Arctic Liquid Freezer II rev3

freeagentt

Limp Gawd
Joined
Dec 5, 2018
Messages
213
Thanks for the answers.

Sorry, but

isn't

in my book.


Your first image will not open to a size I can read.

Your "Physics and thermal transfer, water can't take away heat if the water temp is higher than the CPU tempsr" babble is simple common sense.

OKay, so someone else has told you how little coolant is in a CLC. Sorry, I don't do video reviews. I want written ones, but that's me. Custom loops have reservoirs hold as much often more than CLCs do.

Read what freeagentt posted about his H100 cooling ability over time on hot CPU. His H100 cooling degradation was mostly loss of coolant. Numerous times people ave lower improved CLC performance by removing waterblock base and topping up coolant.

It isn't the increased coolant temps that wears pump but the loss of coolant.
Increased coolant temp means increased temps in hose and plastic which is expansion so higher molecular action thus coolant can move from inside CLC to atmosphere easier .. meaning more coolant loss with higher coolant temps .. meaning less cooling ability of CLC .. and when coolant becomes low enough for air bubbles start moving through pump we have increased pump and seal wear.

.​

Apology accepted. Very easy for posts to be misunderstood. I apologize for how I replied to you.

I probably laughed more than I should have. It was fairly new when I was playing around with it like that. I did leave my D14 for it, so it wasn't horrible. Also was using my Panaflos on it before I took the stock fans out of the package, so a little better than stock performance :D

But.. I have LGMRT now.. no need for a CLC.. although I have looked a couple of times at them.. But they are expensive and I really want either a new CPU or GPU. My overclock cant really be improved on the CPU side of things so not much point.. it can but I don't want to hurt it.
 

doyll

[H]ard|Gawd
Joined
Feb 4, 2012
Messages
1,356
I probably laughed more than I should have. It was fairly new when I was playing around with it like that. I did leave my D14 for it, so it wasn't horrible. Also was using my Panaflos on it before I took the stock fans out of the package, so a little better than stock performance :D

But.. I have LGMRT now.. no need for a CLC.. although I have looked a couple of times at them.. But they are expensive and I really want either a new CPU or GPU. My overclock cant really be improved on the CPU side of things so not much point.. it can but I don't want to hurt it.
Le Grand Macho RT is one of the very best! I've used one quite a bit myself, but use TRUE Spirit 140 Power more. Performs as well for much less money. ;)
 

TheSlySyl

[H]ard|Gawd
Joined
May 30, 2018
Messages
1,220
Did anyone here get the AF 420? I'm curious if its performance matches its gigantic size.
 

Spartacus09

[H]ard|Gawd
Joined
Apr 21, 2018
Messages
1,809
The graph is in the article I thought I linked in a better size but I think it copied the small size pic twice here's the better one: https://www.overclockers.com/wp-content/uploads/2010/06/HWLabs-SR1360_HeatLoadChart-2.jpg
Your "Physics and thermal transfer, water can't take away heat if the water temp is higher than the CPU tempsr" babble is simple common sense.
Exactly you're asking for data, I'm just relaying the reasoning and common sense of my thought process.
OKay, so someone else has told you how little coolant is in a CLC. Sorry, I don't do video reviews. I want written ones, but that's me. Custom loops have reservoirs hold as much often more than CLCs do.
He does both, but they arent going to always have the same info, I don't think he mentions the amount of liquid in the CLC (the video i linked at the time starts when he drains it he notates the amount about a minute past that)
https://www.gamersnexus.net/hwreviews/3571-arctic-liquid-freezer-ii-cooler-review-benchmark
It isn't the increased coolant temps that wears pump but the loss of coolant.
Increased coolant temp means increased temps in hose and plastic which is expansion so higher molecular action thus coolant can move from inside CLC to atmosphere easier .. meaning more coolant loss with higher coolant temps .. meaning less cooling ability of CLC .. and when coolant becomes low enough for air bubbles start moving through pump we have increased pump and seal wear.
I feel like these things would correlate together though, it doesn't seem logical that a 10-15c over ambient or even 20-25c (assuming basically 100% workload 24/7) would cause such a substantial increase in permeation and/or evaporation (addtionally if it is getting that high I think better static pressure fans and/or higher rpm are in order :p).
I can't help, but still think with a top mount or a tubes down orientation the cooling and liquid to the pump would still be adequate (albeit decreased capability) even with a double digit percentage evaporation of the liquid.
I would think you would need close to 25% liquid evaporation to reach the point where it might start affecting it and introducing air/bubbles to the pump in those orientations.
At the same time if the water is getting to 40-45c then the CPU has bound to be running at near triple digit temps, so its either undersized aio or too large of a heat source.
IIRC most aio are only designed for a couple hundred watts of dissipation since they have such thin radiators, even the AFII is rated for only about 300w I believe and that'd be running full speed fans.
Apology accepted. Very easy for posts to be misunderstood. I apologize for how I replied to you.
I appreciate as well.

Did anyone here get the AF 420? I'm curious if its performance matches its gigantic size.
Pretty sure its just going to increase the heat soak time not necessarily increase performance unless it has a different pump and/or tubes.
 

Nenu

[H]ardened
Joined
Apr 28, 2007
Messages
19,626
Did anyone here get the AF 420? I'm curious if its performance matches its gigantic size.
Yeah its as good as the radiator size increase.
I got the best overclock yet on my 6700K and memory, both, with this cooler.
Before that, I could either max out my memory OR cpu overclock, never both together.
With the AF420 I got an even higher overclock on both, at the same time.
 

doyll

[H]ard|Gawd
Joined
Feb 4, 2012
Messages
1,356
Did anyone here get the AF 420? I'm curious if its performance matches its gigantic size.
With CLCs larger radiator sizes do help, but not as much as in custom loops. While I can't prove it I believe the reason is their pump's extremely low (40-60 L/h) coolant flowrate doesn't move coolant through waterblock fast enough compoounded by cheap waterblocks that don't transfer heat from block to coolant. as well as waterblocks that cost as much or more than entire CLC costs.

Spartacus09

I will reply to you in the morning. It's midnight where I am and I'm up at 6.
 

Nenu

[H]ardened
Joined
Apr 28, 2007
Messages
19,626
With CLCs larger radiator sizes do help, but not as much as in custom loops. While I can't prove it I believe the reason is their pump's extremely low (40-60 L/h) coolant flowrate doesn't move coolant through waterblock fast enough compoounded by cheap waterblocks that don't transfer heat from block to coolant. as well as waterblocks that cost as much or more than entire CLC costs.

Spartacus09

I will reply to you in the morning. It's midnight where I am and I'm up at 6.
The rate of flow has to be offset with the surface area of the block vs surface area of the radiator, as well as the amount of water in the system.
If any one of them gets saturated a lot more than the other, cooling will be reduced.
 

Danja

Limp Gawd
Joined
Dec 7, 2010
Messages
195
I'm getting max CPU temps during gaming (Skyrim VR) of 89c with a 5800x + AF240. Cinebench R23 Multicore also gets it to 88-89 (modded SkyrimVR is very CPU single core-intensive). My room temperature is ~19c. This is with all stock settings in the BIOS aside from enabling DOCP, and switching the CPU fan curve to silent (but the curve still maxes out by ~70c IIRC). Idle temp is at 60c, but was ~45-50 before I made the fan curve silent. Am I wrong to think that seems a bit high, even for a 5800x? Also, are you guys getting per-core monitoring in HWMonitor? I used to on my 4790k, but now I have a bunch of sensors that don't seem to obviously match up to cores.
 
Last edited:

doyll

[H]ard|Gawd
Joined
Feb 4, 2012
Messages
1,356
The graph is in the article I thought I linked in a better size but I think it copied the small size pic twice here's the better one: https://www.overclockers.com/wp-content/uploads/2010/06/HWLabs-SR1360_HeatLoadChart-2.jpg

Exactly you're asking for data, I'm just relaying the reasoning and common sense of my thought process.

He does both, but they arent going to always have the same info, I don't think he mentions the amount of liquid in the CLC (the video i linked at the time starts when he drains it he notates the amount about a minute past that)
https://www.gamersnexus.net/hwreviews/3571-arctic-liquid-freezer-ii-cooler-review-benchmark

I feel like these things would correlate together though, it doesn't seem logical that a 10-15c over ambient or even 20-25c (assuming basically 100% workload 24/7) would cause such a substantial increase in permeation and/or evaporation (addtionally if it is getting that high I think better static pressure fans and/or higher rpm are in order :p).
I can't help, but still think with a top mount or a tubes down orientation the cooling and liquid to the pump would still be adequate (albeit decreased capability) even with a double digit percentage evaporation of the liquid.
I would think you would need close to 25% liquid evaporation to reach the point where it might start affecting it and introducing air/bubbles to the pump in those orientations.
At the same time if the water is getting to 40-45c then the CPU has bound to be running at near triple digit temps, so its either undersized aio or too large of a heat source.
IIRC most aio are only designed for a couple hundred watts of dissipation since they have such thin radiators, even the AFII is rated for only about 300w I believe and that'd be running full speed fans.

I appreciate as well.


Pretty sure its just going to increase the heat soak time not necessarily increase performance unless it has a different pump and/or tubes.
Top down is best location, but problem isn't just location. But that doesn't change the fact CLCs are the absolute cheapest possible build cost & quality liquid cooling that can be made only just having enough cooling ability to do the job. It's not uncommon to be able for outflow from waterblock/pump to feel warmer than intake hose with high wattage CPU on CLCs. This never happens in custom loops.

Advertising hype uses the fact water cooling (custom loops) are best cooling available to push the sale. But comparing a custom loop to CLC is like comparing a Rolls Royse to a rickshaw.

A 25% loss of coolant will leave some of the fin area filled with air instead of coolant with rad in horizontal (top) position, so rad is no longer functioning at 100%.

Advertising hype uses the fact water cooling (custom loops) are best cooling available to push the sale. But comparing a custom loop to CLC is like comparing a Rolls Royse to a rickshaw.


Entire CLC costs about the same as one quality component in a custom loop.

CLC pump flowrate is about the same as average health adult urine flowrate.
 

Nenu

[H]ardened
Joined
Apr 28, 2007
Messages
19,626
I'm getting max CPU temps during gaming (Skyrim VR) of 89c with a 5800x + AF240. Cinebench R23 Multicore also gets it to 88-89 (modded SkyrimVR is very CPU single core-intensive). My room temperature is ~19c. This is with all stock settings in the BIOS aside from enabling DOCP, and switching the CPU fan curve to silent (but the curve still maxes out by ~70c IIRC). Idle temp is at 60c, but was ~45-50 before I made the fan curve silent. Am I wrong to think that seems a bit high, even for a 5800x? Also, are you guys getting per-core monitoring in HWMonitor? I used to on my 4790k, but now I have a bunch of sensors that don't seem to obviously match up to cores.
Thats terrible for a water cooled CPU.
Try placing the radiator outside your PC and see how much difference that makes.
Work from there.

ie
If it makes little difference its a problem with mounting to the CPU or the cooler has a problem.
If it makes a big difference your case and/or coolers fan airflow is up the spout.
 

freeagentt

Limp Gawd
Joined
Dec 5, 2018
Messages
213
How old is your power supply? It looks like its almost time to replace it. I have a couple of old 850s that will dip down to 11.4v. Can only run X58 with X5690 @ stock. Gaming kinda iffy. Be careful.
 

Danja

Limp Gawd
Joined
Dec 7, 2010
Messages
195
How old is your power supply? It looks like its almost time to replace it. I have a couple of old 850s that will dip down to 11.4v. Can only run X58 with X5690 @ stock. Gaming kinda iffy. Be careful.
Is this in reference to my post?

Thats terrible for a water cooled CPU.
Try placing the radiator outside your PC and see how much difference that makes.
Work from there.

ie
If it makes little difference its a problem with mounting to the CPU or the cooler has a problem.
If it makes a big difference your case and/or coolers fan airflow is up the spout.

It'll be tricky to place the rad outside the PC but I'll give it a shot. The chip itself is kind of weird though. I changed the BIOS fan curve from silent to 100%. If I just boot up and launch HW monitor, after a minute the chip settles down to 33c. But as I'm typing now it's been bouncing between 45 and 55c. It seems really variable.

I currently have the AIO plugged into the CPU fan socket rather than the AIO socket, as suggested by the manufacturer. It's a bit annoying that there's only one cable - it looks nice, but I'm assuming the pump also slows down when the fans do; is that correct?
 

Grebuloner

Gawd
Joined
Jul 31, 2009
Messages
967
Did anyone here get the AF 420? I'm curious if its performance matches its gigantic size.
I've got one on my 10980XE. Fantastic performance and not too noisy, cores range 70s-80s C at 330W, and that isn't even at full speed.

My D15 was over 100C with the same load and 100% fans.
 

Spartacus09

[H]ard|Gawd
Joined
Apr 21, 2018
Messages
1,809
A 25% loss of coolant will leave some of the fin area filled with air instead of coolant with rad in horizontal (top) position, so rad is no longer functioning at 100%.
Not refuting that, I'm sure it does suffer a substantial performance hit, but my argument is that that it is likely to continue to functioning in that orientation as the pump isn't running completely dry.
 
Top