Now that you are talking about a different flow, it's entirely different. The fact that we don't know how the block does at different flow rates is the reason that JoeC's reviews aren't the greatest, but, on the other hand, if you don't know what flow you'll get (like in pHaestus'), then the curve is useless. The fact that Bill Adams' reviews show pressure drop and performance at different flow rates is what makes his testing the best.

faith is one thing, reason quite another

"Will it do as well as a block with a c/w of .13? Not likely as long as all other system variables are unchanged....but even then I would have to say it IS possible."

rogerdugans
can you address what you said above ?
not some 'other' situation, explain how you concluded the above

I too wish to visit the world where worse can be better

I apologize for the digression from the original thread topic that I seem to have inadvertently started.

But I will make an attempt to explain the statement you quoted in its entirety:

"Will it do as well as a block with a c/w of .13? Not likely as long as all other system variables are unchanged"
Given what I have learned about water cooling, I would expect a waterblock that performs well at medium to high flow rates and has a c/w of .13 to provide lower cpu temps than a block that has a c/w of .17 when used in a system that maintains medium to high flow rates.

I don't think that is the portion of the statement that anybody has a problem with, however...although I could be wrong. ;)
Qualifier to the above portion of the statement:
"....but even then I would have to say it IS possible."

I have read reports from users of small ID tube systems that work quite well with low flow water blocks and pumps that increasing flow rate made no difference or worsened temps. I also recall that some of the blocks mentioned in those posts by users have been tested/rated and have c/w ratings in the area of .25 or more.
And that some of those systems had been tested with blocks with lower c/w and still not gotten better results.
(I will have to see if I can find some of the old posts/reports.....it's been a while.)

I have NOT tested these results myself, and questioned them at the time: however, as I am NOT a scientist, nor a "know it all" type, I did NOT then say that they were full of it.
Nor will I now.

I will not say anything is impossible unless I would be willing to bet my life on it- unlikely, yes. Impossible, no.
I would be willing to bet money that a block with a lower c/w WILL perform better than a block with a higher c/w.....I don't see where I gave any other impression, either?

What I DID allow in that statement is the fact that I do NOT know everything.
Never have, never will.

And that, my friends is why I posted that statement.

I will see if I can find some links to the posts of people claiming better results in low-flow systems with low-flow pumps and blocks.

Edit- ate dinner and looked for some linkage- only one I can find so far is HERE.
There are some other links in there that go to German sites- testing was done there as well.

I do not make any claims that there testing is accurate, only that such posts and claims are/were enough to make me believe that many things are possible...no mater if I still find them unlikely. ;)

Thanks for not getting cranky and flame-happy, guys. That's all I wanted to avoid.

hey - that's a new graph on me......
RBX sure comes off well in that comparison.... and what's an MCW5003g?

I would hope to hell its a new graph, lol (or someone can pull stuff from my computer)
the RBX does well ? - perhaps, depends on the pump capacity though (and sys design)
MCW5003g ?? come on now, think a bit . . . .
betcha Cathar can say exactly what it is
but compare it to the MCW6000, why would we mfgr it ?

rogerdugans
ck the graph; the Innovatek XX specifically as it is a low flow wb
does it appear to have superlative low flow performance ?
hardly
I can easily explain the European 'disconnect' re tubing size/flow rate, this is not my question

the burden of reconciling "Not likely as long as all other system variables are unchanged" and "....but even then I would have to say it IS possible." is yours
- let it be if its too difficult to explain, I understand that those 2 assertions cannot both be true for the same system

To tell the truth, I don't think it is that difficult to explain:

The first states what I believe and expect to be true: a lower c/w will give better temps.
The second is an admission that I do not know everything and admit that there may well be factors that I am unaware of that could change this...although I find it to be unlikely.

I also don't believe in ghosts haunting houses, but since I lack evidence to prove that it is impossible, I will merely say that I don't find it likely to happen. ;)

ah ha ! agreement !
my position too regarding religion, er superstition

I agree wholeheartedly and without reservation on that one too, Mr. Unregistered. ;)

But that is a whole different ball of wax and controversy that I think I will leave alone before someone else comes after me, lol!

well it appears to beat everything, even the cascade, from a relatively attainable 0.5M headloss (a little over 1gpm i believe, since that is 16.3" H20 according to joe's test at overclockers) which certainly doesn't require anything exceptional in the pump line.... and certainly isn't what i'd have expected from pHaestus' tests....

think?? i'll try but it is nearly 2am over here.... ;)
presumably it's a further developement of the 5000>5002 line..... i can't think of what the "g" might be for though....
it appears to outperform the 6000 at all headlosses above 0.25M, which would seem to include any system with an mcp600 - so on that basis (not knowing what it is) i presume that the mcw6000 is a lot cheaper to make...

hmm - it just occured to me... you do have the key right on this graph???

on each curve, each tic represents the progression 0.3, 0.5, 1.0, 1.5, 2.0, and 3.0 gpm
so the RBX flow rate at 0.5m head loss is 1.5gpm

key ? on the graph ?
well hell yes, just read the name of the wb

Cathar, these guys need help

ahh - suddenly all is clear!... so to see the whitewater & cascade at their best we need the rest of it at over 1m....
and the 5003's disadvantage is that it is more restrictive than a 6000 - or even a cascade.....?? but the C/W vs Headloss graph flatters the less restrictive blocks (since the balance of the system would be a higher proportin of the total resistance and so a given pump would give a lower headloss to a less restrictive block...) so then the 5003g is better than i thought - :confused:

Why can't we stick to the old flow rate versus thermal resistance? Ack, this is probably much more useful, if we could see the full thing. I can't think tonight... It would seem that the RBX is the best, but that's at a much higher flow than the Cascade and WW are at in that chart. Do you still test, Bill, even though you work for Swiftech? Possibly for the reason of comparing? Is that the reason this graph exists?

Edit: Well, obviously you must still test, for this to exist. "Why?" is a much better question.

He tests because he is compelled to do so. Bill you should REALLY mess with people and make up an x,y,z graph with C/W, flow rate, and head loss as axes :)

If no tests are made, how do you know where you stand on your design (or others on theirs) ?.

Bill, I'd suggest a package deal for the MCW5003g and MCP1200.

hey; since neither exist, I'll bet its benchmarked at 110%

I'm quite into the numbers game, but truly we are discussing small increments; for any real user less than the mounting variation or the effects of hose bending on the TIM joint's compression

Okay..... heres a diagram I threw together on how I was planning on placing everything. It goes- Pump >> CPU >> GPU >> NB >> Radiator >> Reservoir >> Pump. I've tried to draw everything to their actual size as best I could without getting real technical:rolleyes:. The Reservoir in the diagram is the Tank-O-Matic and the Radiator is the Black Ice Xtreme II. I'm not saying that these are for sure the parts I'm buying, I just used their measurements for the diagram.
Alright.... now the big question. If I had those 120mm fans exhausting (pulling air through the radiator which is what I've heard is best), and seeing that the radiator is pretty close to the dvd-rom...... how much do you think this would this hurt my cooling capabilities to leave the dvd-rom there? I think the space would be about a half inch at the most. So, which would be better.... a single 120mm fan sized radiator/heatercore? Or a Dual 120mm fan sized radiator/heatercore with that very small gap on one side? :shrug: :shrug:

https://home.comcast.net/~snowrydr/diagram.jpg

This case sure wasn't built for watercooling lol...

https://home.comcast.net/~snowrydr/diagramplacement.jpg

you might want to consider splitting the loop after the CPU block and putting the Y adapter
after GPU & NB back to the Radiator
my 2¢

Can that MCP600 handle that? I was thinking that since the pump has good head pressure but not a very good gallons per minute rating that the other way would be better?? I don't know.... I'm no expert.

I'm no expert either..
i was just thinking it would be less restrictive that way...
if i'm wrong I'm shure some one else will pipe up and say so...

Parallel is generally bad, and always complicated. The more resistance you have on your system, the more the MCP 600 shines, relative to must pumps (most being everything but the MD-20RZT)

Alright, here's another question. I just went to the Hardware labs website (the company that makes the Black Ice radiators) and came up with these performance spec's (I underlined the heat ratings so they'd stand out more)......

So my question is how do these ratings compare to that of a heatercore? Everyone says heatercores are better, but I've never seen actual comparisons between the two.

I would think it would be the other way around...
the less restrictive i think would be better..
if the more resistance was better...
then i would kink all my hoses untill i had almost no flow... max resistance...

well I went for BIX rev2 and BIXmicro2 knowing full well that they wern't as eficient as a heater core.
the only reason i went with that instead is due to size...
i was unable to find a heatercore that would fit that size.
complete listing of heatercores listed here

if in this list you can find one that meets your requirements go with that instead.
I have a mini tower, 15" tall and in it i have the 2 above listed heatercores as well as 2x120mm fans in a push pull effect on the BIX and 4x80mm fans on the BIXmicro2 and add 2x1" for the shrowds for each BIX, leaves almost no space. but i got it all to fit :)
but if it wasn't for the size I would have gone with a heatercore. which is what i'm am planning to do with my server(Full height case).

Those radiator dissipation rating are a triumph of marketing. 1838 watts? With what, 80 C water and 90 dB 400 cfm fan? Their numbers are meaningless without knowing the test conditions.

And why the hell do they alway try to tack something to the end of watts? Power isn't watt-hours or watts per hour, it's just plain watts. Damn, I'd like to kick a marketor in the balls right now. :mad: