Newer Test Results
 

I think some of you have seen these by now, but I figured I'd go ahead and post them for everyone. I am waiting on mfgr comments to finish the reviews; Joe said he'd post them on Sunday when he got back in town from work. Anyway here are the results for the Dangerden Maze4 and the Swiftech MCW5000-A:

http://phaestus.procooling.com/temp/wbtests_s.jpg

Earlier Swiftech results were using older springs (presumably weaker). With the newest springs the MCW5000-A looks pretty good at lower flow rates (such as you'd see with multiple blocks or that 3/8" tubing that Swiftech likes). Their strategy for cooling system design makes a lot more sense to me now.

The Maze4 is less restrictive than most of the other waterblocks and does reasonably well at high flow rates. I wouldn't recommend it with anything less than a Mag3/MCP600 though and I wouldn't use it in a loop where multiple blocks are lowering flow rate below 2GPM.

That is some wierd points on that MCW5000-A? Why does it do sligtly better at 1.25GMP than 1.5GPM? Other than that it it performs very much like I thought it would. I wasn't expecting more flow rate to significantly increase performace as the other one's do because of the design.

jaydee: That's called "real data". Not statistically significant (note the error bars).

I just finished installing the aquajoe block and promptly had to pull it because of a leak. I plumbed in #rotor's block in the mean time; should start testing that one tomorrow :)

looks great pH... the new springs made quite a big difference (around 2C?) as it seems ,compared to your prelim test results with the older springs

Um leak? I thought that the Aquajoe was already under test. Did the leak develop during the testing? Did it start to leak on it’s own or was it a case of loosing or tightening of the barbs as which was the case when the WW leak. If so will you make the same attempts to fix the leak or not? While not impossible it is highly improbable that the unit left Aquajoe with a leak, as all blocks are pressure tested for leaks before shipping. Any additional information would be helpful.

On Sunday night I installed all the mounting hardware for the aquajoe but didn't plumb the block into the loop. Last night I did so and the block was leaking around the barbs. Everything's so tight in that area with 1/2" thickwall hose that I am not completely sure if it's a leak between hose and barb or a leak from barb to block. The barbs weren't even completely fingertight (but they were taped up) so I dunno.

At any rate the leak occurred before I had even mounted the block onto my system and so no harm. I'll definitely get the problem solved and then test the aquajoe out. It isn't a big deal and could very well have been that my hose was a little expanded from going on and off 4-5 other blocks.

Not to rush you pHaestus but when are we going to see the original WW data? This is what I am most interested in. Did D-Tek live up to the original :confused:

In the same line it will be good to see results from MCW5002 block. Just to see real improvements over the MCW5000.

The MCW5000-A and the MCW5002 perform identically as a function of flow rate. The difference is in their flow resistance only.

Am confused
How is the MCW5000-A diffeerent from the MCW5000 tested here:
http://www.swiftnets.com/assets/imag...te-300x250.gif

The original MCW5000 had a top that allowed the bulk of the water to avoid the pins. Bill's design change was the "dropped deck" top that forces water to actually go through the copper pins of the base. That modification is the MCW5000-A. The MCW5002 is the same MCW5000-A design with the 1/2" push in fittings replaced with 3/8" NPT threads for hose barbs.

I think anyway

Ta ,sounds right.

Ok thx, didn't know that.

Heh, Danger Den is gonna think you're out to get them pH. First ya point out the mount issue with the RBX & now the Maze 4 comes in second to the Swifty MCW5000A.

Glad to see more testing from you so soon, great job pH! We really do need a toasting smilie here Joe.:D

I look forward to reading the complete review! :drool:

I have no interest in angering Dangerden; Jeremy seems like a nice guy and has been very helpful to this site for a long time.

I was actually surprised that the Maze4 does so well; roughly 2.5-3C worse than an RBX at the same flow rate (and less restrictive so it should have a higher overall flow rate). Not bad really. I don't completely follow their design strategy with the Maze4 or the RBX though; it seems like they are tuning the blocks for larger pumps than most people use at the expense of performance at lower flow rates (which most people have). Swiftech's approach seems entirely sensible given they use 3/8" hose and push multiple blocks/multiple smaller radiators for complete systems. This means though that people who are only cooling CPUs with decent pumps and lower resistance loops could do better (Whitewater or RBX)

I'll finish testing the #rotor block tonight and probably finish writing up the Maze4 review as well. That leaves tomorrow/Sunday for the original Whitewater (I hear your pleas Nikhsub1). Not 100% sure I'll finish that testing before I go out of town for a week but I'll do my best.

Looks like the maze4 really needs a high flow system to perform great from the graph. I am looking forward to seeing the #rotor block's performance in your review also.

http://3rotor.homelinux.com/images/m...e_a_Number.gif

:D:D

http://phaestus.procooling.com/temp/secrate.jpg

shhh don't tell anyone I posted that.

About to catch a flight; be back this afternoon from Wisconsin :D

Heeehaaa I'm almost not last :D

Awesome job

Hey, until the gpm goes past 1gpm the #rotor block is better performing than the maze4 :cool: It follows nearly the same performance curve as the swiftech but the #rotor seems ~1/1.5C higher which isnt bad at all.

I have an idea, how about testing the dtek TC-4 v2~Spiral aluminum top waterblocks since dtek dropped the price to like $25. I have seen many people wanting to know how those perform compared to the newer blocks because of the new low price. I really havnt seen many reviews of the TC-4 v2 compared to many other blocks either It may become much more popular because of the price and people remembering back in the day the TC-4 was a bit better than the Spiral...atleast thats what I remember. Testing blocks takes alot of time and I can understand if it takes a while but testing those older blocks will show sort of a baseline and show how far ne w waterblocks have come?

jip The grid is optimized for 40x40mm TEC use.... So I did not expect anything Stellar, I'm actually pleasantly surprised at that Graph.

and I had that kinda performance since almost 4 years ago.... Guess that means I'll have to pull my head out my ass and get some development going again... :D

Hehe That was my goal to make an optimized T.E.C block.
Ph I am wondering with all these blocks that you are testing have you noticed or learned anything new that you did not know in the past about block testing and or performance? If so would you share the information? And are you taking pictures to post when the reviews are posted.

PH a few quick questions

Is delta T measured in degrees c?

Also when measuring my block compared to the RBX it is just over 2 degrees c worse with respect to the water temperature but almost identical with respect to the air temperature. I guess this means that the radiator is the limit to performance even though it is quite a large heatercore has anyone else experienced anything like this?

Delta T or Delta anything, for that matter.... is calculated and is always in the unit of measure for the two values in the calculation.
as in 11 apples - 01 apple = 10 apples.... that kinda thing....

it can not be apples - oranges = mangos...... or some such cleverness :) that don't work...

sorry my question wasn't that clear it was the third time I tried to post it and it gradually got shorter and shorter until it din't make any sense.
I was just wondering if it was apples, oranges or mangos (Centigrade, Fahrenheit or Kelvin) :)

I know Centigrade and Kelvin will be the same in terms of delta T, I assumed it would be one of these scales but the flow rate is in GPM so I thought that the temp might be mearsured using the Farenheit scale. This might be obvious to some people due to values but I really have no idea when it comes to Farenheit

As always pH, great job!

WAJ UK,

You say your water temps are 2c worse as compared to the RBX. I take this to mean the water is 2c warmer, which would be a problem at the rad as you say.

Could you post the following regarding you rad set up: Rad's WxHxD, single or dual shrouds, fans diameter & thickness + cfm & pressure? Is the air flowing through the rad outside room temp air, or warmer case air? Many influences in good rad performance, and a large heater core should perform well if set up right.

From what you posted regarding water temps your problem may well be the fan(s) or how the fans are set up, not the heater core.

I caught what you are saying:
If you are using Celsius should you not also use LPM rather than GPM. Well I think that is what you are trying to say :p

Everything is in degrees celsius. I am aware that I should really graph in terms of LPM, but my flowmeter reports gallons per minute and I am lazy :)

I hope that Joe will soon have the website addon done that will dynamically graph information as people request it (just tick off the wbs of interest and it will generate comparative performance vs. flow rate curves). I would suspect that it will be easy to add a "metric" or "US" option for that.

sorry Blackeagle another case of the question getting shortened until it made no sense, I seem to be good at that. I won't get into the rad thing here, it's a bit off topic. I think it might be to do with dodgy measurements and not having a suitable test environment where the ambient temp is nice and steady. I think my results might be completely **** anyway. My delta t is about 4-5 degrees lower than PH's