Rad Rating Thread.

[Talonman]

UPDATED: We are assuming a few variables about your system...

Assume: 1.5GPM Flow Rate - (This would probably be close to what most systems run, and differences result in very little difference in radiator thermal performances).

Assume: 10C Water to Ambient Delta - This is pretty standard, I've seen a few cases of 15C for the 80mm rads, but 10C is pretty common.
Note:This is just how much hotter the coolant will be than the ambient temperature. People need to understand it's just a number, and if you want better cooling, this is one of the most important areas you can improve upon. It's not that hard to keep water/ambient deltas less than 5C if you have plenty of radiator capacity.

Assume: Nexxus Fans at 12V - While we probably see more people using Yate Loon D12SL12s, most of the available thermal curves for radiators have used the Nexus fans at 12V, yate loon data just is not available.


RAD USED________________________________ Approximated Thermal Dissipation.
ThermoChill PA120.1 Single 120mm fan ------------------------- 175w
ThermoChill PA160 Single 120mm fan --------------------------- 220w
ThermoChill PA120.2 Dual 120mm fans -------------------------- 345w
ThermoChill PA120.3 Triple 120mm fans ------------------------ 490w to 510w

Swiftech MCR120-QP Single 120mm fan ----------------------- 170w
Swiftech MCR120-QP-K Single 120mm fan --------------------- 170w
Swiftech MCR220-QP Dual 120mm fans ------------------------ 320w
Swiftech MCR220-QP-K Dual 120mm fans ---------------------- 320w
Swiftech MCR320-QP Triple 120mm fans ----------------------- 430w
Swiftech MCR320-QP-K Triple 120mm fans --------------------- 430w

Black Ice GT Stealth 120 Single 120mm fan -------------------- 000w
Black Ice GT Stealth 240 Dual 120mm fans --------------------- 000w
Black Ice GT Stealth 360 Triple 120mm fans -------------------- 000w
Black Ice GT Stealth 120 X-Flow Single 120mm fan ------------- 000w
Black Ice GT Stealth 240 X-Flow Dual 120mm fans -------------- 000w
Black Ice GT Stealth 360 X-Flow Triple 120mm fans ------------- 000w
Black Ice Pro Single 120mm fan -------------------------------- 000w
Black Ice Pro II Dual 120mm fans ------------------------------- 000w
Black Ice Pro III Triple 120mm fans ----------------------------- 000w
Black Ice Xtreme Single 120mm fan ---------------------------- 000w
Black Ice Xtreme 2 Dual 120mm fans --------------------------- 000w
Black Ice Xtreme X-Flow Single 120mm fan --------------------- 000w
Black Ice Xtreme 2 X-Flow Dual 120mm fans -------------------- 000w
Black Ice Xtreme 3 X-Flow Triple 120mm fans ------------------- 000w
Black Ice GTX Gen Two Xtreme 120 Single 120mm fan ----------- 000w
Black Ice GTX Gen Two Xtreme 240 Dual 120mm fans ------------ 000w
Black Ice GTX Gen Two Xtreme 360 Triple 120mm fans ----------- 000w
Black Ice GTX Gen Two Xtreme 480 Quad 120mm fans ----------- 000w

Koolance HX-360 Single 120mm fan ----------------------------- 000w
Koolance HX-362 Single 120mm fan ----------------------------- 000w
Koolance HX-420 Single 120mm fan ----------------------------- 000w
Koolance HX-720 Dual 120mm fan ------------------------------- 000w

MagiCool Xtreme Nova 1080 (9 X 120mm fans) ------------------- 000w
MagiCool Xtreme Quad 480 (4 X 120mm fans) -------------------- 000w
MagiCool Xtreme Slim Profile Single 120mm fan ------------------- 000w
MagiCool Xtreme Slim Profile Dual 120mm fans -------------------- 000w
MagiCool Xtreme Slim Profile Triple 120mm fans ------------------- 000w
MagiCool Pro Slim Profile Single 120mm fan ----------------------- 000w
MagiCool Pro Slim Profile Dual 120mm fans ----------------------- 000w
MagiCool Pro Slim Profile Triple 120mm fans ---------------------- 000w
MagiCool 1 X 120mm MC-120UVB Single 120mm fan -------------- 000w
MagiCool 2 X 120mm MC-240UVB Dual 120mm fans --------------- 000w
MagiCool 2 X 120mm MC-360UVB Triple 120mm fans -------------- 000w

Alphacool NexXxoS HC 240-LE Dual 120mm fans ------------------ 000w
Alphacool NexXxoS HC 360-LE Triple 120mm fans ----------------- 000w

Evercool WC-R12 Single 120mm fan ----------------------------- 000w

Glacial-EXM Dual Pass Single 120mm fan -------------------------- 000w
Glacial-S Single 120mm fan -------------------------------------- 000w

Thermaltake Aqua TMG1 Single 120mm fan ------------------------ 000w
Thermaltake Aqua TMG2 Dual 120mm fans ------------------------- 000w
Thermaltake AquaBay M2 Single 120mm fan ----------------------- 000w

[Burn]

I can tell this is going to be a great thread :D I am especially interested in the GT Stealth 360 and the MCR320, the former I own and the latter I'm interested in :D

Anyway, I'd contact MartinM, he has a ton of information regarding radiator and pressure drop testing which I think might factor into maximum thermal dissipation potential (what you're after).

Good luck and I hope this thread becomes a really handy reference :)

[Mekrel]

I don't believe this is possible, as you would need the performance curves for each radiator and they would need to come off the exact same test bed.

[Talonman]

What if we said no blocks, or all rads would be getting 6lpm of flow....

I would like to take the blocks affecting the flow totally out of the equasion.

I just would love to know the max cooling ability of each rad.

Again, even a close rateing would help with buying decissions. :)

[Clockwork_Apple]

What's the difference between the MCR-QP and MCR-QP-K?

[Talonman]

Not sure, but they have them listed as different items on the sites selling them. :)

The QP Single is $43.95, http://www.performance-pcs.com/catalog/index.php?main_page=product_info&cPath=200&products_id=22813

And the QP-K Single is $34.95, http://www.performance-pcs.com/catalog/index.php?main_page=product_info&cPath=200&products_id=3495

I'm guessing the QP is better?

[mak1skav]

Not sure, but they have them listed as different items on the sites selling them. :)

The QP Single is $43.95, http://www.performance-pcs.com/catalog/index.php?main_page=product_info&cPath=200&products_id=22813

And the QP-K Single is $34.95, http://www.performance-pcs.com/catalog/index.php?main_page=product_info&cPath=200&products_id=3495

I'm guessing the QP is better?

The first one that costs $43.95 it has an integrated reservoir.

[Martinm210]

I would love to know this too..:up:

Like you already said, we'd just need to set a fixed flow rate and also a fixed water to ambient delta.

I think it's pretty standard to use a 10C water to ambient delta for heat dissipation curves, so that's a give me.

Flow rate average is probably around 1.5 GPM.

Fans as you mentioned is the 70.5cfm yate loons.


Just for the heck of it, lets see if I can extract anything from my own persnal results:Unfortunately I only have figures for the slow speed 47cfm types in my own system.

I see 4.0C delta water to ambient for about 125 watts(per the CPU calculator and an assumed 16 watts for pump heat dump) on an MCR320

If I solve that (assuming a linear relationship) and 10C, that would be 312 watts.

Now if I assume a linear relationship between CFM (which I know is wrong) that would be 312 x 70.5/47 cfm = 468 watts.

But, I think that is off somewhat.:down: The PA120.3 is a better radiator with more surface area, better pressure drop, and possibly better air flow pressure drop. So either my temperatures are slightly off or my heat dissipated is slightly off. Also saying a fan with 70cfm is 70/47 times better is incorrect, fans work just like pumps with a pressure/flow type of curve and pressure is equally as important especially with dense radiators. Anyhow it's tough trying to get a good "Watts dissipated" from anything other than using the exact same fans and having really good equipment to measure accurate temperature differences.


But, swiftech does provide heat dissipated curves for the MCR320, just need to look it up here..:up:

This is with delta fans that are 75 cfm, so it's might be a little off with the two different fan types and pressure differences, but it should be in the ballpark)
http://www.swiftech.com/assets/images/products/MCR320/mcr320_HD_vs_FR.gif

MCR220:
http://www.swiftech.com/assets/images/products/MCRad-220/MCR220QP_HD_VS_FR.PNG

MCR120:
http://www.swiftech.com/assets/images/products/MCR120-QP/MCR120QP_HD_VS_FR.PNG

With 75cfm delta fans:
MCR320 would be "about" 430 watts.
MCR220 would be "about" 320 watts.
MCR120 would be "about" 170 watts.

The trouble is there really isn't much information out there for the other radiators other than swiftechs and thermochills, although I heard HWlabs may be doing some updates to there site soon with more data like this.:up:

[Talonman]

Thanks Martin, Your a scholar and a gentleman.

Origional post updated. :up:

Do you think the QP-K's would also be rated about the same?

[Martinm210]

From Thermochills graphs

Using the Delta 75cfm fans at 12V:
PA120.1
http://www.thermochill.com/PATesting/PA1201cwVSFlowrateGPM.jpg
PA160
http://www.thermochill.com/PATesting/PA160HeatdissVSFlowrateGPM.jpg
PA120.2
http://www.thermochill.com/PATesting/PA1202HeatdissVSFlowrateGPM.jpg
PA120.3
http://www.thermochill.com/PATesting/PA1203HeatdissVSFlowrateGPM.jpg

Results with Delta 1212M at 12V (75cfm), 1.5 GPM:
PA120.1 = 175 watts
PA160 = 220 watts
PA120.2 = 345 watts
PA120.3 = 510 watts

[Martinm210]

Thanks Martin, Your a scholar and a gentleman.

Origional post updated. :up:

Do you think the QP-K's would also be rated about the same?

If the surface area is the same, it's probably very similar would be my guess.

[Talonman]

Thanks again buddy!! :)

[Origin_Unknown]

awesome thread idea...

but i thought the pa120.3 was capable of 700w disapation.... maybe i drempt it?

[Talonman]

Thanks .....

I have never heard it rated that high. :p:

[Isriam]

my pa120.2 has a hard time handling 200 watts.

i'd say use 40 cfm fans (middle ground) that way people could expect more or less depending on what fans they used.

[Xilikon]

awesome thread idea...

but i thought the pa120.3 was capable of 700w disapation.... maybe i drempt it?

Actually, it's more like 600w top but with fans over 100 cfm only.

[Talonman]

my pa120.2 has a hard time handling 200 watts.

i'd say use 40 cfm fans (middle ground) that way people could expect more or less depending on what fans they used.

What block(s) are you using in your loop, that is accounting for your 200watts, and what is it cooling? A Q6600 makes around 200w to 250w of heat alone.

This is where the various pumps and fans could also come into play.

If 40cfm fans are the (middle ground), what is the CFM rating of the low speed fans they are using with their rads?

[DarthBeavis]

I wish someone would make a calculator like the PSU calculator for watercooling . . .I am just not smart enough to figure this stuff out.

[Martinm210]

awesome thread idea...

but i thought the pa120.3 was capable of 700w disapation.... maybe i drempt it?

That's the scary part of talking watts. We should emphasis the "10C water delta", "fans" used, and flow rate (although flow rate has minimal change).

The PA 120.3 is capable of dissipating 700watts, it just depends on the fans and water to air delta in temperature. The 490 watts is with 70cfm yate loons(10C delta, 1.5GPM), and the 510watts is with 75cfm Deltas(10C delta, 1.5 GPM).

With a 40C delta and a 220cfm delta (http://www.sidewindercomputers.com/de12gf.html), we could probably make an MCR120 a 2000 watt radiator..:ROTF:

I'm a silence is bliss old fart I guess, I like my 47cfm yates..:D

[MaxxxRacer]

That's the scary part of talking watts. We should emphasis the "10C water delta", "fans" used, and flow rate (although flow rate has minimal change).

The PA 120.3 is capable of dissipating 700watts, it just depends on the fans and water to air delta in temperature. The 490 watts is with 70cfm yate loons(10C delta, 1.5GPM), and the 510watts is with 75cfm Deltas(10C delta, 1.5 GPM).

With a 40C delta and a 220cfm delta (http://www.sidewindercomputers.com/de12gf.html), we could probably make an MCR120 a 2000 watt radiator..:ROTF:

I'm a silence is bliss old fart I guess, I like my 47cfm yates..:D

And because of this, I never really persued the Idea of trying to make a chart like this and is why I never state Wattage number for any piece of cooling gear becasue as you mentioned you can turn a MCR120 into a 2000watt rad with enough heat.

Sadly, and understandably, many users feel that reccomendations are not enough and they need to see hard numbers. Hopefully those same people will understand that these numbers need to be taken into context.

With that said Talonman should probably add (preferably in large letters) the conditions at which the rads were tested and a disclaimer stating that testng the rads under different conditions will yield different wattage numbers.

[Talonman]

That is why I stated:

Suppose we used a D5 pump, and Yate Loons rated at 70.5cfm, could we figure the Thermal Dissipation for all the rads listed? If so it might be handy for a quick reference chart, to help aid in buying decissions.

RAD USED________________________________ Approximated Thermal Dissipation.


I understand there are lot's of variables that can change the performance of a rad, but reject the idea that performance can't be charted, when just looking for a ballpark figure.
A similar chart worked well for us to help determine the Loaded NB temp on the stock Maximus SE.

I suppose we could ask for a rad recomendation each time, but I favor a chart myself.

Keep the data flowing in. I think some will use the chart if we ever get it fully populated with our combined best guess. It would have the potential to keep a few bad purchases from being made. :)

[Isriam]

the problem is, you're going to have people going "this rad says it will support 500 watts and it isn't cooling my cpu and vidx3" with stats like this. i find it hard to believe a pa120.2 can cool 350 watts in any configuration.

[Talonman]

I still wanted to know how you came to your 200w figure, that your pa120.2 is having a hard time cooling.

What block(s) are you using in your loop, that is accounting for your 200watts, and what is it cooling? A Q6600 makes around 200w to 250w of heat alone.

This is where the various pumps and fans could also come into play.

If 40cfm fans are the (middle ground), what is the CFM rating of the low speed fans they are using with their rads?

I bet Marci could confirm our approximated data on the ThermoChill rads, using the fans I specified?

He has never had a problem doing that.

Here are some if my favorite posts made by, him that are related to rad performance:

YateLoon D12SM-12 = 70.5cfm @ $5.95 (http://www.petrastechshop.com/12yalod1cafa1.html), and used with a PA120.3 and liq flow around 6lpm, would give approx 490w of Thermal Dissipation.
My favorite post!!


With what fans??
That's the difference between them. Core is thicker on the GTX (dual row) thus requires higher pressure fans to penetrate it adequately.

Pro and GTS = lower CFM fans
Extreme and GTX = upper CFM fans

A GTX vs a GTS with lowspeed fans up to a specific heatload, the GTS will win, but it's margin over the GTX will become smaller as you approach that specific heatload, at which point, the GTX will win as it has slightly higher thermal capacity. With Highspeed fans, the GTX will win more or less all the time.

GT and GTS version - these are both the same thing. GT Stealth = GTS(tealth)
The GTX is better than a GT or GTS with highspeed fans.



Not really, but there is a way to work it out (in a VERY rough fashion).

1 - Working out the heatload - ie: how many watts of heat you need to be able to cool.
Max Power Consumption = Max Heat Output. Go here: http://www.extreme.outervision.com/psucalculatorlite.jsp

Fill in only the CPU and GPU fields (ie: the fields for those items upon which you intend to mount a waterblock), and the pump field (as your pump dumps heat into the system too). Hit calculate, and knock 38w off the result. What you have left is your heatload. If you want to include chipset cooling, just add another 100w for sake of argument and assume you'll be left with headroom. I recommend googling to find out what kind of overclock folks get with similar stepping CPUs and at what vCore, and completing the "overclock my cpu" field with the relevant details, as heatload gets much higher once you start overclocking it.

2 - Work out system flowrate
Use the Flowrate Estimator (http://www.xtremesystems.org/forums/showthread.php?t=151627) to work out what your total liquid flowrate will be based on your selection of blocks etc etc. That'll give you an answer in gallons per minute.

3 - Work out roughly what CFM you need combined with liquid flowrate to cool the radiator adequately.
http://www.thermochill.com/PATesting/PA1203HeatdissVSFlowrateGPM.jpg

Take the answer from 1), and plot that on the vertical axis. Take the answer from 2) and plot that on the horizontal. Where the two cross, look for the nearest colored line. Each line represents one of 3 fans at either 12v or 7v.

Pf H = Panaflo FBA12G12-H1A - 105cfm, 6.8mmH2O, 41.5dBA @ 12v (Factory spec)
1212M = Delta WFB1212M - 72.4cfm, 3.4mmH2O, 34dBA @ 12v (Factory spec)
Nexus = Nexus D12SL-12 - 36.8cfm, 22.8dBA @12v (Factory spec)

You should be able to guage approximately what CFM-per-fan you need to be moving thru the radiator from some fudgy guesswork and approximations by dividing the space between the lines into equal increments etc.

4) Choose your fans
Once you know roughly what kind of CFM per fan you want to be using, you now need to find a fan that will produce that amount within your noise tolerance. Refer to the following data...

http://www.employees.org/~slf/curves/noise/flow-noise-amb.png

Look along the horizontal axis for the CFM you need. Go up til you hit the first line. This is the quietest fan for the job. Now find that fan in the links below... and look at the green line on it's graph. Find your CFM on the horizontal axis, go straight up til you hit the green line. Now go straight across to the right. That's the rpm the fan needs to go at. Now go left from the rpm til you hit the BLUE line, and go straight up. That'll tell you what voltage you need to run that fan on to achieve that rpm, and thus achieve the desired airflow at the lowest noiselevel possible.

Individual fan cfm vs noise vs rpm vs voltage graphs:
Acoustifan AFDP-12025 [Dustproof] (http://www.employees.org/~slf/curves/noise/acoustifan-afdp12025.png) ArcticCooling AF12025PWM (http://www.employees.org/~slf/curves/noise/arcticcooling-af12025pwm.png) CoolerMaster AAF-B12-E1 (http://www.employees.org/~slf/curves/noise/coolermaster-aafb12e1.png) Delta TFB-1212GHE (http://www.employees.org/~slf/curves/noise/delta-tfb1212ghe.png) EBMPapst 4312L (http://www.employees.org/~slf/curves/noise/ebmpapst-4312l.png) EBMPapst 4412FGML (http://www.employees.org/~slf/curves/noise/ebmpapst-4412fgml.png) Frankenfan - Tricod Loon (http://www.employees.org/~slf/curves/noise/tricodloon-d12sl12-c.png) Panaflo FBA12G12H (http://www.employees.org/~slf/curves/noise/panaflo-fba12g12h.png) Scythe S-Flex SFF21F (http://www.employees.org/~slf/curves/noise/scythe-sflex-sff21f.png) Sharkoon SilentEagle 2000 [Golfball] (http://www.employees.org/~slf/curves/noise/sharkoon-se2000.png) ThermalTake R121225BUT [Smart Case Fan II] (http://www.employees.org/~slf/curves/noise/thermaltake-r121225but.png) Tricod Science SPDL1225S (http://www.employees.org/~slf/curves/noise/tricodscience-spdl1225s.png) Yate Loon D12SL-12 [curved/Nexus blade design] (http://www.employees.org/~slf/curves/noise/yateloon-d12sl12-c.png) Zalman ZF1225ASH [ZM-F3] (http://www.employees.org/~slf/curves/noise/zalman-zf1225ash.png)

Source: http://www.xtremesystems.org/forums/showthread.php?t=150261


Doneskis. Your system should now be capable of returning coolant to [ambient + 10degC] (ish)

If this results in your only choice being fans you consider to be excessively noisy, you can recalculate it settling for a final coolant temp of [ambient + 15degC] (so roughly 5 deg C higher temps all round basically - you always sacrifice temperature for noise and vice-versa... low temp = high noise, high temp = low noise) by substituting the PA120.3 dissipation graph above with the one below... but you'll need to use the "Liquid Flow Equivalents (http://www.deltainstrumentation.com/calcs.html)" converter to convert the answer from 2) into lpm, as I can't be arsed to translate the graph from lpm into gpm (sorry!!).

http://www.thermochill.com/PATesting/dT15/pa1203.jpg

I think by taking out all blocks, the Thermal Dissipation should be easer to calculate....



Get your facts straight before you start throwing such weighty words about please...

1 - The charts for BOTH products were produced by the SAME independant tester, Bill Adams... who undertakes independant testing for a fee. See www.thermal-management-testing.com - The Swiftech data is not represented on that site as the testing on their rads was undertaken whilst he was in employment by Swiftech (he also designed their MCR series rads). Since he was under employment by them, Swiftech own all the data produced so he cannot legally reproduce it on his own site which covers all the work he's done independantly to Swiftech (also note, Bill also worked on the Coolingworks radiator design)...

2 - The data used for those charts was produced by two different test rigs at two different points in time... as the tester who produced the data has stated, these two charts are NOT directly comparable due to differences in the testing rigs. The tester (BillA) retested the MCR220 vs the PA120.2 for scaling of results... see HERE (http://www.thermochill.com/PATesting/ALLcwVSFlowrate.jpg) and the quote below.
Long n' short, there are no charts you can use to compare the radiators in question due to testrig differences... and the only testing done of the two against each other in equal scenarios was by Radical_53. Whether this can be determined as conclusive evidence is entirely up to the end-user.

In some scenarios (very low airflow - sub Nexus@12v) the PA120.2 beats the MCR320. In other scenarios (higher airflow - Delta1212 range and upwards) the MCR320 beats the PA120.2 - so, the question can only be answered without any fanboyism influencing the decision by providing us with the following info: What fans is one intending to use on the radiator? and What heatload do you intend to cool?
Between the MCR320 and PA120.2, it appears to be a close race?


Long n' short of it - the lesson to learn when testing large format radiators (Triple or Quadruple Fan Footprint) is that you must be able to produce a seriously large heatload for any differences between the two to become distinguishable... If the heatload is below the capacity of the radiators being tested, then both radiators will appear to perform very much the same, and any advantages of one over the other will be very difficult to detect. The miniscule differences in water temp vs the larger difference in temp at the CPU highlight this - a miniscule change in the temps of variables at the radiator can have a MUCH larger (relatively) affect on things elsewhere in the system - hence when Bill does radiator testing, he uses thermal probes to measure the coolant temps accurate to 6 decimal places. Flowrates measured in GPM to 3 decimal places.

And hence radiator testing is a bit of a minefield for anyone to step into... :D
I believe that.... and think that is why we should just keep it to approximations only.


PA120.3 vs GTX480 with 50cfm fans, PA120.3 wins. With 70cfm fans, PA120.3 wins. Soon as you move to 80CFM fans on both rads, the GTX480 pulls into the lead ever so slightly (less than a degree at 80cfm, and at 100cfm GTX probably wins by 2 or 3 degrees) but you also have the increased space it takes up for those 2 degrees, so there are tradeoffs. But, BOTH rads have the same fans on, and thus produce the same noise. OBVIOUSLY any rad of same footprint or larger, with 100cfm fans, WILL outperform any other rad with 50cfm fans. It is the amount of airflow that determines how much heat the radiator can remove (assuming flowrates are constant)...
Again, the importance of specifying what fans you are using.

[billb]

So you go through this lengthy post stating that there many variables that must be controlled (in order to get meaningful results) and then want to generate approximations? I won't say "It aint gonna happen", but garbage in garbage out!

Oh-where-oh-where is Bill A. .............rolling over in his beer soaked coffin?

[Talonman]

I wan't exact numbers if I could get them.....

Anything, even approximations would be better than nothing...

Kind of like your post billb. I had hopes it actually furthered our cause when I saw activity in this thread.

Turned out to be just a troll post.

sad!

[MaxxxRacer]

So you go through this lengthy post stating that there many variables that must be controlled (in order to get meaningful results) and then want to generate approximations? I won't say "It aint gonna happen", but garbage in garbage out!

Oh-where-oh-where is Bill A. .............rolling over in his beer soaked coffin?

lol Mr Adams.

[eXa]

Hmm this would be hard work...

I wonder.. could u make an crude test bed wich roughly finds the wattage a radiator can dissipate?

Like. take 2 peltiers, a adjustable psu, 2 waterblocks, voltage\amp meter (to kontroll the peltier power usage) something to monitor temps here and there, D5 and a bunch of rads to test.

Then u tune in by adjusting the voltage to the peltiers to the point where the rads keep the water at an steady 10c above ambient(or whatever delta u wish to test at).
Then u can just look at the voltage and amp drawn to se aproximity how much wattage the rad dissipates
Only thing im not shure of is what to do with the cold side? i gues it needs to be kept as close to ambient as possible? can be abit hard...

This would ofcourse not be exact science but would give u a rough idea. And fairly easy test to perform once the test rig is set up.

[Talonman]

I would love to see that done.... :up:

There has got to be a way to rate these things.

[Xilikon]

I think it's not necessary to be a peltier but just a heating element. just feed 120v constant (or monitor the actual voltage if you want to be that accurate) and adjust the amperage till you reach a delta of 10C. Read the amperage required to attain this and do the math (V x A = W) to get a accurate heatload.

I said it's better to use a heating element because the power loss would be 100% heat so you can correlate the power feed to the heat dissipation. However, to be efficient, we need to build a custom heating element with 2 barbs so water can circulate on it in a behavior similar to a normal waterblock but it's just a theorical guess and I'm sure someone well versed in thermodynamics and fluid dynamics can find how.

[Martinm210]

I think it's not necessary to be a peltier but just a heating element. just feed 120v constant (or monitor the actual voltage if you want to be that accurate) and adjust the amperage till you reach a delta of 10C. Read the amperage required to attain this and do the math (V x A = W) to get a accurate heatload.

I said it's better to use a heating element because the power loss would be 100% heat so you can correlate the power feed to the heat dissipation. However, to be efficient, we need to build a custom heating element with 2 barbs so water can circulate on it in a behavior similar to a normal waterblock but it's just a theorical guess and I'm sure someone well versed in thermodynamics and fluid dynamics can find how.

I've been thinking about trying exactly that. But using aquarium heater elements to do the heating.

It would be really easy to stick 3 or 4 down my test reservoir and switch on several in different steps to get a good comparison. Perhaps in 100watt heater elements, or a larger one with a variac to control voltage.

It would provide a great basis of comparison, the trickier part is trying to determine the actual wattage dissipated by radiator. The tricky part is measuring the coolant temperature in and temperature out accurately enough.

For light loads, it could be less than one degree, so your typical thermal probe that is only accurate to within 1 degree just isn't going to be good enough. You could just record air in vs the stabilized coolant temperature and at least compare those. You could also use the variac and a larger heater to adjust the voltage until 10C delta was reached, but I would guess that's a bit tricky. A good variac is also pretty expensive.

Regardless, I plan to tinker around with this after I get a few other projects out of the way..


FYI,
These guys go over the right way to do radiator testing, of coarse that's some nicer than I'll ever have equipment:
http://translate.google.com/translate?u=http%3A%2F%2Fwww.cooling-masters.com%2Farticles-38-0.html&langpair=fr%7Cen&hl=en&safe=off&ie=UTF-8&oe=UTF-8&prev=%2Flanguage_tools


Q = M x x dT Cp (Equation 1)

Q = power dissipated by the radiator in W (calculate)
M = mass flow of water kg / s (measured)
Cp = heat capacity of water J / kg K, it is 4186 J / kg K here (known)
DT = temperature difference between the entrance and exit of the radiator for Water in K (1 K = 1 ° C) (measured)


You always dissipate whatever has gone in:
http://www.cooling-masters.com/images/articles/gts240/images/evo.png

It's just that the better radiator will keep the coolant temperature lower, that's why they measure directly at the inlet and outlet of the radiator, because it eliminates any losses in tubing or other areas.

Still it would be fun to play with even if just measuring the coolant temperature delta from ambient, it would be a useful comparison with other rads tested the same way.

[eXa]

I just used peltiers cus it was the only thing i could think of where u pretty accurately could adjust the wattage...

But yeah a heater element would be better since u dont have to deal with the coild side thingy and u could have the element directly in the water.

[Xilikon]

It's not a question of life or death over a difference of 5W so we can afford to have something accurate to within ±10-15W since this often translate to less than 0.2C discrepancy. What matter here is to know if radiator X can dissipate Y watts and radiator Z can dissipate W watts to help pick the right one for our needs. It will also help us separate the best from the rest (If radiator X dissipate Y watts and cost 40$ and radiator Z dissipate more and cost 42$, you can work a W per $ ratio).

Another tricky part would be to always stick to the same selection of fans (no rheobus, only straight 12v or straight 7v).

[Xilikon]

I also think one of the tricky parts will be to get all those rad in at one time and one place for testing:)

Just a thought,

andyc

Tag with Alex and this might be solved :D

[Xilikon]

I know, I was just kidding :p:

Actually, the best way is to work on a good testing setup and put the method in writing. Anyone who want to provide results on radiators not tested can do it with the same methodology. The tricky part is to make sure the parts you use for testing can be found or build very easily worldwide.

[Talonman]

Such fine minds trying to get us valuable info to help us better plan our loops just gives me goosebumps. Good job boys!! I love team players. :up:

[billb]

Turned out to be just a troll post.

sad!Don't get me wrong, I'm on your side, probably more so than you can imagine*. It's just that the people that will understand what a D5 pump, and Yate Loons rated at 70.5cfm, ThermoChill PA120.2 Dual 120mm fans - 345w means, probably already have the WCing smarts and math skills to generate the data they need, and the nOObs will only be confused at best, or end up with totally screwy systems at worst.

Here's some data on some more rads:
http://translate.google.com/translate?u=http%3A%2F%2Fwww.effizienzgurus.de%2Fm ain%2Findex.php%3Fcontent%3Darticle%26action%3Dvie w_spec_article%26article_id%3D222&langpair=de%7Cen&hl=en&ie=UTF-8
Alpha Cool NexXxos Extreme III Rev.2
Hardware Labs Black Ice GT Stealth 360 chrome Hardware Labs Black Ice GT 360 Stealth chrome
Hardware Labs Black Ice GT Xtreme 360 Hardware Labs Black Ice Xtreme GT 360
Hardware Labs Black Ice GT Xtreme 360 Lite Hardware Labs Black Ice GT Xtreme Lite 360
MagiCool Slim Triple Slim Triple MagiCool
ThermoChill PA 120.3 Thermo Chill PA 120.3
Watercool HTF3 Triple X Cool Water HTF3 Triple X

But, from your data:

RAD USED________________________________ Approximated Thermal Dissipation.
ThermoChill PA120.1 Single 120mm fan ------------------------- 175w

What would be the flow rate be? And even if, by some pulling a rabbit out of a hat trick, I could determine that, what if I wanted to substitute in a more/less restrictive block into my system. How would that effect (hard data please) the results I potentially would see?

Hmmm, flow rate goes down, but block picks up more heat ...does the air water DeltaT go up or go down? Hmmmm, can I substitute quieter (lower CFM) fans and still get the same C/W ratio?

Basically, what you're giving people is the attached chart.

* "I'm new to WCing, I wanna WC every thing in my case including the PSU, memory(W/LEDs), and I wanna add flow meters, have three pumps and 2 Rads, What parts should I buy? And Where can I find a Rez W/LEDs?" ...all with no mention of what CPU, GPU or overclocking!!!!!!!! How can this person be helped? With a baseball bat? Sent back to school to redo third grade? ...maybe a few days of waterboarding!

[billb]

lol Mr Adams.Yea, been "accused" more than a few times of being a pseudonym for billa ..now that's a LOL

One of his last public utterances...a classic

but really bs pr is all that the enthusiast community requires as their purchases are based on anything other than FACTS (as there are almost none), opinion is sufficient - after all this is a hobby, not science


Emphasis added........

[Talonman]

Thanks for the post Bill...

I just am sensitive to when we get a thread started, some people who don't see the value, have to post negitive things. It would be better for them to just pass it by and let the members who would do see the value, let it run it's cource to completion. The nay sayers just might learn a thing or two and be suprised! I am glad you are on board with the Rad Rateing thread, and see it's value too. I can see you know your stuff. You might be able to help us get even more accurate numbers.... :up:

Thanks for the post.

All you need to do is add togeather how many watts of heat your system is generating, baised on what exactly you are cooling in your loop. Then buy the correct RAD that can deal with it.

Check out Martins flow rate estimator, some of your questions might be answered. Very little difference in RAD's cooling ability, after you get above a decent flow, that a D5 pump can normally generate. Yes if you overload your loop, you could drop below that number, but most won't.

If we don't know how the RAD's are rated, you can't make a wise decission on your RAD selection.

[Martinm210]

I think the fact that we can't even begin to fill out 80% of the list a good indication how little scientific data we have to work with on radiators, and unfortunately we are buying on opinions and biased recommendations.

I can only assume if the data isn't available, that it's not good. I'd suggest a big 0 next to the blanks...:D

[Talonman]

Don't get me wrong, I'm on your side, probably more so than you can imagine*.

* "I'm new to WCing, I wanna WC every thing in my case including the PSU, memory(W/LEDs), and I wanna add flow meters, have three pumps and 2 Rads, What parts should I buy? And Where can I find a Rez W/LEDs?" ...all with no mention of what CPU, GPU or overclocking!!!!!!!! How can this person be helped? With a baseball bat? Sent back to school to redo third grade? ...maybe a few days of waterboarding!

This would be my advice to this hypothecial user:

PSU = Dont get a water cooled PSU. There is no advantage to this, and it will just dump extra heat in your loop, and decress your flow.

Memory: Probably a bad move too, but if you must add in about 30 watts of heat.

Flow meters don't add heat, just reduce your flow. If you must add only one, not multiple.

You must give us the CPU you are running, but if it's a Q6600, add in 200 to 250 Watts.

We would need to know What GPU you intend to cool, but it will be less than 200 Watts of heat if a single.

So far we are up to 430 watts...

So far 1 PA120.3 will do it!!

[billb]

I just am sensitive to when we get a thread started, some people who don't see the value, have to post negitive things.

...acerbic and churlish I am...........

All you need to do is add togeather how many watts of heat your system is generating, baised on what exactly you are cooling in your loop. Then buy the correct RAD that can deal with it.

...how else would you do it? Ouija board?

''Check out Martains flow rate estimator,

...had my own four years ago.....


If we don't know how the RAD's are rated, you can't make a wise decission on your RAD selection. My point is that what you're giving people is one point on a graph. Yea, if all the variables are controlled, that point can be compared to others, if not controlled, the data is up for grabs. I'm just not sure if, in this case, "some data" is actually better than no data.

< a long pause >

... well, upon reflection, I'd have to say some dsta is better than none if you understand the limitations of the data. If you don't ...buy some LEDs and be happy! Like you say, I'm in!

Black Ice Pro Single
http://www.overclockers.com/articles1140/
Black Ice Pro III Triple
http://www.overclockers.com/articles1163/

[Xilikon]

Bill has a point. While the data can be used as a reference to narrow the choices, you cannot provide a exact answer because the performance depend on flow.

However, if we can outline the same parameters (like telling which fan and which flow rate), it can be valid enough to make choice since the graphs isn't in a bell shape but a light curve with a sharp change under 1 GPM.

[Martinm210]

I agree, some information is better than no information. We just need to emphasize the variables.

I would recommend:

Assume: 1.5GPM (This would probably be close to what most systems run, and differences result in very little difference in radiator thermal performances).

Assume: 10C Water to Ambient Delta - This is pretty standard, I've seen a few cases of 15C for the 80mm rads, but 10C is pretty common.

Assume: Nexxus Fans at 12V - While we probably see more people using Yate Loon D12SL12s, most of the available thermal curves for radiators have used the Nexus fans at 12V, yate loon data just is not available.

With those three assumptions, we can pick a point. I really don't think flow rate makes enough difference, but we should emphasize the difference in fans. Using Panaflo 105cfm fans at 12V will litterally double the performance compared to some slower speed nexus fans. I think that variable is extremely important to understand.

It would probably also be helpful to explain what the 10C delta means. This is just how much hotter the coolant will be than the ambient temperature. People need to understand it's just a number, and if you want better cooling, this is one of the most important areas you can improve upon. It's not that hard to keep water/ambient deltas less than 5C if you have plenty of radiator capacity.

We have a wide audience here, I'm all for some good rules of thumb as long as we spell out some of the assumptions. In the end, it will introduce more good than harm. The XS way is to show me the numbers, so lets do it! Everything in life is neither black or white, but if you can't move on with some assumptions and risk, you'll be spinning your wheels on the details forever.

Some data is better than no data..:D

[Talonman]

I am on board with that... :up:

OP updated.

[Martinm210]

I am on board with that... :up:

OP updated.

I like it, particularly the big 000w! Maybe we can encourage a few more manufacturers to "PROVIDE SCIENTIFIC DATA!":D ..:up:

[Talonman]

Thanks!

We will keep the heat on them, and see how they disapate it...:rofl:

[SottaM]

Could anyone tell me about the heat dissipation on a Thermaltake Aqua TMG2 (Plus TT P500 Pump+reservoir)? I just bought a TT Armor LCS and it came together. Im using the "stock" settings, i mean i just read the manual and followed the steps. Im just a beginner on watercooling :(. Anyway i did the leak test on the setup and it passed without any problems :p:.

My loop is only for the CPU, it's a X2 3800+ (939) @ 2.750 and 1.41v (the lad hangs on 1.45vcore, bad cpu) on the vcore.
sig updated! :ROTF:

---
Another question regarding future upgrades: :yepp:

Ahh i forgot to mention, im thinking on a upgrade to my wc system because i'm going to build a new rig in next year. I was impressed with MCP655 by Swifttech GPH/LPH rating, so im thinking about to use 2xMCP655 (cpu/gpu, and what radiators/reservoirs for MAXIMUMMM (yeah to teh MAX) dissipation?) to this loop: going to vmod my 88GT to run it like 800/19xx/1xxx on the clocks at least, buy another one to SLI'them (with the same vmod) and maybe a Lynnsfield on the loop too, heavily overclocked. :up::up::up:

aaaaah, should i make a new topic regarding my last question?

[orclev]

Would also be nice to get some data on the new Feser X-Changer 120, 240, and 360 radiators. If you believe the hype on performance-pcs.com, they perform better than the PA120.X radiators, but without any hard numbers it's all hearsay.

Eventually it would be cool if we could run test on each of the variables for the various radiators and come up with something like Martins (most excellent) flow estimator. I'm picturing a spreadsheet you punch your thermal load in on (or delta to ambient I guess), your CFM, and possibly your GPM (although that's not much of a factor), and then based on the tables of data for various models it could generate a graph. Unfortunately as has been pointed out several times already we have all of about 1/100th of one of those tables, so there's a very long way to go will we reach that point.

For what it's worth I'm thinking I'm going to pick up a X-Changer 360, an X-Changer 120, and possibly and X-Changer 240 for my next build, so if I can swing a set of acceptable fans and cobble together some sort of heating element/measurement system I could try to get a bit of experimental data. It would be very limited based on exactly the CFM fans I'd have on hand, but it's better than nothing.

[RCG_Bex]

What about the new TFC rads? We not going to include them?

I know the quads beat the GTX rads when it comes to low CFM fans (<1800RPM) and seems to do the same against the Thermochill PA but I've not seen a full comparision...

~Bex

[Martinm210]

I would just be hesitant to include heat dissipated numbers in the same table tested by two different test beds. My own testing for example is on an open air system without any insulating of tubing or the reservoir, or the rest of the system. So in the end, I am dissipating some heat through the tubing and such.

I consider that ok because you'll do the same in any real world watercooling system or similar anyhow. The problem is mixing numbers from different sources, we just don't know where they came from and performance at times is so close you just can't really compare unless they are tested on the same setup.

I'm already running out of steam after only a handfull of radiator tests, so I don't think I'm willing to do a whole bunch more myself. Unfortunately it just takes way too much time. I probably spend 20hours on one radiator, so I could easily buy a half a dozen radiators for the time spent vs the one sample aquired.

In addition regarding the heat dissipated table, I've already found several radiators to be optimal for one fan over another(One prefers and is better with high speed fans and the other is better with the low speed fans), so to be fair you'd have to build several heat dissipated tables for several different fan scenarios. It all gets very complex different depending on the fan used.

Your best bet is to compare by characteristics. The single most important one is the frontal surface area, how much area does the initial cold air hit when entering the radiator...Bigger is BETTER. Second is probably the fin density and you want more open for slow speed fans and more dense for high speed fans. Third with be the thickness where the double thickness rads edge out a little bit more from the air particularly at higher fan performances. You also get a little more efficiency from thinner flow tubes or fewer of them at the cost of higher restriction. That'll get you in the ballpark anyhow.