Overclock.net › Forums › Cooling › Water Cooling › (56K Warning)*NEW* AQUACOMPUTER Airplex Modularity System Radiators
New Posts  All Forums:Forum Nav:

(56K Warning)*NEW* AQUACOMPUTER Airplex Modularity System Radiators - Page 18

post #171 of 279
Quote:
Originally Posted by Chunky_Chimp View Post

What about the red and blue versions of the rads? Are we ever going to see those?
You can already get them. Because of the massive amount of possible combinations it will take a while until everything is listed in the shop. If you want one with a specific color, just add a note at the end of the comment field. Available colors are red, black and blue. Here is a photo to get an idea how it looks like:

ams_farbig.jpg

ams_farbig_schwarz.jpg

ams_farbig_rot.jpg

ams_farbig_blau.jpg

ams_farbig_edelstahl.jpg
post #172 of 279
Hmm, that red isn't as bright as I was expecting, but still good. smile.gif
Deep Blood
(16 items)
 
  
CPUMotherboardGraphicsRAM
Intel Core i7 920 D0 (Batch # 3901A240) Foxconn X58 Blood Rage (non-GTI) eVGA GeForce GTX670 2GB 6GB Corsair Dominator 1.6GHz TR3X6G1600C8D Rev 3.2 
Hard DriveOptical DriveCoolingOS
Samsung SSD830 128GB / Seagate 640GB+1TB+2TB+4TB Samsung SH-S223Q 22X XTTH | OCZ Freeze | 9G1212H1011 Windows 7 Ultimate x64 SP1 
MonitorKeyboardPowerCase
Dell U2410f (rev. A02) Code 104 w/O-rings + Vortex doubleshot keycaps Antec HCPP-850 CaseLabs SM8 
MouseMouse PadAudioOther
Logitech G500 Sharkoon 1337 Asus Xonar DX + Swan D1080MkII08 + MXL Tempo Lifecam Cinema 
  hide details  
Reply
Deep Blood
(16 items)
 
  
CPUMotherboardGraphicsRAM
Intel Core i7 920 D0 (Batch # 3901A240) Foxconn X58 Blood Rage (non-GTI) eVGA GeForce GTX670 2GB 6GB Corsair Dominator 1.6GHz TR3X6G1600C8D Rev 3.2 
Hard DriveOptical DriveCoolingOS
Samsung SSD830 128GB / Seagate 640GB+1TB+2TB+4TB Samsung SH-S223Q 22X XTTH | OCZ Freeze | 9G1212H1011 Windows 7 Ultimate x64 SP1 
MonitorKeyboardPowerCase
Dell U2410f (rev. A02) Code 104 w/O-rings + Vortex doubleshot keycaps Antec HCPP-850 CaseLabs SM8 
MouseMouse PadAudioOther
Logitech G500 Sharkoon 1337 Asus Xonar DX + Swan D1080MkII08 + MXL Tempo Lifecam Cinema 
  hide details  
Reply
post #173 of 279
Quote:
Originally Posted by cpachris View Post

I expressed my admiration for Martin's efforts in a previous post, so I'm glad you agree with me on that. But you can have admiration for the efforts and dedication of someone, and at the same time use your own brain to interpret the results he brings us. By equalizing pumping power for rads that have very different levels of restrictiveness, the test will favor less restrictive rads. If you instead equalized flow rate, the test would favor more restrictive rads. Either way you chose to test, it would favor a certain group of rads. There is really almost no way around getting some skewing of the results. Martin chose a method that he believes represents the most "real world" scenario. And it probably does...for most people. But for those on OCN that have two strong pumps, the testing method won't show how well a restrictive rad could perform. My guess is Martin would agree.

The way I see it, you can't control flow in a loop. You can change your blocks, pumps, rads and whatever but, still, you can't control your flow. Which means that the logic way to test it is to assume a certain pumping power and see how different rads fare. Sure, one could argue that it ain't fair that this rad gets lower flow than the others but, in reality, it will ALWAYS be like that, because its a lot more restrictive than the others.

It is what it is wink.gif


Quote:
Originally Posted by Martinm210 View Post

Wow, just wanted to stop by and say howdy all and thanks for all the kind comments. For what it's worth, while my flow meter reading capability failed, using the pressure drop curves I would estimate my flow rate on any of the radiators was never much below 1.3GPM or higher than 1.6GPM. I purposely used a fixed pumping condition so the flow rates could vary as I feel that's behaving more similarly to the real world than fixing a single flow rate. Also fixing a single flow rate if anything in my opinion skews the results toward the higher restriction radiator and more artificial than real world. If 1.3GPM is a flow rate inadequate for any heat exchange component (block or radiator), that is too bad on the design side.thumb.gif
As with anyone that doesn't like a particular test...opinions are worthless without the data to back it up and I invite anyone to share something constructive. Put up (As in get off your moldy rear and start testing and sharing). Opinions are a worthless waste of cyberspace..smile.gif
No my testing is not absolute or perfect and yes you too can contribute something for others to review and criticize. If you choose to complain without sharing some of your own work for sake of comparison, then you have not put on your big boy panties..wink.gif Some go straight from Huggies to Depends and never step up and share..

Hey Martin, good to see you around wink.gif
I remember that I asked some time ago about how water flow affects radiator performance and was shown this graphs (was told by Vapor that they are yours):

badiz.png

Any chance you still have the whole article to look into it? thumb.gif Or could you explain this a little more? wink.gif

Quote:
Originally Posted by YOUDFDO View Post

I tune it for block effeincy . There is only 1 right way.

You do it wrong, then. The more flow you have, the better. Of course, there is a point in which the pumps (which dump more heat the more flow they produce) dump more heat than the gains you get from such flow increase, but there is no tuning to do, as flow never hurts.

Quote:
Originally Posted by YOUDFDO View Post

Heres what it really comes down to . lets use serial pumps vs parr. Serials pumps create zero extra pressure and zero extra flow . but you have redundancy. If I want redundancy I will create it in the system I will not run 2 pumps in serial . If you have a pressure drop or flow stoppage use an alarm . easly set up with pressure guages . Most preferr flow gauges and well they should . I preferr a pressure guage once I know the flow of a loop I can check pressure and use that from that point on. Point is people debate this all the time . It is the dumbest PC water cooling debate of all time.

You are just proving once and again how limited your knowledge is.

1) Serial pumps: when the pumps are in serial, you get DOUBLE pressure. This means that you will get some added flow as a consequence, but the max flow is probably the same as if you were running 1 pump and had a 0 restriction loop.

2) Parallel pumps: you get DOUBLE max flow. The problem is that given our loops that are fairly restrictive the gains might not be that good (if any) and thus why its better to run the pumps in serial.

In both cases, you get redundancy.

seriesparallelconclusion2.png?w=614

Another great article from Martin wink.gif

http://martinsliquidlab.org/2011/04/26/pump-setup-series-vs-parallel/2/
My System
(16 items)
 
  
CPUMotherboardGraphicsGraphics
i5 2500k MSI P67 GD65 GTX460 1GB GTX460 1GB 
RAMHard DriveCoolingOS
G. Skill Ripjaws Crucial C300 Prolimatech Megahalems + 2 x GT AP15 Win 7 x64 
MonitorKeyboardPowerCase
Dell U2311H Noppoo Choc mini Corsair HX620 Microcool Banchetto rev. 2.0 black aluminium 
MouseMouse PadAudio
Zowie EC1 Zowie Fiio E7 + E9 / Sennheiser HD 25 1 II 
  hide details  
Reply
My System
(16 items)
 
  
CPUMotherboardGraphicsGraphics
i5 2500k MSI P67 GD65 GTX460 1GB GTX460 1GB 
RAMHard DriveCoolingOS
G. Skill Ripjaws Crucial C300 Prolimatech Megahalems + 2 x GT AP15 Win 7 x64 
MonitorKeyboardPowerCase
Dell U2311H Noppoo Choc mini Corsair HX620 Microcool Banchetto rev. 2.0 black aluminium 
MouseMouse PadAudio
Zowie EC1 Zowie Fiio E7 + E9 / Sennheiser HD 25 1 II 
  hide details  
Reply
post #174 of 279
Since AMS is such a restrictive radiator and will reduce flow in your cooling loop compared to using other triple rads regardless of whether you use 1 pump or 2 pumps, you also have to factor in the fact this reduced flow vs using other rads will reduce performance at cpu block.

The loss in performance at cpu waterblock using AMS with 1 pump or 2 pumps compared to much less restrictive rads in loop is certain, but any gains at rad side is still speculative, but even if you do get some small improvement if AMS is much more responsive to higher flow than other rads, such gains are likely to be lost at cpu waterblock.

Especially considering Martin said his tests had minimum of 1.3GPM with AMS rad. My loop has 2 mcp 355 with full cover gpu block, cpu block, 2 rads, and I have 2.1 gpm flow. With AMS rad, I would have much less flow than 2.1gpm, I just dont see AMS rad making up for any lost performance at cpu block in a real loop with 2 pumps.
My System
(15 items)
 
  
CPUMotherboardGraphicsRAM
6900K 4.4ghz  MSI X99A Krait Titan XP Gskill 32GB 3200 C14 
Hard DriveHard DriveCoolingOS
Samsung 960 Pro 1TB OS Samsung 850 pro 1TB games D5, XT45 280 rad, Nemesis 280 rad,  Windows 10 64bit 
MonitorPowerCase
Acer Predator XB271HU 144hz Seasonic Prime Titanium 850 Fractal Design R5 
  hide details  
Reply
My System
(15 items)
 
  
CPUMotherboardGraphicsRAM
6900K 4.4ghz  MSI X99A Krait Titan XP Gskill 32GB 3200 C14 
Hard DriveHard DriveCoolingOS
Samsung 960 Pro 1TB OS Samsung 850 pro 1TB games D5, XT45 280 rad, Nemesis 280 rad,  Windows 10 64bit 
MonitorPowerCase
Acer Predator XB271HU 144hz Seasonic Prime Titanium 850 Fractal Design R5 
  hide details  
Reply
post #175 of 279
366

So here's a question... If the water goes in 5 pipes, out 5 pipes, in 5 more, and back out the last 5; why in the world are those side channels done linearly? The pipes will be running the water in parallel, so wouldn't it increase the flow dramatically if you just routed out that plastic bar some that forces the water to run linearly? I'll see if I can diagram what I mean:
Code:
Current:
  .-.-.  .-.-.  .-.-.
==| | |==| | |==| | |==
==| | |==| | |==| | |==
==| | |==| | |==| | |==
==| | |==| | |==| | |==
==|   |==|   |==|   |==
  '---'  '---'  '---'


Better Flow:
  .-.  .-.  .-.
==| |==| |==| |==
==| |==| |==| |==
==| |==| |==| |==
==| |==| |==| |==
==| |==| |==| |==
  '-'  '-'  '-'


[Edit]
If it were me, instead of: Image (Click to show)
366

I would do: Diagram (Click to show)
Code:
.--------------.
| 4 3 3 3 3 3 4|
|4 2 2 2 2 2 4 |
| 4 1 1 1 1 1 4|
'--------------'

This way, you could design the end-caps to not restrict the water at all except for the 4th (last) pass, which would be a minimum of half the restriction of the current design.
Edited by Electrocutor - 6/6/12 at 6:14am
post #176 of 279
Yes, that flow rate sensetivity data is something I did a long time ago with my old radiator v1 test bench. Again one of those learning moments when I figured out that measuring water on the outlet only is bad if you vary flow rates at all because the water temp drop across a rad changes. It was also a big reason why I chose to ignore flow rate because it had relatively little to no change in the 1.5gpm range. I suppose that could be different with round tube radiators, but the more than two pass design is there to make up for that. I considered do more of that with V2 but lost my ability to measure flow rate when the XT died hope to buy another when I win the lottery, but I am really not interested enough...not even remotely interested in doing more really. I am on vacation and have done more than my share and really don't care. Bottom line..didn't write an article but I learned that you need to measure water at the inlet and the outlet and average the two and also that flow rate generally didn't have much influence. Fan speed was always the big variable so I focused my efforts on capturing that first and foremost.

Anyhow, I do wish we could get a few more reviews out there, rather than argue or complain about what we have..put the energy toward something constructive and share some data.thumb.gif

The more data and testing the better regardless of test methods or detail. Need more people to step up..thumb.gif

Perhaps flow or a particular fan type or whatever particular of any condition is a bigger variable than anticipated. We can sit on our butts and speculate and argue or we can do something to answer the question that is scientific and contsructive. The choice and opportunity is out there...
Edited by Martinm210 - 6/6/12 at 6:34am
    
CPUMotherboardGraphicsRAM
3930K ASrock X79 Ext7 Evga GTX570 CORSAIR Vengeance 16GB 
Hard DriveOptical DriveOSMonitor
Phoenix SSD+2TB Seagate drives ASUS DVD-RW combo Windows 7 28" Hanns-G HZ281 
KeyboardPowerCaseMouse
MS Natural Corsair TX650 Danger Den Torture Rack Logitech MX518 
Mouse Pad
Desk 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
3930K ASrock X79 Ext7 Evga GTX570 CORSAIR Vengeance 16GB 
Hard DriveOptical DriveOSMonitor
Phoenix SSD+2TB Seagate drives ASUS DVD-RW combo Windows 7 28" Hanns-G HZ281 
KeyboardPowerCaseMouse
MS Natural Corsair TX650 Danger Den Torture Rack Logitech MX518 
Mouse Pad
Desk 
  hide details  
Reply
post #177 of 279
Nope, because the idea is that the water with the most temperature difference use opposite sites.

550x366px-LL-9a500814_aquacomputer_airplex_15.jpeg

In this image, either the blue or the red get the water to the cap. Lets assume its the blue: once it reaches the end, it goes back through the left orange tubes, and returns through the right orange ones, to go again with the red ones.

In a conventional design, the rad goes all the way in, and then goes all the way back. This means that all the water uses 1/2 of the tubes at once, instead of this design in which the water uses 1/4 at once, and thus why the restriction is so high. Also, because the tubes aren't flat, I bet you don't get that good heat-transfer as they thought they would, and thus why not only its not a good performer, its also a too restrictive design.
My System
(16 items)
 
  
CPUMotherboardGraphicsGraphics
i5 2500k MSI P67 GD65 GTX460 1GB GTX460 1GB 
RAMHard DriveCoolingOS
G. Skill Ripjaws Crucial C300 Prolimatech Megahalems + 2 x GT AP15 Win 7 x64 
MonitorKeyboardPowerCase
Dell U2311H Noppoo Choc mini Corsair HX620 Microcool Banchetto rev. 2.0 black aluminium 
MouseMouse PadAudio
Zowie EC1 Zowie Fiio E7 + E9 / Sennheiser HD 25 1 II 
  hide details  
Reply
My System
(16 items)
 
  
CPUMotherboardGraphicsGraphics
i5 2500k MSI P67 GD65 GTX460 1GB GTX460 1GB 
RAMHard DriveCoolingOS
G. Skill Ripjaws Crucial C300 Prolimatech Megahalems + 2 x GT AP15 Win 7 x64 
MonitorKeyboardPowerCase
Dell U2311H Noppoo Choc mini Corsair HX620 Microcool Banchetto rev. 2.0 black aluminium 
MouseMouse PadAudio
Zowie EC1 Zowie Fiio E7 + E9 / Sennheiser HD 25 1 II 
  hide details  
Reply
post #178 of 279
Quote:
Originally Posted by Chunky_Chimp View Post

Hmm, that red isn't as bright as I was expecting, but still good. smile.gif

Since these panels are easily removed by user, you could paint/powdercoat yourself for that perfect color. I just might do that!
post #179 of 279
Quote:
Originally Posted by Martinm210 View Post

Anyhow, I do wish we could get a few more reviews out there, rather than argue or complain about what we have..put the energy toward something constructive and share some data.thumb.gif
The more data and testing the better regardless of test methods or detail. Need more people to step up..thumb.gif

That is definitely true. Bundymania had said months ago that he had already tested the AMS, and he stated that it was not the best performing rad, ie more along lines of your testing. But still waiting to see his actual results/flow,etc, have searched many times, but still havent seen yet. Though he may be taken a needed vacation as you. And we all appreciate all the work you guys have done, takes an enormous amount of thought and time.

His post is here.
269

I am wondering if some of lack of testing of AMS versus other rads, is lack of manufacturer willing to send out to be tested from those willing or lack of vendor sponsors willing to because of the cost of rad. And testers rightfully so dont want to fork out that kind of money for doing free testing.
My System
(15 items)
 
  
CPUMotherboardGraphicsRAM
6900K 4.4ghz  MSI X99A Krait Titan XP Gskill 32GB 3200 C14 
Hard DriveHard DriveCoolingOS
Samsung 960 Pro 1TB OS Samsung 850 pro 1TB games D5, XT45 280 rad, Nemesis 280 rad,  Windows 10 64bit 
MonitorPowerCase
Acer Predator XB271HU 144hz Seasonic Prime Titanium 850 Fractal Design R5 
  hide details  
Reply
My System
(15 items)
 
  
CPUMotherboardGraphicsRAM
6900K 4.4ghz  MSI X99A Krait Titan XP Gskill 32GB 3200 C14 
Hard DriveHard DriveCoolingOS
Samsung 960 Pro 1TB OS Samsung 850 pro 1TB games D5, XT45 280 rad, Nemesis 280 rad,  Windows 10 64bit 
MonitorPowerCase
Acer Predator XB271HU 144hz Seasonic Prime Titanium 850 Fractal Design R5 
  hide details  
Reply
post #180 of 279
Quote:
Originally Posted by prava View Post

The way I see it, you can't control flow in a loop. You can change your blocks, pumps, rads and whatever but, still, you can't control your flow. Which means that the logic way to test it is to assume a certain pumping power and see how different rads fare. Sure, one could argue that it ain't fair that this rad gets lower flow than the others but, in reality, it will ALWAYS be like that, because its a lot more restrictive than the others.
It is what it is wink.gif
Hey Martin, good to see you around wink.gif
I remember that I asked some time ago about how water flow affects radiator performance and was shown this graphs (was told by Vapor that they are yours):
badiz.png
Any chance you still have the whole article to look into it? thumb.gif Or could you explain this a little more? wink.gif
You do it wrong, then. The more flow you have, the better. Of course, there is a point in which the pumps (which dump more heat the more flow they produce) dump more heat than the gains you get from such flow increase, but there is no tuning to do, as flow never hurts.
You are just proving once and again how limited your knowledge is.
1) Serial pumps: when the pumps are in serial, you get DOUBLE pressure. This means that you will get some added flow as a consequence, but the max flow is probably the same as if you were running 1 pump and had a 0 restriction loop.
2) Parallel pumps: you get DOUBLE max flow. The problem is that given our loops that are fairly restrictive the gains might not be that good (if any) and thus why its better to run the pumps in serial.
In both cases, you get redundancy.
seriesparallelconclusion2.png?w=614
Another great article from Martin wink.gif
http://martinsliquidlab.org/2011/04/26/pump-setup-series-vs-parallel/2/


Sorry but it is possiable to change flow in a loop. I like The Aqua AMS raid . So much so I built a system around it . These are expensive systems. . My daughter will begin using the pump balance bar system in here builds were higher flowing raid are used. We actuall tune flow fopr the best flow efficiency threw the block . Since the pump reservoir system is desighned for very restrictive rads . we need away to control flow of higher flowing rads . And we have. Controling flow in a loop is vary easy to do We just took it to the next level . Increasing head in this maNNER IS USLESS. UNLESS YOU RUNNING SMALL PUMPS THAT CAN'T SUPPLY ENOUGH FLOW AND PRESSURE TO BLOCKS . Building a system out of spec for a single pump is wasteful. It can also be said that running 2 pumps in parr. will not produce 2x flow., This is a true statement under specific uses. Your serial pump adding head without flow is useless. It only means your pumps were to small to start with if you see a performance gain . . In serial lets put 1 pump at head of loop and the other at the rear of loop . What are the results. They say you shouldn't run a smaller pump with a larger pump in serial as it could cav. Both pumps need to run at same speed . Is this a true statement . Yes and know . Plenty of 2 stage pumps around saying thats a false statement and the fact I worked with these pumps for years.Raising head is ok but without flow its useless in a system that is correctly sized for the loop. I have built hydralic systems that run 4 50 HP 2 stage pumps. in parr. /serial all 4 pumps us a single header. turn on pump 1,100 psi turn on pump 2 1,1oo psi no change is head pressure . turn on pump 3 1.100psi no change is pressure . turn on pump 4 1,100 psi no change in pressure all 4 pumps are run into same headert no pressure change none zero none. The psi gauge at the end of loop read 900 psi for each pump and never changes as the others are turned on . But with only 1 pump I can't run the system not enough flow. I can run some machines but not the whole system . It takes 3 pumps to run the system. To say you can't change flow in a simple water cooling PC setup is nonsense . What exactly is it the these cpu block makers are doing with the input side of nozzel. There changing flow and pressure at the block base. So If the cpu block makers are tuning their blocks for efficiency why bother with restricting flow when the cpu block does the same. Thats a fair question that I won't ans.

Martin because you work with these pumps and like helping out the big companies. Do some water pump dead head test . I have never seen ya do this type of testing. Trust me martin it has its uses . Dead head 2 pumps in serial. and than do a single at same settings . (pressure Test Only) Martin The last chart above. Am I reading that correctly. The parr, system is running in a sigle loop ? WHY? That is a disfunctional way to go about it . Are we changing the rules to fit the PC world perspective or what how parr, series pumps work and the correct application of said configerations. In my parr. system you do not get redundancy and never will . Pump replacement at time of breakdown . Ya don't want people 1 pumping a system that needs 2. Also how you getting double flow with parr. and redundancy . It has to be the same loop and you are not doubling the flow its basicly a serial setup.

Ok martin You said I should show some test . I am going to do Dead head test in serial and single . I all ready know the results . You hint at them but you say so much more. I won't show my utube results until I see your dead head results. Than I think we can continue this discussion. Ya I know the results will be the same . BUT so many believe differantly . How did that happen Martin. Wife just chimed in she wants me to repeat the riser test . Not a big deal since the tubing is still on the telephone poll. Serial vs single head riser test . I believe Martin likes the med setting or the 3 setting so I will use that..

So @ the 3 setting It should be arounf 12 foot rise. For serial I should see a 24 foot rise / I will bet anybody any thing it doesn't even come close to 24foot rise.

God Martin you got betty angry, She wants to do parr. into single loop test water flow . We done all these test she knows you can not double flow if both pumps feed the same loop.
Edited by YOUDFDO - 6/7/12 at 9:46am
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Water Cooling
Overclock.net › Forums › Cooling › Water Cooling › (56K Warning)*NEW* AQUACOMPUTER Airplex Modularity System Radiators