Overclock.net banner
41 - 60 of 121 Posts

·
Registered
Joined
·
2,065 Posts
The Apogee XL has worse hydraulics than the HD? That's weird, Swiftech claims the other way around:


Stren's results show the HD having more restriction:
 

·
Registered
Joined
·
2,065 Posts
Quote:
Originally Posted by Moonmanovich View Post

Hi!

Strange, i didn't notice this before so I went and compile some numbers from myself, stren and Martin.

Martin and @stren appear to have similiar method for flow testing and therefore their results are not directly comparable to mine (except for the order - best to worst). Stren vs. Martin numbers are however quite comparable and this is where I have spotted this (stren is solid ______, martin is dashed _ _ _ _ _):



- Raystorm CPU block results are in-line.
- stren's testing shows Apogee HD to be much more restrictive than Martin's testing (control result).
- stren's testing shows Koolance CPU-370 to be much less restrictive than Martin's testing (control result) and on par with Raystorm, which both mine as well as Martin's testing disprove. Koolance CPU-370 is more restrictive than Raystorm for sure.

I have now re-assembled the block and re-ran the test and got repetitive results. I'd hate to say but I think something is off with some of the stren's numbers.
confused.gif
Thanks for assembling that graph!

I do trust your numbers and method, I'm simply interested in learning why there's this difference. I know the same waterblock can behave thermally differently because the bowing is not always the same, and the CPU IHS plays an important role there too... But such a large difference in waterflow/restriction must be something else.

Could imperfections of the fin matrix be the origin? Or the acetal tops (aren't those like clones)?

/summon @Martinm210 @stren @stephenm
 

·
Registered
Joined
·
2,065 Posts
Quote:
Originally Posted by Moonmanovich View Post

I know, it's really puzzling. I see a few points where an error could occur:
  • clogged water block due to 'plastificator' or debris from radiator - I cleaned mine before flow testing them.
  • water loop not bled of air bubbles - if the loop if full of air the flow rates are lower. I bled the system of air bubbles before taking measurement.
  • manometer not zeroed after each run - i zeroed once the system was bled of air.
Fin matrix imperfection should not be the case. In my opinion even if there are some bent fins they would not impact the result by THAT much.

tiredsmiley.gif
I guess we'll never know then...

As for the fins, you're right, it wouldn't even affects temps, let alone flow.
 

·
Registered
Joined
·
170 Posts
Moonmanovich (jel s'to ti, Luka Mesec?
biggrin.gif
), can you make an educated guess as to where the older EK Supreme HF block would
place in this roundup?

I have it in my system for a few years (running on i5 2500k), so I'm wondering if there would be any point in replacing it with the
Supremacy (EVO).
 

·
Premium Member
Joined
·
5,583 Posts
Quote:
Originally Posted by Moonmanovich View Post

Hi!

Strange, i didn't notice this before so I went and compile some numbers from myself, stren and Martin.

Martin and @stren appear to have similiar method for flow testing and therefore their results are not directly comparable to mine (except for the order - best to worst). Stren vs. Martin numbers are however quite comparable and this is where I have spotted this (stren is solid ______, martin is dashed _ _ _ _ _):



- Raystorm CPU block results are in-line.
- stren's testing shows Apogee HD to be much more restrictive than Martin's testing (control result).
- stren's testing shows Koolance CPU-370 to be much less restrictive than Martin's testing (control result) and on par with Raystorm, which both mine as well as Martin's testing disprove. Koolance CPU-370 is more restrictive than Raystorm for sure.

I have now re-assembled the block and re-ran the test and got repetitive results. I'd hate to say but I think something is off with some of the stren's numbers.
confused.gif
Sometimes I have got a funny block from Swiftech - their titan for example showed bizarrely high restriction in comparison to the identically designed hydrocopper.

Your EVO data is weird though, how can it be less restrictive than the original supremacy? Even Niko agreed it should be a bit more restrictive I thought, though he was suprised how restrictive my EVO was measuring, just as he was surprised by my thermal results. Which jetplates were you running on each - perhaps that is the difference?

The HK3 is also quite bizarre on your charts - running very high flow when all of watercool's products are pretty restrictive.

FWIW I try and clean out the blocks before running flow on them as crap can build up. As for water loop not bled - I use a continuous feed of water from a faucet which is why I'm able to run up to much higher pressures and flow rates than when using a pump. The air introduced to the system is quickly cleaned out.
 

·
Call me VSG
Joined
·
12,870 Posts
For what it's worth, I am running a "low" restriction test loop with the Evo in it and it is the single most restrictive piece in it by far as measured by a Dwyer manometer. The jetplates definitely affect it but more so the insert than anything else- bleeding it takes way longer than I imagined.
 

·
Registered
Joined
·
2,065 Posts
Quote:
Originally Posted by geggeg View Post

For what it's worth, I am running a "low" restriction test loop with the Evo in it and it is the single most restrictive piece in it by far as measured by a Dwyer manometer. The jetplates definitely affect it but more so the insert than anything else- bleeding it takes way longer than I imagined.
I think I won't buy the EVO upgrade kit...
 

·
Call me VSG
Joined
·
12,870 Posts
Quote:
Originally Posted by cloppy007 View Post

I think I won't buy the EVO upgrade kit...
Why is that? Bleeding is a one time thing, and longer than expected doesn't mean a long time. Usually more restrictive blocks also tend to have more performance which, in this case, seems to be true so I would not mind one bit- CPU blocks aren't very restrictive in general as it is.
 

·
Registered
Joined
·
2,065 Posts
Quote:
Originally Posted by geggeg View Post

Why is that? Bleeding is a one time thing, and longer than expected doesn't mean a long time. Usually more restrictive blocks also tend to have more performance which, in this case, seems to be true so I would not mind one bit- CPU blocks aren't very restrictive in general as it is.
Too little of a thermal improvement while sacrificing hydraulics. Plus, it costs more than half a new block (35€).
 

·
Call me VSG
Joined
·
12,870 Posts
Quote:
Originally Posted by cloppy007 View Post

Too little of a thermal improvement while sacrificing hydraulics. Plus, it costs more than half a new block (35€).
Valid reasons, all. Honestly I don't believe anyone with a current flagship CPU block should "upgrade" to this unless they want a different aesthetics choice- otherwise, it makes little sense.
 

·
Registered
Joined
·
2,065 Posts
Quote:
Originally Posted by geggeg View Post

Valid reasons, all. Honestly I don't believe anyone with a current flagship CPU block should "upgrade" to this unless they want a different aesthetics choice- otherwise, it makes little sense.
I agree, "upgrading" makes little sense usually. I downgraded from an Apogee HD and my temps are slightly worse (and so is the contact between WB and IHS).

 

·
Premium Member
Joined
·
5,583 Posts
Quote:
Originally Posted by Moonmanovich View Post

Hello stren,

I think you are not reading my data correctly, please have a look at it again. The EVO is a tad more restrictive than old Supremacy, but still on par with Koolance CPU-380i.
Supremacy was running jet plate J1 while Supremacy EVO was running J2. I don't think it makes much of a difference.
There are couple of variants of Heatkiller 3.0, are you sure we are talking about the same one? The one I have tested is pretty high flow, as opposed to Aquacomputer Cuplex Kryos Pro.
Disclaimer: I take apart all water blocks after testing and clean them. I am also using Tygon Norprene tubing for extra precaution.
Hello geggeg,

yes, any water block is the most restrictive element in the loop. I don't think it is restrictive at all, if you look at my charts they are on par with Supremacy or Koolance CPU-380i, which are considered very high flow water blocks. I think longer bleeding time is due to the design of the insert.
Hello cloppy,

what CPU is that?
Lulz I could have sworn the evo and supremacy were the other way around when I saw it last
thumb.gif
I must be getting old. My supremacy was running the 0.7mm beta jetplate out of the triple stac while my EVO was running the 1150 setup. Jeez I get confused between all of them, but my EVO was way more restrictive than yours it seems. I did have worse results on it than you did too :/

GPU blocks can often be more restrictive than cpu blocks, as can some of the smaller QDCs. And some CPU blocks are super low restrictive (though none that are in production anymore e.g. iceforce and 5noz/sniper.

I'm not sure what was up with the 370 data, I still have it so maybe I'll re run at some point.
 

·
Registered
Joined
·
2,065 Posts
Quote:
Originally Posted by Moonmanovich View Post

Hello cloppy,

what CPU is that?
3570k
 
41 - 60 of 121 Posts
Top