Status Report

[pHaestus]

Joe is going to be out of town all February, and so it will be difficult to get reviews posted. I am doing my best to get the following writeups to him this week:

Swiftech MCWChill
Dangerden RBX
Little River Cascade
Dtek Whitewater

All tests are completed. I will probably just post test results in my worklog for the month of February rather than waiting a month to post reviews. We'll see though. Swamped at work this week so it should be fun times.

[Incoherent]

Looking forward to it. Immensely.

Don't burn yourself out.

Cheers

Incoherent

[#Rotor]

expect this in a snailmailbox near you, shortly

insedently... my avatar is the insides of that actual block

[nikhsub1]

Originally Posted by pHaestus

Joe is going to be out of town all February, and so it will be difficult to get reviews posted. I am doing my best to get the following writeups to him this week:

Swiftech MCWChill
Dangerden RBX
Little River Cascade
Dtek Whitewater

All tests are completed. I will probably just post test results in my worklog for the month of February rather than waiting a month to post reviews. We'll see though. Swamped at work this week so it should be fun times.

Not even gonna post some graphs for us? Oh the horror!

[pHaestus]

I'll make you a deal. I am writing up the wb reviews this minute. If I can't get a draft finished tonight then I'll just post a graph in this thread

[KnightElite]

Failure to comply noted .

[BillA]

Originally Posted by KnightElite

Failure to comply noted .

ah, KE
but he carefully did not say when he would post a graph
- he's learning from the flaky vendors (or Joe)

[pHaestus]

Having an awful time ftping from work to procooling; was trying to post a smaller version of the graph for your perusal. FINALLY it loaded:



Points are averages (from 8-10 runs) and error bars represent std. deviation (from those runs). Note (a) the excellent performance of Cascade throughout (b) the poor low flow performance of the RBX (c) the crossover point around 2.25GPM where the RBX (essentially a modified whitewater) finally performs better than the standard whitewater. Most users don't have pumps like mine and so relative performance would be Cascade>Whitewater>RBX.

Look to the 1gpm numbers for 3/8" systems and to the 2GPM numbers for 1/2" loops with no other blocks.

//edit to add smaller version of graph inline

[jaydee]

I got a couple (4 ) blocks to send aswell. PM me your mailing addy pH and I will get them out sometime.

[Blackeagle]

Does this mean we won't be getting the full write ups for each block for another month?

[pHaestus]

Nope I sent a review of these three waterblocks to Joe about 15 minutes ago. I'd expect it in the next few days after BrianS reviews it and Joe formats it.

You may or may not get full writeups for the next few blocks I am testing posted as separate articles though right away. Instead they may be just posted in my work log as a new thread rather than on front page as articles. We'll see though. Could be Joe will still be able to format my reviews without issue (or we'll get the new review format coded up so I can post them myself)

[killernoodle]

Instead of simply doing a round-up style review, would it be possible to implement a rolling round up? Kinda like an OC.com heatsink/waterblock compilation? I would like to see a review between many different waterblocks over time all using the same test conditions and variables.

Pretty please?

Also, if I built a block and sent it to ya, would you be willing to review it and compare it to the other blocks?

[pHaestus]

killernoodle:

That's exactly how we are setting it up, but I wanted to begin with 3 reviews at once so that there were some blocks to compare

I will also be testing blocks from our DIY forum yes

[KnightElite]

pH, are you guys implementing that JGraph thinger that Wizzard came up with? One thing we discovered though (if you do it), is that you need to cache the graphs, because if you don't, it generates them every single hit, and server ressource usage jumps way up .

[Incoherent]

Originally Posted by pHaestus

Points are averages (from 8-10 runs) and error bars represent std. deviation (from those runs).

Beautiful.

A lot of work in that simple chart.

Cheers

Incoherent

[geoffman]

Probably a dumb n00b question (but if you never ask you never learn!)

On the y-axis I understand delta T, but what is the meaning of (CPU-wb in)?

Also what nozzle was used with the RBX (just curious)?

[pHaestus]

the delta T is the difference in temperature between the CPU diode (max6655) and the water temperature going into the block (Digitec 5810).

I was using the nozzle that the RBX shipped with; I think that is #1? I have been told that #4 nozzle may perform better at the cost of more restriction. Testing of different nozzles will occur at some point soon.

[Blackeagle]

One question that came to me.

Why 9 points on the graph for the White Water and the RBX, but only 6 for the Cascade?

I thought at first that the Cascade's resistence was high enough to prevent you testing at higher flows. But there is one gap in the Cascade graph which made me doubt that was the reason.

A great graph you posted though pH, lots of points make for a nice smooth, and very clear, curve. Should be helpfull to all who have a idea of the real flow rate of the system they are using.

[pHaestus]

I had initially planned to test at 0.5, 0.75, 1, 1.25, 1.5, 2, 2.5 gpm. The Cascade was first block I tested, and it couldnt make it to 2.5.

[geoffman]

Originally Posted by pHaestus

the delta T is the difference in temperature between the CPU diode (max6655) and the water temperature going into the block (Digitec).

Ahhhh it's so obvious! Why didn't I figure that out? (I'm sooo stoopid sometimes), thanks for not laughing at me pHaestus.

[pHaestus]

no problem. Guess that is something I must clarify in the article. Hopefully BrianS will take a look at it today/tonight

[Blackeagle]

Any time line as to when you'll have time to test that #4 nozzle pH?

I agree it should offer better cooling with your powerfull pump. But it will also raise the RBX's head loss to a level greater than the White Water's.

Both the differance in head loss and temps would be of interest. I'm wondering if my MD-15 would do better with a RBX than a White Water. And the Danner Mag 3 would perform about the same as the MD-15, and more people use the Mag 3.

[Blackeagle]

I was just reading a thread at overclockers in regards to the graph.

GRRRRRR, one guy was claiming that the lack of testing of more of the nozzles means the data is of little use.

Even though that's the only nozzle shipped with it, you pay Xtra for the others, which many won't do!

And if the #4 nozzle makes the head loss higher than the White Water, fewer pumps will be able to drive the RBX to as high a flow rate. Which could still leave the White Water ahead, when you factor in higher flow for the same head loss. And as I understand it, from your posts pH, and other reviews I've read, the #4 noxxle is the ONLY one of the other nozzles that should increase the RBX's performance with even a strong pump.

[pHaestus]

can't comment on nozzles yet other than saying that I suspect ROKK knows what he is talking about as the RBX designer. Here are my comments in that thread:

"I tested the RBX with the nozzle that came with it (#1). I was later told by Dangerden that this isn't the best nozzle for cooling a single CPU but instead is a balance for users with multiple blocks. Dangerden recommends the #4 for better performance when using only a CPU block. I have the nozzle package for the RBX and I will eventually write a separate review testing them all. I spend 3-4 8 hour days of testing per waterblock (or per nozzle in this case) and so it will take me several weeks of testing to compare all of the nozzles (I can't run 7 days/week with all my other duties in life).

I felt that the readers would be better served by testing several common blocks first and then returning to the effects of nozzle on RBX."

And here is the thread in question:

http://www.ocforums.com/showthread.p...hreadid=266384

[Richdog]

geoffman, ditto for me too, lol. I've only just entered the murky realms of watercooling and i'm seeing all of these unfamiliar terms flying around, can be hard to gain a real understanding without knowing them all. But yes, it does seem bloody obvious when it's explained in laymens terms, lol.

Is there a guide on the forum that explains all of them? As soon as I get the underlying principles and terms i'll be set to go, and hopefully it'll all start to come naturally when I read a graph or set of results, instead of plodding through it.

But i'm eagerly awaiting this round-up as just about all the reviews of waterblocks i've seen on the internet (with the exception of BA @ overclockers.com) tend to be hideously biased and woefully lacking in detail. Throw in 1'000 different testbeds and conditions and you have a real mess. Having found these forums I now read them religiously... there's an enormous amount of knowledgeable people and experts gathered here, and most importantly for watercooling beginners like me, they're very accessible.

Having just got my hands on a cascade and a good set-up, i'm set to hit watercooling big-time. I just pray my wallet can stand the pace.