids generally but in a all copper/brass/nickel system you should be fine, nothing should build enough to blow a pop.

 

Having said that running for a couple of hours for testing is good practice generally 

Also thought id add, i have inspected the waterblock and other parts. No build up yet. Im not quite sure but i was thinking the metal reaction wont build pressure itself will it? Is that not a side effect of clogged components which i don't have. or does the pressure also build due to reactions. 

 

Is there a non opaque white coolant I can use, i just dont want to lose the aesthetics of the rgb fittings. 

[For Science!]

20 minutes ago, Powell said:

 

"White" is by definition non opaque, so unfortunately there is no coolant that will satisfy this condition. However you can sand down tubing to produce a satin finish, this will diffuse RGB lighting and functionally be similar to having opaque coolant without having it.

 

https://bit-tech.net/features/modding/watercooling/how-to-make-your-own-satin-hardline-tubes/1/

 

The pressure if probably coming from gas release as a biproduct of the reaction between the sacrificial zinc layer in the aluminium radiator (or perhaps the alu itself) and the coolant. As I posted above, at least one other person has had pressure build up in a static loop (i.e. it was just sitting there closed, without running)

 

[Powell]

On 11/21/2019 at 12:35 AM, For Science! said:

"White" is by definition non opaque, so unfortunately there is no coolant that will satisfy this condition. However you can sand down tubing to produce a satin finish, this will diffuse RGB lighting and functionally be similar to having opaque coolant without having it.

 

https://bit-tech.net/features/modding/watercooling/how-t

On 11/21/2019 at 12:35 AM, For Science! said:

"White" is by definition non opaque, so unfortunately there is no coolant that will satisfy this condition. However you can sand down tubing to produce a satin finish, this will diffuse RGB lighting and functionally be similar to having opaque coolant without having it.

 

https://bit-tech.net/features/modding/watercooling/how-to-make-your-own-satin-hardline-tubes/1/

 

The pressure if probably coming from gas release as a biproduct of the reaction between the sacrificial zinc layer in the aluminium radiator (or perhaps the alu itself) and the coolant. As I posted above, at least one other person has had pressure build up in a static loop (i.e. it was just sitting there closed, without running)

 

o-make-your-own-satin-hardline-tubes/1/

 

The pressure if probably coming from gas release as a biproduct of the reaction between the sacrificial zinc layer in the aluminium radiator (or perhaps the alu itself) and the coolant. As I posted above, at least one other person has had pressure build up in a static loop (i.e. it was just sitting there closed, without running)

 

Hey thanks again, if tou happen to have any suggestions on coolant that might work well with the rgb im all ears. Im going to run the cryofuel for now as i already have it and really dont mind a bit of preventative maintenance.

[For Science!]

33 minutes ago, Powell said:

Hey thanks again, if tou happen to have any suggestions on coolant that might work well with the rgb im all ears. Im going to run the cryofuel for now as i already have it and really dont mind a bit of preventative maintenance.

EK-Cryofuel Clear (Transparent) + Satin finish tubing would be my recommendation.