SOLVED Self-contained liquid cooling?

[rrohbeck]

Just wondering: Has there ever been any comparison between liquid cooling and high end air? The only one I can remember was the H50 which totally sucked in frostytech's list. Now I'm wondering if any 120mm rad is worth it.
Asetek specifies 0.124 K/W for the 570LC, i.e. H80 but frostytech's measurements indicate 0.09 K/W for high end air.

 

[Knufire]

AFAIK, H80 and H100 are the only ones who beat high end air, and just by a few degrees, making them priced exactly what they're worth.

 

[rrohbeck]

AFAIK, H80 and H100 are the only ones who beat high end air, and just by a few degrees, making them priced exactly what they're worth.

But according to the data, the H80 should be worse than high end air.
As long as Corsair doesn't send one to frostytech I don't want it.

 

[muddocktor]

If you look at any of my later reviews published on the front page, you will see how the Corsair H80 stacks up against the other heatsinks I tested at 4 GHz on my test system. The test system is the same setup for all heatsinks listed in each review and are temperature compensated on the average temps, so they are directly comparable.

 

[rrohbeck]

If you look at any of my later reviews published on the front page, you will see how the Corsair H80 stacks up against the other heatsinks I tested at 4 GHz on my test system. The test system is the same setup for all heatsinks listed in each review and are temperature compensated on the average temps, so they are directly comparable.

Ah cool. I didn't know there was such a big difference between the H70 and H80. Now I'll have to make up my mind if replacing my NH-D14 is worth it...

 

[Knufire]

If you already have an NH-D14, IMO it's not worth the money to replace it. You're only getting a few degrees cooler...

 

[RockstarX74]

Yes. There is a Molex to the pump header, and the fans come off the pump itself. Meaning ALL power to the pump and the fans come off the one molex.

Nice!

The H80 can't be any easier to install. Bet my wife can do it. With her checking email and her crafting forum, don't think she would have any issues installing an H80 or worrying about heat issues. She's pretty smart......

What every man needs! A good woman who knows her way around a PC!

 

[RockstarX74]

At the risk of sounding like even more of a yo-yo then I already do... Would I be better off going with the H100 and a set of Delta 120 x 25mm High-Speed PWM Fan - 150.33 CFM(yeah I found some decent 25mm high-pressure fans ) than my previous decided upon setup?

 

[rrohbeck]

At the risk of sounding like even more of a yo-yo then I already do... Would I be better off going with the H100 and a set of Delta 120 x 25mm High-Speed PWM Fan - 150.33 CFM(yeah I found some decent 25mm high-pressure fans ) than my previous decided upon setup?

I learned a lot from muddocktor's H80 review at http://www.overclockers.com/corsair-h80-review.
Putting fast fans on the H80 made 5C difference under high load. Now the question is if you'd be pushing the H100's cooling capacity. I doubt that TBH.

 

[rrohbeck]

If you already have an NH-D14, IMO it's not worth the money to replace it. You're only getting a few degrees cooler...

Aye. I searched a little more and came across this: http://www.anandtech.com/show/5054/corsair-hydro-series-h60-h80-and-h100-reviewed/5

This basically tells me that with stock fans H80, H100 and high end air are basically equivalent and *maybe* you can push liquid coolers farther with high powered fans.
What do you folks say about the dBA/CFM tradeoff for the SanAce 9CR1212P0G03 monster versus the 38mm Delta fans? I'll try to optimize the fan setup on my NH-D14 to get maximum airflow at a reasonable noise level. At full speed the Delta AFC1212DE-PWM is just a bit too loud for my taste, and of course the question is how much airflow the NH-D14 can actually use before being limited only by the heatpipes' thermal resistance, not heat exchange at the fins. I guess I'll have to run more experiments. Maybe a shroud for the Delta fan will be next.

 

[mjw21a]

If you do in fact go with those Delta monsters, make sure you get longer screws at your local hardware store. The ones which come with the H100 aren't long enough for a 38mm fan.

I ended up setting my H100 to the lowest performance setting in order to make my Delta's into something I could live with. They DO have awesome static air pressure, thing is my system didn't exactly run very hot to begin with. The H100 is actually quite adequate with its stock fans.

It's just that I've come full circle to where I was about 10 years ago. I now have 3x computers for various different uses, and I just wanted the best possible cooling for my gaming rig that didn't involve ging to a full custom loop. Noise isn't such an issue for a gaming rig as you've always got your headset on, however yes, you can hear the fans through the headset.

The Delta's are awesome just to play around with, along with the fact people ask whats in the rig. Take the side off and there are two nice big fat fans with wire grilles on them, look pretty tough ;op

EDIT: I'd be concerned about that 150CFM fan you linked to. It has the potential to blow your mainboards fan headers. The 253CFM fans don't have that issue due to having 2 pins for PWM control and a separate connector to draw power direct from the PSU.

Ah cool. I didn't know there was such a big difference between the H70 and H80. Now I'll have to make up my mind if replacing my NH-D14 is worth it...

Not really mate. I don't think there's any reason for you to replace that cooler at all. It's still one of the best on the market.

 

[RockstarX74]

I learned a lot from muddocktor's H80 review at http://www.overclockers.com/corsair-h80-review.
Putting fast fans on the H80 made 5C difference under high load. Now the question is if you'd be pushing the H100's cooling capacity. I doubt that TBH.

I noticed in the review that one of the conditions of the test was:

"All energy saving features of the motherboard and processor were turned off to keep it from down clocking the processor speed and vcore"

Do the latest motherboards have a tendency of doing this? (ie. AUSUS P8Z68-V Pro with the TPU and EPU turned on) I watched a video recently that gave me the impression this wouldn't hinder overclocking.

 

[RockstarX74]

If you do in fact go with those Delta monsters, make sure you get longer screws at your local hardware store. The ones which come with the H100 aren't long enough for a 38mm fan.

I was planning on using those with the H80. The 150CFM fans are only 25mm so I thought they would work great with the H100 considering I only have 60mm's of clearance at the top of my case before they would interfere with the motherboard. but not to ignore your point if I do go the H80 route with the 38mm Delta's i'll be sure to pick up some screws that will work with them.

EDIT: I'd be concerned about that 150CFM fan you linked to. It has the potential to blow your mainboards fan headers. The 253CFM fans don't have that issue due to having 2 pins for PWM control and a separate connector to draw power direct from the PSU.

I wasn't planning on hooking them directly onto the motherboard. I planned on hooking them into the PWM plugs on the H100 pump block. Would they also prove to be problematic going that route?

 

[mjw21a]

Not sure. I know I went with the Delta's linked on eBay as I was concerned with killing the Corsair H100, and it's not exactly cheap. I wasn't willing to take the risk.

One thing I've noticed is that high airflow, high pressure fans don't make the same difference they used to 8 or so years ago. Back then, you could take something like the ThermalTake PIPE101 (a lapped on pictured in my avatar) and fit a high airflow cooler on it for great performance. Anything less and the air wouldn't actually penetrate the tightly packed fins. That was great for an overclock back then and once lapped beat out the old Thermalright cooler that was king back then running the same fan. Considering how coolers now seem to have a lower fin density and the fins don't actually go as deep (they're being designed for low noise operation) things don't seem to make as much of a difference. Low noise operation seems to be the primary consideration these days.

I wonder, if a tower heatsink we designed with the same fin density as the old PIPE101, how well would it perform? I suspect that if fitted with a high airflow fan it'd smash the Corsair H100 for a six.

Do they not make coolers for people who don't care about noise anymore?

 

[muddocktor]

I noticed in the review that one of the conditions of the test was: Do the latest motherboards have a tendency of doing this? (ie. AUSUS P8Z68-V Pro with the TPU and EPU turned on) I watched a video recently that gave me the impression this wouldn't hinder overclocking.

I still turn off the energy saving features on most of my systems, just from habit and the fact that I also run DC projects (BOINC/Rosetta), which keep the processor loaded to 100% anyways. But I do have a couple of SB systems that I haven't disabled C1E on that don't seem to have been affected on their ability to overclock to any great extent. As for my test system, I have to disable all energy saving features, since I am testing as a worst case scenario for heat production and also to keep the heatload as constant as possible. Otherwise my test results wouldn't mean much, since I wouldn't be keeping the maximum amount of heat being input into the heatsink or LCLC system. And that is also the reason why I didn't do my actual H80 testing with the fans plugged into the pump/coldplate unit for power, but rather plugged into the motherboard headers or fed from a molex connector in the case of the AP-31 fans. I looked at the unit with it set up for maximum heat removal and wasn't worried about noise testing.

 

[RockstarX74]

One thing I've noticed is that high airflow, high pressure fans don't make the same difference they used to 8 or so years ago.

My guess would be that heatsink fin/rad designs have become much more efficient over the years so the difference gap isn't there like it used to be. But I'm not an airflow egnineer so i may be way off base!

 

[mjw21a]

My guess would be that heatsink fin/rad designs have become much more efficient over the years so the difference gap isn't there like it used to be. But I'm not an airflow egnineer so i may be way off base!

My belief is that cpu heat hasn't increased at the same rated as heatsink progress. Probably the biggest leap forward was the tower design which heatpipes allowed. Allows the air to pass through with much less resistance.

I still believe however that if you packed the fins a lot tighter and deeper then you'd get extremely effective cooling with a powerful fan, however it would be very poor when combined with a regular fan.

I'd like to see this done though it really would be a niche product.

 

[RockstarX74]

As for my test system, I have to disable all energy saving features, since I am testing as a worst case scenario for heat production and also to keep the heatload as constant as possible. Otherwise my test results wouldn't mean much, since I wouldn't be keeping the maximum amount of heat being input into the heatsink or LCLC system.

Ahhh... makes sense. It was just the way the statement was worded gave me the impression that these energy efficiency aspects of the motherboard were a hindrance to getting the most out of overclocking.

 

[RockstarX74]

My belief is that cpu heat hasn't increased at the same rated as heatsink progress. Probably the biggest leap forward was the tower design which heatpipes allowed. Allows the air to pass through with much less resistance.

A very plausible explanation... Todays CPU's & MB's are much more energy efficient as well which means less heat.

Give it time and self-contained water cooling solutions will rival traditional water cooling. I'm astounded at how fast technology has come since I bought my first PC back in '96. I thought I was just styling with a 166Mhz CPU, 256MB RAM and a 1.6GB HDD

 

[mjw21a]

Lol, a bit like my first rig. A 33mhz 386sx (the dx had cache), 2mb RAM, 90mb hard drive and woot, I had 256 colours and a SoundBlaster 16 with a double speed CD ROM drive. Cutting edge stuff I tell ya ;o)

Could play cool games like Commander Keen, Lemmings etc