Do close looped coolers ever leak

Comixbooks

Comixbooks

Fully [H]
Joined
Jun 7, 2008
Messages
22,019
I seen a CPU on overclock.net saying when Liquid cooling and CPUs don't mix (showing a CPU damaged from water cooling) thinking of picking up a Intel watercooler on the Retail Edge website.
 
Though rare, yes, they have been known to leak. Though typically it isn't a leak as much as it is a hose bursting. However, the later generations of CLCs tend to have more robust tubing, mitigating that risk.

On the same token, CPUs damaged by watercooling are extremely rare as well, and most damage that occurs is either due to inexperience or negligence rather than a part failure.
 
Tsumi said:
Though rare, yes, they have been known to leak. Though typically it isn't a leak as much as it is a hose bursting. However, the later generations of CLCs tend to have more robust tubing, mitigating that risk.

On the same token, CPUs damaged by watercooling are extremely rare as well, and most damage that occurs is either due to inexperience or negligence rather than a part failure.
Click to expand...

^This.
 
Some manufacturers offer very long warranties, and even replace hardware damaged by a broken cooler on a case-by-case basis. Corsair is good about this, and has a 5-year warranty on their all-in-one / closed loop water cooling units.

One thing about traditional water cooling is that it requires maintenance. You can't just set it up once and run it for 5+ years without a second thought. With AIO / closed loop water coolers, you can.
 
I pretty much did just that. Only reason my loop was disturbed was due to a pump failure. You can get away with minimal to no maintenance if you pick components that will not degrade.
 
Last edited:
CLCs are not as great as some like to think. They cool about the same as top tier air coolers, cost more, generally make much more noise and the most common failures are pumps. Pumps wear out with use and coolant does escape from systems right through the hosing so after a year or so systems tend to need topped up, and coolant needs lubricant, conditioner, inhibitors replenished every year or two. But CLCs have no provision for replacing worn out pumps or topping up coolant, etc. So when coolant level gets low or pump fails, it's buy a complete new system .. which is a major expense, not to mention you have no cooling and can't use your system until replaced.

Now a good air cooler only has the fan/s that can fail. If a fan does fail, system will still work at low load and any fan, even a case, fan can be attach with rubber bands or whatever until suitable replacement arrives.
 
doyll said:
CLCs are not as great as some like to think. They cool about the same as top tier air coolers, cost more, generally make much more noise and the most common failures are pumps. Pumps wear out with use and coolant does escape from systems right through the hosing so after a year or so systems tend to need topped up, and coolant needs lubricant, conditioner, inhibitors replenished every year or two. But CLCs have no provision for replacing worn out pumps or topping up coolant, etc. So when coolant level gets low or pump fails, it's buy a complete new system .. which is a major expense, not to mention you have no cooling and can't use your system until replaced.

Now a good air cooler only has the fan/s that can fail. If a fan does fail, system will still work at low load and any fan, even a case, fan can be attach with rubber bands or whatever until suitable replacement arrives.
Click to expand...

I think there's more than a few people here with CLC's that disprove your opinion. Due to the layout of my system, if I was low on coolant I would get temp throttle and shutdown immediately as my CPU is the highest point in the loop. 2.5 years of use, and I'm still running the same temps as I did when I first installed it.
 
Jeremy C said:
I think there's more than a few people here with CLC's that disprove your opinion. Due to the layout of my system, if I was low on coolant I would get temp throttle and shutdown immediately as my CPU is the highest point in the loop. 2.5 years of use, and I'm still running the same temps as I did when I first installed it.
Click to expand...

Quite a few people have the misconception that CLCs cool better than air, which is not the case. All the points he brings up are valid concerns that people typically don't think about when buying a CLC. However, I believe that coolant loss would not be significant until about 10 years or longer afterwards. Pump failure is the primary point of failure on CLCs.
 
It has happened, but it is not a regular occurrence.

I have seen very few catastrophic failures. In fact, I can only think of one, in which a user on these forums had a pump seize and it burst soaking the inside of the case.

From what I recall, Corsair showed what good customers service is like, stood behind their product and replaced all of his damaged parts.

They have been very robust, much better than I would have expected, based on the number of complaints I have seen online.

I have had several and never had a problem. Right now I am using a H110i GTX for my CPU, and I am about to install two H90's, one for each GPU.

I feel perfectly comfortable using them. At the same time we know the AIO pumps don't last forever. Corsairs models are typically warrantied for 5 years. I wouldn't use them much beyond that. Everything wears out over time, and when there is water involved, you don';t want to take any unnecessary risks.

I'd replace them after the warranty runs out.
 
Tsumi said:
Quite a few people have the misconception that CLCs cool better than air, which is not the case. All the points he brings up are valid concerns that people typically don't think about when buying a CLC. However, I believe that coolant loss would not be significant until about 10 years or longer afterwards. Pump failure is the primary point of failure on CLCs.
Click to expand...
I agree pumps are the most common failure point. CLCs will probably last 3-5 years, and coolant loss is slow enough to last that long. Pump flow slows with impeller wear, again, 2-5 years is reasonable expectation.

Cost of CLCs is significantly higher than air coolers. We can build a great air cooled system with case fans running automatic speed control just like CPU and GPU cooler fans so system is silent under low load use and definitely not loud even at full load.

Most of the people who tout how great they are have never used one for more than a year before getting something else. Some of my air coolers have been in constant use for 7 or 8 years.
 
doyll said:
CLCs are not as great as some like to think. They cool about the same as top tier air coolers, cost more, generally make much more noise and the most common failures are pumps. Pumps wear out with use and coolant does escape from systems right through the hosing so after a year or so systems tend to need topped up, and coolant needs lubricant, conditioner, inhibitors replenished every year or two. But CLCs have no provision for replacing worn out pumps or topping up coolant, etc. So when coolant level gets low or pump fails, it's buy a complete new system .. which is a major expense, not to mention you have no cooling and can't use your system until replaced.

Now a good air cooler only has the fan/s that can fail. If a fan does fail, system will still work at low load and any fan, even a case, fan can be attach with rubber bands or whatever until suitable replacement arrives.
Click to expand...

I don't know about louder. Mine are pretty much silent.

I hear some people complain about pump noise, but i have never heard the pump on any of mine, without putting my ear right on it.

Efficiency wise air coolers are more efficient at cooling.

For the same size radiator, with the same case air temperature, an air cooler will be more effective, as there are fewer efficiency losses in a simple direct system.

The benefits of water coolers is that you can fit a bigger radiator than you can with an air cooler, and that if venting hot air out of the case, so the internal temperature doesn't rise as much.

These can more than make up for the efficiency difference.

I'm not aware of any air cooler that is competitive with my h110i gtx.
 
Tsumi said:
Quite a few people have the misconception that CLCs cool better than air, which is not the case. All the points he brings up are valid concerns that people typically don't think about when buying a CLC. However, I believe that coolant loss would not be significant until about 10 years or longer afterwards. Pump failure is the primary point of failure on CLCs.
Click to expand...

"Quite a few people have the misconception that CLCs cool better than air"

Is the NH-D15 still considered the best air cooler on the market? Assuming it is, then there are several AIOs that outperform it easily and not costing much more. Unless of course I'm missing something in the meaning of "cool better".
 
Eldata said:
Some AIOs are even LOA (leaking on arrival) like here;

http://forum.corsair.com/v3/showthread.php?t=153068
Click to expand...

Oh, come on, don't scare him needlessly :p

That was just for that one model. It had a defect in which if it reached really cold temperatures during shipping (cold truck in freezing weather, or airplane cargo hold at 30,000ft) the coolant could freeze and crack some of the plastic fittings resulting in leaks.

Corsair did a full recall on all of them, and have since replaced them with fixed units.

Just to be on the safe side, check before installing, but it would be very unlikely to come across one.
 
Zarathustra[H];1041948718 said:
Oh, come on, don't scare him needlessly :p

That was just for that one model. It had a defect in which if it reached really cold temperatures during shipping (cold truck in freezing weather, or airplane cargo hold at 30,000ft) the coolant could freeze and crack some of the plastic fittings resulting in leaks.

Corsair did a full recall on all of them, and have since replaced them with fixed units.

Just to be on the safe side, check before installing, but it would be very unlikely to come across one.
Click to expand...

Actually it was the H110i GT (not the H100i GTX) that had that issue and recall;

http://forum.corsair.com/v3/showthread.php?t=137973

I agree with those who argue that liquid cooling is a more perilous option but disagree with them when it comes to cooling performance. Compare the form factor, cooling ability and price between say an H100i GTX and NH-D15...No way in hell would I want an NH-D15.

http://tinyurl.com/pyf4282
http://tinyurl.com/oh8y3y3
http://tinyurl.com/pgh7cmu
 
Here is testing including actual air temp going into cooler. While it out performs air at full speed (sounding 6x as loud), H110i is only 1c cooler than top air which are 2-7c quieter.
Georges%20data%20w%20all%20coolers_zpscmcly46l.png



@Eldata
The chart above shows R1, PH-TC14PE and NH-D15 all perform the same. We can add several others similar like Silver Arrow SB-E, Silver Arrow IB-E, HE01, etc. Others with performance not more than 3c warmer are Assassin and Assassin 2, be quiet! Dark Rock Pro 2 & 3, TRUE Spirit 140 Rev.A & Power, Archon, NH-D14, NH-U14S, Megahalems, etc.
 
doyll said:
Here is testing including actual air temp going into cooler. While it out performs air at full speed (sounding 6x as loud), H110i is only 1c cooler than top air which are 2-7c quieter.
Georges%20data%20w%20all%20coolers_zpscmcly46l.png



@Eldata
The chart above shows R1, PH-TC14PE and NH-D15 all perform the same. We can add several others similar like Silver Arrow SB-E, Silver Arrow IB-E, HE01, etc. Others with performance not more than 3c warmer are Assassin and Assassin 2, be quiet! Dark Rock Pro 2 & 3, TRUE Spirit 140 Rev.A & Power, Archon, NH-D14, NH-U14S, Megahalems, etc.
Click to expand...

Who is George Cella?
 
Eldata said:
Actually it was the H110i GT (not the H100i GTX) that had that issue and recall;

http://forum.corsair.com/v3/showthread.php?t=137973

I agree with those who argue that liquid cooling is a more perilous option but disagree with them when it comes to cooling performance. Compare the form factor, cooling ability and price between say an H100i GTX and NH-D15...No way in hell would I want an NH-D15.

http://tinyurl.com/pyf4282
http://tinyurl.com/oh8y3y3
http://tinyurl.com/pgh7cmu
Click to expand...

So you would spend $20 more for something that lasts only 3 - 5 years and gives you what 5 degrees more cooling? If that 5 degrees is the difference in stability sure I see your point, but I would not go that route myself.

I have no issue with anyone choosing water cooling AIO or otherwise, but I do not see any proof that the AIO is better this is just personal choice.

I have air coolers that have been running for 8 years and still do thier job well enough, I prefer the set it and forget it setup. Each have their pros and cons, one is not better
 
Eldata said:
Who is George Cella?
Click to expand...

OCN user ciarlatano, of course! Says so right in the chart! :p

Either way, who cares who he is.

If we look at some comparison charts from the latest [H] cooler testing, we see that the best air cooler (The Cryorig R1 Ultimate) is below the middle of the pack.

1444758383plyIjgazLB_3_2.png


Now, it could be that the [H] just haven't tested any better ones, but if you have any ideas on which coolers to look into, I'd be curious.

AS I mentioned above, I'm with you. Air coolers are simpler, single stage, and as such are more efficient with the same size radiator, and the same temperature differential.

The trick with AIO coolers is that - while inherently less efficient - they make up for this by having more swept radiator area, and the fact that they blow the heat outside the case (or suck cold air in) meaning that you don't have case heat buildup that worsens the performance.
 
zaniix said:
So you would spend $20 more for something that lasts only 3 - 5 years and gives you what 5 degrees more cooling? If that 5 degrees is the difference in stability sure I see your point, but I would not go that route myself.

I have no issue with anyone choosing water cooling AIO or otherwise, but I do not see any proof that the AIO is better this is just personal choice.

I have air coolers that have been running for 8 years and still do thier job well enough, I prefer the set it and forget it setup. Each have their pros and cons, one is not better
Click to expand...

You're preaching to the choir.
 
George used to review for a well know review site.
http://lmgtfy.com/?q=George+Cella

The chart posted by Zarathustra[H] with clcs' being significantly better than best air is very misleading. I'll bet a pint that the baseline air temp is room, not the actual air temp going into cooler .. and that the case is not flowing enough air and air coolers are recycling their own heated air trying to cool the CPU. When air cooler and CLCs are all ingesting the same temperature air the results do not look like that.

And I'll bet another pint those CLC temps almost all come at the price of 2x -8x more noise.
 
Last edited:
Zarathustra[H];1041948830 said:
OCN user ciarlatano, of course! Says so right in the chart! :p

Either way, who cares who he is.

If we look at some comparison charts from the latest [H] cooler testing, we see that the best air cooler (The Cryorig R1 Ultimate) is below the middle of the pack.

Now, it could be that the [H] just haven't tested any better ones, but if you have any ideas on which coolers to look into, I'd be curious.

AS I mentioned above, I'm with you. Air coolers are simpler, single stage, and as such are more efficient with the same size radiator, and the same temperature differential.

The trick with AIO coolers is that - while inherently less efficient - they make up for this by having more swept radiator area, and the fact that they blow the heat outside the case (or suck cold air in) meaning that you don't have case heat buildup that worsens the performance.
Click to expand...

Is there even a case in most cooler reviews?
 
doyll said:
George used to review for a well know review site.
http://lmgtfy.com/?q=George+Cella

The chart posted by Zarathustra[H] with clcs' being significantly better than best air is very misleading. I'll bet a pint that the baseline air temp is room, not the actual air temp going into cooler .. and that the case is not flowing enough air and air coolers are recycling their own heated air trying to cool the CPU. When air cooler and CLCs are all ingesting the same temperature air the results do not look like that.

And I'll bet another pint those CLC temps almost all come at the price of 2x -8x more noise.
Click to expand...

Why do you believe a case is used for cooler reviews and comparisons?
 
The [H] believes very strongly in real world testing.

As such they use a Corsair Carbide case with good airflow for their testing.

As far as what case fans are used and how many, we'd have to ask the cooling editor.

I am pretty confident that they have selected a case and case fans providing decent air flow.
 
For what it's worth, my "old" Corsair H100 (1st generation) AIO served me well, cooled better than most air coolers and I never had a failure or leak in the last 4 years. I upgraded to the EK Predator 360 with this new build because I feared it wouldn't last much longer and the EK can be completely disassembled/drained or upgraded with the upcoming pre-filled QDC GPU blocks, unlike most other AIOs.
 
Eldata said:
"Quite a few people have the misconception that CLCs cool better than air"

Is the NH-D15 still considered the best air cooler on the market? Assuming it is, then there are several AIOs that outperform it easily and not costing much more. Unless of course I'm missing something in the meaning of "cool better".
Click to expand...

Outperform it easily? Not so fast. The reviews that show those CLCs typically have 2000-2500 RPM fans strapped to them set on high, while the NH-D15 has 1200-1500 RPM fans. Put both to the same noise level, and the NH-D15 cools as well, if not better than the top CLCs.

Eldata said:
Actually it was the H110i GT (not the H100i GTX) that had that issue and recall;

http://forum.corsair.com/v3/showthread.php?t=137973

I agree with those who argue that liquid cooling is a more perilous option but disagree with them when it comes to cooling performance. Compare the form factor, cooling ability and price between say an H100i GTX and NH-D15...No way in hell would I want an NH-D15.

http://tinyurl.com/pyf4282
http://tinyurl.com/oh8y3y3
http://tinyurl.com/pgh7cmu
Click to expand...

Again, the metric generally ignored is noise/performance. CLCs do not have that advantage.

Zarathustra[H];1041948830 said:
OCN user ciarlatano, of course! Says so right in the chart! :p

Either way, who cares who he is.

If we look at some comparison charts from the latest [H] cooler testing, we see that the best air cooler (The Cryorig R1 Ultimate) is below the middle of the pack.

1444758383plyIjgazLB_3_2.png


Now, it could be that the [H] just haven't tested any better ones, but if you have any ideas on which coolers to look into, I'd be curious.

AS I mentioned above, I'm with you. Air coolers are simpler, single stage, and as such are more efficient with the same size radiator, and the same temperature differential.

The trick with AIO coolers is that - while inherently less efficient - they make up for this by having more swept radiator area, and the fact that they blow the heat outside the case (or suck cold air in) meaning that you don't have case heat buildup that worsens the performance.
Click to expand...

Once again, the noise/performance metric is something that is typically ignored when it comes to reviews of CPU coolers, especially when just the stock fans are used (air coolers typically have 1500 RPM max fans, while CLCs typically have 2500 RPM max fans).

They could be so much more efficient if they used better components, but that would drive up the costs. Aluminum for radiator fins is something the watercooling community ditched a long time ago, but found its way back with CLCs due to their much cheaper cost. I also doubt that the waterblock is optimized to transfer heat well, it's optimized for cheap manufacturing and low restriction for the weak pump.
 
Tsumi said:
Quite a few people have the misconception that CLCs cool better than air, which is not the case.
Click to expand...

Really, I don't see the distinction at all. They are both ultimately using air to perform the cooling. Most modern air coolers already have heatpipes. I tend to think of the water tubes in an AIO as just big heat-pipes. Overall the biggest benefits of AIO Closed loop coolers are being able to have a larger cooling surface than what would fit directly over the CPU, not worrying about ram clearance issues, improving overall case airflow, etc.
 
Eldata said:
"Put both to the same noise level, and the NH-D15 cools as well, if not better than the top CLCs."

Performance:
http://tinyurl.com/pgh7cmu

Noise:
http://tinyurl.com/o83knjt

My preferences:
Extreme OC - LC
Mild OC - Simple compact air heat pipe costing less than $30.00
No OC - Stock cooler
Click to expand...

You just proved my point. The NH-D15 is 1 C hotter for 1 dB less noise compared to the H100i. The only CLC it flat out loses to is the H110i, by a small margin. 3 C hotter at the same noise.

GotNoRice said:
Really, I don't see the distinction at all. They are both ultimately using air to perform the cooling. Most modern air coolers already have heatpipes. I tend to think of the water tubes in an AIO as just big heat-pipes. Overall the biggest benefits of AIO Closed loop coolers are being able to have a larger cooling surface than what would fit directly over the CPU, not worrying about ram clearance issues, improving overall case airflow, etc.
Click to expand...

Ultimately almost all computer cooling systems are using air. However, the method of heat transfer is what makes the difference, so it is important to say there is a distinction. And it is important to note that in terms of raw performance, CLCs typically have the edge only because they use high speed fans. At the same noise levels, they typically perform worse than equivalent cost air coolers. It is also important to note that because of their low price, shortcuts had to be made, and their performance is typically lower than a custom loop of the same size, though the custom loop will almost definitely cost 1.5 to 2 times as much.

I guess what it comes down to is that you get what you pay for. With a CLC, you're trading performance for convenience and a different form factor as compared to an equivalent cost air cooler. You also won't get the performance of a real watercooling system with a CLC.
 
What is Guru3D's test system?
They say air temp is room ambient. I much prefer the temp of air going into radiator or fin pack. Even bench testing the room air temp and actual intake air temp are often not the same.
 
"CLCs typically have the edge only because they use high speed fans. At the same noise levels, they typically perform worse than equivalent cost air coolers"

Someone who's into serious OC is probably going to want better cooling performance, regardless of noise, at some point. This review highlights the shortcoming of air coolers when it comes to that since it focuses more on raw performance;

http://tinyurl.com/p92hde9

OCCT just about disqualifies all the air coolers from the extreme CPU test..."Finally our 4.6GHz test. Don't be fooled, this is an extreme test and the graph reflects this, you will only see the very best featured in this graph. If we really want to measure outright performance, this is where we do it."

After the dust settles, there is nothing stopping one from retuning the AIO's fans for less noise. Who really cares if the CPU is running just as hot or few degrees hotter than the best air cooler provided noise levels are satisfactory and heat is properly contained?

Noise!..The last refuge of the "air benders".
 
yes if your an extreme overclocker then water is your choice.

The point I have tried to make is they are both situational, the average user with stock to mild overclock is going to have an easier time with Air cooling.

You might also be someone who wants a large radiator on a silent system at stock or small overclock and can tune the fans way down.

These things give us choices. Personally my ears pick up the pump noise from most if not all of these units and it drives me batshit crazy. I hate the sound it grate son my nerves. Maybe I need a silent case.


For me the #1 reason to pick air cooling is not ever having to worry about replacing the entire cooler because of pump failure or reduced performance due to evaporation. I am not an extreme overclocker either, if I was I would go with custom unit so I could add more fluid and replace only the pump when needed rather than the entire thing. it also looks way cooler.
 
@Eldata
Not saying they didnt' get those results, but they are not testing coolers on their own merits. Unless they are monitoring the air temperature at the intake of radiator / cooler fin pack (they are not because their ambinet temps are obviously the room), they are only testing their system's performance with different coolers, so unless we are using exactly the same system the results mean little. Also, there is no noise levels graphed to show how extremely loud the H100 is .. and the H100, H100i, H100i GTX, or whatever version is not as good as many other CLCs, yet their tests show it as one of the best.[/INDENT]
 
Not that I want to get TOO deep into the debate I will say that when I got my H100 it was only $55. It took my underload temps from 80 to 38 and my idle from 50 to 27. I may not have had the best air cooler as it was just the stock one, but I have a hard time believing an air cooler could drop temps by half for the small price of $55.
 
auntjemima said:
Not that I want to get TOO deep into the debate I will say that when I got my H100 it was only $55. It took my underload temps from 80 to 38 and my idle from 50 to 27. I may not have had the best air cooler as it was just the stock one, but I have a hard time believing an air cooler could drop temps by half for the small price of $55.
Click to expand...
It's stories like yours that mislead others. :D
Sorry, but you stepped right into the deep end with your 42c / 23c lower temps. :D
For temps to go from 80&50c to 38&27c tells me you had a serious airflow or TIM problem with your system. Instead of fixing the problem you replaced the cooling system and expound on how great the H100 is.
The reality is unless your case had almost no airflow a have decent air cooler would have done the job just fine.
Add to this that for $55 I'm better it was a refurb, meaning short warranty and no coverage for component damage at all.
And where do you think all those "refurbs'" come from? They are either CLCs that failed or they are being liquidated as "reburb" to facilitate a much shorter and more limited warranty. ;)
 
I had good air flow before but went from just 2 fans removing hot air on top of my tower to 4 fans doing a push pull through the 240mm rad.

I am sure this helped.
 
To add to that, I didnt find even the 80 degrees to be too steep. I replaced the cooler simply because of the $55 cost. Dropping temps was solely a bonus for me. We could argue what caused the change in temperatures all day. Those that want to believe one side or the other are unlikely to be swayed here on HardForums.

My pump makes no noise that I can tell from putting my ear up to it.. but my 4 H1A fans certainly do. Its like a jet engine in that case.
 
LOL. Yeah, they are loud.

CLCs have their place (like in the junk pile). :D there are a few decent CLCs. AIOs with the ability to add / change components (not CLCs) are even better. I have a H80i, used it with GT fans for awhile, now it's in a box somewhere. Would like to try AIO like Swiftech H240-X or EK WB Predator 360, but probably won't due to their cost being about 3x what a good air cooler is.
 
To me, the biggest advantage of AIO coolers is that you can expel the hot air straight out of the case.

In my planned build, I have a H110i GTX for my CPU, and two H90's (one for each GPU, using the HG10 N980's I just got)

This way I never need to worry about air circulation inside the case. The heat from the CPU will never reach the GPU's and the heat from the GPU's will never reach the CPU.

Of what value is it that your air cooler is slightly quieter, if you need to make up for the fact that you are dumping heat inside the case, by running additional noisy case fans? :p
 
That seems to be the sole argument for an air cooler. Noise.

But really, you could have LESS fans and run them slower with an AIO. Atleast thats my opinion on that.
 
You must log in or register to reply here.
Back
Top