Extremely lownoise heatpipe-towers

CeeJay · Post by **CeeJay** » Sat Aug 28, 2004 3:38 am

I wanted to point out this recent test by Madshrimps.be :

http://www.madshrimps.be/?action=getarticle&articID=213

In it they mostly test various heatpipe-tower-coolers and find them to perform great .. also at low-airflow/low-noise settings.

One of them .. the Sharkoon HSP1 , outperforms even the Thermalright SP-97 at low-noise settings.

Also from a modders point of view I find theese towercoolers very interesting because they seem to offer more possibilities for making good airducts or multiple fan setups .. which may make them the ultimate choice for lownoise aircooling.

Many of them are also much cheaper than the Thermalright SP-97 :)

Bluefront · Post by **Bluefront** » Sat Aug 28, 2004 11:59 am

Very interesting....anybody seen those Sharkoon models in the USA?

josephclemente · Post by **josephclemente** » Sat Aug 28, 2004 3:43 pm

I couldn't read the article for long... The black background with white text was driving my eyes crazy and all the pictures were huge with no thumbnails. I just had to press the close button before the headache would start...

I'll have to wait for the SPCR review.

CeeJay · Post by **CeeJay** » Sun Aug 29, 2004 4:22 am

josephclemente wrote: I couldn't read the article for long... The black background with white text was driving my eyes crazy and all the pictures were huge with no thumbnails. I just had to press the close button before the headache would start...

I'll have to wait for the SPCR review. :)

A SPCR review would be really nice , but until that happens , I suggest you override the page's colors with your own.

Mozilla Firefox (and probably other good browsers) can do this.
You can then choose black on white and perhaps a larger font if you prefer.

CeeJay · Post by **CeeJay** » Sun Aug 29, 2004 4:43 am

Sharkoon mostly have dealers in Europe (I believe Sharkoon is a german manufacturer) , but their website does list one dealer in the US :

http://www.sharkoon.com/dealer/deal_usa.htm

That dealers website is down for the time being though, and they might not carry the HSP1, since Sharkoon have so far mostly sold flashy colorful casemodding gear.

Tweakup.dk also likes this cooler and the fact that it looks really nice in a modded case with neonlight and a casewindow :
http://www.tweakup.dk/article/287/dk/

The article is not available in anything but danish though (sometimes they also translate some articles to english)
If you really want to know what it says then I can translate it (I'm danish)

jmke · Post by **jmke** » Sun Aug 29, 2004 5:00 am

josephclemente wrote:I couldn't read the article for long... The black background with white text was driving my eyes crazy )

no offence, but I hate white/light background with black text more, its more straining then black background white text anyway :p

CeeJay · Post by **CeeJay** » Sun Aug 29, 2004 6:16 am

You must have problems viewing most pages on the net then .. including SPCR's pages.

Are you sure you're not running in a too high resolution with a low monitor-refreshrate ?

I don't have trouble viewing pages on the net.
I'm using a good 19" monitor at 1024 x 768 with a refreshrate of 100hz .. I also make sure to have other more stable lightsources illuminating my computerarea to allow my eyes to relax and I make sure I have no monitorglares to annoy me.

Stable lightsources are ordinary lightbulbs , LED's and sunlight.

jmke · Post by **jmke** » Sun Aug 29, 2004 6:19 am

" including SPCR's pages. " yep.. too bright for me, although I do like the grey back-back ground as you see around the logo..

I'm using HQ TFT screens, white background is more taxing, it's like staring at a lightbulb for extended periods of time. I like darker colours better as they are easier on the eyes.

Anyway, that's all offtopic

got 14 P4 coolers to test for our next roundup, any requests?

CeeJay · Post by **CeeJay** » Sun Aug 29, 2004 6:52 am

jmke wrote:

" including SPCR's pages. " yep.. too bright for me, although I do like the grey back-back ground as you see around the logo..

I'm using HQ TFT screens, white background is more taxing, it's like staring at a lightbulb for extended periods of time. I like darker colours better as they are easier on the eyes.

Anyway, that's all offtopic
got 14 P4 coolers to test for our next roundup, any requests?

Adjust your brightness

I have a Samsung Syncmaster with MagicBright so I can easily set different modes of brightness to suit what I'm doing with the computer at any given moment.

I've got some ideas for interesting coolers to test :
The heatpipes towers mentioned above + other new heatpipe towers.
NorthQ f.x have just made a new variation of the heatpipe tower idea .. they've made a towercooler and put the fan inside the construction so it sucks and blows at the same time.
The NorthQ 3300 Double Dragon :
http://www.northq.dk/products/coolers/nq3300.html

Thermaltake have their Thermaltake Silent Tower Cooler which also looks like it might be worth reviewing .. its another heatpipetower.

More interestingly Thermaltake have also launched new products with "Fanless 0db Heatpipe Cooling" as they call it :

The Fanless103 cooler :
http://www.thermaltake.com/coolers/cl-p ... ess103.htm

and the Silent PurePower Series Fanless PSU ( whoops not a CPU cooler

)

http://www.thermaltake.com/purepower/w0 ... esspfc.htm

Personally I really like all theese new heatpipecooler .. I'm also toying with the idea that you might be able to mod an ordinary PSU to be silent by inserting a heatpipetower into it and letting it stick out the back of the computer so it's cooled by fresh cool air.

The same thing could probably also be done by bending the heatpipes of a cooler like the Fanless103 so the radiatorpart was sticking out of the case.
This could cool anything from the CPU to the PSU (including the GPU) .. as long as the heatpipes were long enough and you could bend them without breaking them.

jmke · Post by **jmke** » Sun Aug 29, 2004 7:06 am

here's the list of coolers in test:

* Aerocool HT-101 [H]
* TTIC NPH-101 [H]
* Thermalright SP-94 [H]
* Thermalright SLK948U
* Coolermaster Hyper6 [H]
* Coolermaster Vortex
* Thermalright SB-2
* Thermalright XP120 [H]
* Scythe NCU2000 [H]
* Scythe FCS50 [H]
* Evercool CUW8-715CA
* Swiftech MCX478-V

might miss some, just quickly written down, those with [H] have heatpipes.

I tried contactng thermaltake several times without succes :-/

silvervarg · Post by **silvervarg** » Mon Aug 30, 2004 6:01 am

I think Madshrimps made a rather good review here. Far better than I would ahve expected to find. Unfortunately they do not try to run all heatsinks with the same fan, so they are not testing the heatsinks that much, they are rather testing the heatsink+fan combination.

The Sharkoon HSP1 with the included fan looks very good for budget cooling in the $35 range. The question is if you would choose to run that one in a tight budget or if you would go with something a lot cheaper, like the Arctic Cooler TC2 for ~$10.

It is very nice to see that the review suggests that aluminium fins works almost as good as copper fins with a much ligther overall weight (and especially less top weight).
They miss part of the point by using different fan sizes and not measuring noise level. Since many of the coolers are rather close this can shift the results completely. The test is also done out of case with motherboard flat. Since most of us will run them in a tower case that also shifts the results quite a bit.

Remarks like this one:

On a DFI Infinity board most of them could only be installed with only 1 fan blowing in 1 direction, quite a disappointment, as using two low speed fans in blow/suck configuration will equal performance of one fan blowing at full speed.!

...is nothing but sad. I guess they missed that day in physics class.

I still think the best heatpipe cooler has not been buildt yet.
It should have most of the features with AeroCool DP-102, like the possibility to choose mounting direction.
It still needs the following small changes:
Fins should start slightly higher up (a few mm) to give more clearence.
Width should be increased a little bit and it should be made to accept 80mm fan only.
Sides should be covered (like the extra duct he made in the review).
Fins should be aluminium to keep cost and weight down and be made square instead of round to increase surface area).

jmke · Post by **jmke** » Mon Aug 30, 2004 7:25 am

Silvervarg.. did you read the review at all? I'm thinking you did not, here's why:

silvervarg wrote:Unfortunately they do not try to run all heatsinks with the same fan, so they are not testing the heatsinks that much, they are rather testing the heatsink+fan combination.

Please QUOTE the review where this is said?

here's what IS said in the review:
"Coolink 80mm fan at lowest and highest settings (24-70CFM / 19-44dBA) on all heatsinks unless noted otherwise."

and that "noted otherwise" is for heatsinks which did not work with 80mm fan size, the NPH-2 and default AMD to be exact.. so that's only 2.

silvervarg wrote:They miss part of the point by using different fan sizes and not measuring noise level. Since many of the coolers are rather close this can shift the results completely.

What part exactly? The only heatsink which supports bigger fans in that roundup is the SP-97, which was tested with both 80/92mm fans..

and now the grand finaly

silvervarg wrote: "as using two low speed fans in blow/suck configuration will equal performance of one fan blowing at full speed.!"

...is nothing but sad. I guess they missed that day in physics class.

Do you mean the PRATICAL test PHYSICS class on which I based my statement? Click the words "Full speed" in the review and you are taking to this url : http://www.madshrimps.be/?action=getart ... rticID=168

Where I test a multitude of Fan combinations on a Tower heatsink, and look at the resuls:
2x70mm @ 7v: 47Â°
1x70mm @ 12v: 48Â°

I have nothing against remarks, pointers, and showing of error in one of my reviews, I do however get ticked off when people don't read the text and then go complaining about the tests...

silvervarg · Post by **silvervarg** » Tue Aug 31, 2004 12:58 am

Coolink 80mm fan at lowest and highest settings (24-70CFM / 19-44dBA) on all heatsinks unless noted otherwise.

Sorry. I did read the entire review, but I must have missed that row.
The white text on black background does make it very hard to read, but that is still a lame excuse for missing that fact.
Using 24 CFM as lowest setting and 70 CFM as highest setting is quite shocking in the SPCR world, and that does explain some of the differences.
Fast speed by SPCR standards is typically a L1A @ 12V, so that is roughly close to the 24 CFM as fast speed. So, perhaps I should change that comment to "a very unlucky choise of cooling fan".
I might add that it would have been real nice if you had written how many watts that CPU is putting out with your extreme overclocking settings.
By SPCR standards very few like to overclock much, so these settings as far from preferable in a realistic test of CPU coolers.

Do you mean the PRATICAL test PHYSICS class on which I based my statement? Click the words "Full speed" in the review and you are taking to this url : http://www.madshrimps.be/?action=getart ... rticID=168

Where I test a multitude of Fan combinations on a Tower heatsink, and look at the resuls:
2x70mm @ 7v: 47Â°
1x70mm @ 12v: 48Â°

This does not follow the measurments made by lots of people at SPCR. Including MikeC's very detailed tests. I believe that you get such a small difference that you get roughly equal cooling in that setup, but that is because you are running the fans with airflow way over the heatsinks sweetspot.
E.g. If you run with 100CFM or 110CFM will make almost no difference.
If on the other hand you compare 20CFM vs 18CFM you will noice a difference.
With low CFM a single fan prooves to cool better than two slower fans at the same noise level. Since this note is already becoming rather long I will skip out on the physics and aerodynamic details.

There is a difference to where you comment a review. In the SPCR forum you are expected to read and see things with SPCR eyes. A comment in an overclockers forum would have been compleatly different.

I have nothing against remarks, pointers, and showing of error in one of my reviews, I do however get ticked off when people don't read the text and then go complaining about the tests...

Good that you can take critical comments. I try to be objective, pointing out both the good and bad points. However I can not spend the same time to write a forum reply as you did when writing that review, so the chance the I make a misstake is ofcourse greater.
You on the other hand just jumped on the negative things I wrote, so I hope you do follow up with comments on the positive and constructive things I wrote.
E.g. How should a heatpipe cooler be designed from your point of view? You seem to have lots of experience with them, so you probably have lots of ideas on what is good and what could be better.

Perhaps you could also add if there is a webshop selling TTIC coolers. Last time I checked around the only way I could get them was by direct import from the manufacturer. Needless to say we are not talking about a single cooler here...

Did you notice that I was the first one (and so far the only one) to actually comment a lot on the contents of the review. Even if I didn't do a perfect job, it might not be good to shout out in frustration if you want others to dare to comment on what you write in the future.

After I made lots of comments on the contents you write:

Silvervarg.. did you read the review at all? I'm thinking you did not

I would say that is one very unprofessional comment.

jmke · Post by **jmke** » Tue Aug 31, 2004 1:27 am

I was reacting on the negative/wrong things in your comment, for that I apologise, it was what drew my attention.

"I would say that is one very unprofessional comment."

somehow reflects the statement I read in your 1st reply "I guess they missed that day in physics class.".. do you have a link to MikeC's tests? I ran the tests with those TTIC coolers combined with medium output 70mm fans, nowhere close to 100CFM.. in each test the temperatures obtained with 2 fans running at half the speed of 1 fan equaled the performance, but not the noise ratio.

The advantage of using 2 fans is such a setup is had because you are providing airflow over the whole Tower cooling, that's why in my testing I saw a minor increase in performance when I "ducted" the DP-102, by using 2 slow fans you are guiding the air through the fins of the Tower cooler, because if you use only one fan the air will blow against the middle column and then just go sideways, hardly passing by the other end of the tower.

"Using 24 CFM as lowest setting and 70 CFM as highest setting is quite shocking in the SPCR world, and that does explain some of the differences. "

The advantage of using the SAME fan on all coolers is that you have +/- and idea of how each heatsink will perform at a certain CFM output. if you lower the CFM, you lower performance, those coolers who scored best in the roundup will also score pretty good when a lower CFM fan is used. That's mainly the advantage of using ONE fan for all test. but I do know that using a 100CFM output fan on all coolers ONLY will create a wrong image. That's why I lowered it to 24cfm with the Coolink fan, you can see the heatpipe heatsink pull ahead and the SP-97 loosing ground. I could test all CFM settings between 1 - 100CFM , but that would take me more then just 2 weekends time

24CFM is pretty low imho and for aircooling close to the minium unless you run your AXP CPU at 1.4ghz using 1.45v vcore, in which case you can almost go to passive cooling with the SP-97 and a silent casefan blowing air onto it from nearby.

"I might add that it would have been real nice if you had written how many watts that CPU is putting out with your extreme overclocking settings. "

this is quote from the review, the page where the results are shown:

"This CPU was clocked at 2006Mhz using 1.85v vcore which makes it give of close to 90Watt of heat"

This can hardly qualify for extreme overclocking as those CPU are known to run 2400Mhz using 1.9v without problem, cooling them with aircooled heatsink does become challenging. Not doable silent.

But 2ghz @ 1.85v is doable with the heatpipe coolers, ducting and a good case ventilation setup. An airduct which brings cool air into the case and to the CPU fan will help a lot, and if you can fit a 2nd fan on that Heatsink and duct it to guide hot air out of the case you will have an efficient setup.

This was however out of the scope and abilities of my testing due to time restraints :/ . A seperate project/review with one of the heatpipes used in the review might be done in the future.

TTIC is mainly a reseller and they don't have global representation yet, although they are available in webshops here in Europe. the Sharkoon heatsink is made by TTIC and I'm sure you will find other named products in the US with TTIC parts inside, search for nPowerTek in the US ( http://www.frostytech.com/articleview.c ... cleID=1621 )

Thank you for your comments, I do welcome them

jmke · Post by **jmke** » Tue Aug 31, 2004 4:54 am

josephclemente wrote:I couldn't read the article for long... The black background with white text was driving my eyes crazy)

Here's an SPCR compatible *HACK* of the article, hope you find it useful

http://www.madshrimps.be/index_spcr.php ... rticID=213

Post by **Ralf Hutter** » Tue Aug 31, 2004 7:54 am

jmke wrote:
josephclemente wrote:I couldn't read the article for long... The black background with white text was driving my eyes crazy)
Here's an SPCR compatible *HACK* of the article, hope you find it useful

http://www.madshrimps.be/index_spcr.php ... rticID=213

Without commenting on the editorial content of your review, I do have to say that the "SPCR Hack" version is infinitely more readable to my eyes than the default white-on-black version. Thanks!

mpteach · Post by **mpteach** » Tue Aug 31, 2004 9:09 am

I read the article thoroughly the first time

But i can read the "Hack" version twice as fast, and with half the eyestrain.

jmke · Post by **jmke** » Tue Aug 31, 2004 9:16 am

I will look into skinning options @ the site

added to our evergrowing to do list

silvervarg · Post by **silvervarg** » Tue Aug 31, 2004 11:56 pm

Thanks for the "SPCR compatible HACK".

This CPU was clocked at 2006Mhz using 1.85v vcore which makes it give of close to 90Watt of heat

Doh! I must have read that page 2 or 3 times and still missed that one. Noting that the speed and voltage was printed 3 times on that page. On the places where I apparently was looking a lot more.

TTIC is mainly a reseller and they don't have global representation yet, although they are available in webshops here in Europe. the Sharkoon heatsink is made by TTIC and I'm sure you will find other named products in the US with TTIC parts inside, search for nPowerTek in the US ( http://www.frostytech.com/articleview.c ... cleID=1621 )

Thanks. Since I live in Europe thats fine for me.

The advantage of using 2 fans is such a setup is had because you are providing airflow over the whole Tower cooling

Are you saying a single fan will not create airflow over the entire fin area? (Assuming we have a shroud on the sides of the heatsink as most heatpipe coolers have).
We could possibly have a minimal dead spot on the opposite side of the heatpipe, but that spot will be almost as large when using 2 fans in blow-suck setup.

24CFM is pretty low imho and for aircooling close to the minium

A typical SPCR setup is to use an undervolted 80mm L1A as CPU fan. At 12V it is rated 25CFM, at 7V about 15CFM.
From the posts I have read most run with 7-9V on the CPU fan, but quite a few run the CPU fan at 5V. Personally I typically don't run any fans to cool the CPU. Most people run the CPU with a lot less wattage. Often less than 60W.

Thanks for a great review. Ofcourse we would have prefered if it was done SPCR style, but even though we were not the intended audience for the review we still got quite a few usefull things out of it, and that is quite rare.

jmke · Post by **jmke** » Wed Sep 01, 2004 3:31 am

I see you have your system build into an Antec Overture, quiet and effective cooling is absolutely necessary with a system like that

I wish that Thermalright adopted the XP-120 to AXP, but I'm afraid not. That heatsink is a real performer!

CeeJay · Post by **CeeJay** » Wed Sep 01, 2004 4:22 am

jmke wrote:I will look into skinning options @ the site
added to our evergrowing to do list

It's a simple matter of using stylesheets and providing alternate stylesheets.
http://www.bluesnews.com/ does this and also provides a little styleswitcher on their sidemenu so browser that don't have a native GUI to switch between stylesheets ( IE ) can easily change stylesheets as well.

CeeJay · Post by **CeeJay** » Wed Sep 01, 2004 4:25 am

Another interesting cooler :
http://tw.giga-byte.com/Peripherals/Pro ... ooler).htm

CeeJay · Post by **CeeJay** » Wed Sep 01, 2004 5:02 am

And another one :
http://overclockers.com/articles1094/

CeeJay · Post by **CeeJay** » Wed Sep 01, 2004 7:40 am

Yesterday I helped a friend of mine install his new Sharkoon HSP1.
I had recommended it to him after seeing the review jmke wrote on Madshrimps.be.

After having worked with it and seen it cool his overclocked Athlon XP 1700+ , I can also without a doubt recommend it to everybody .. It really is an excellent cooler.

My friend have lower load temperatures with the Sharkoon cooler than he did in idle with the AMD Barton 2500+ cooler i had given him earlier ( That cooler wasn't half-bad )
He's very happy and have gone on to see if he can overclock higher now .. which it so far seems he can.

He's not a silence-freak so I only got to hear the cooler a fullspeed .. and it wasn't annoying at that speed .. It sounded as if it made less noise than my Zalman 7000AlCu does at full speed.
It could be better than the Zalman at low speeds too.

I truely feel that this cooler is a contender for the first place of best cooler for silent computing.

jmke · Post by **jmke** » Wed Sep 01, 2004 10:12 am

CeeJay wrote:
jmke wrote:I will look into skinning options @ the site
added to our evergrowing to do list
It's a simple matter of using stylesheets and providing alternate stylesheets.
http://www.bluesnews.com/ does this and also provides a little styleswitcher on their sidemenu so browser that don't have a native GUI to switch between stylesheets ( IE ) can easily change stylesheets as well.

already using Style sheets, that's why the hack was so easy

mpteach · Post by **mpteach** » Wed Sep 01, 2004 11:00 am

jmke wrote:
CeeJay wrote:
jmke wrote:I will look into skinning options @ the site
added to our evergrowing to do list
It's a simple matter of using stylesheets and providing alternate stylesheets.
http://www.bluesnews.com/ does this and also provides a little styleswitcher on their sidemenu so browser that don't have a native GUI to switch between stylesheets ( IE ) can easily change stylesheets as well.
already using Style sheets, that's why the hack was so easy

Yes, but wheres your all-powerfull style swicher button?

I guess theres more than one way to do it, blue news doesnt require the page to reload when you change colors, but my college homepage does when you change themes or pictures. www.students.uconn.edu

CeeJay · Post by **CeeJay** » Wed Sep 01, 2004 2:53 pm

Bluesnews offers an alternate stylesheet the standard way.

Code: Select all

<link title="default" href="/inserts/styles.css" rel="stylesheet" type="text/css">
<link title="alt" href="/inserts/retrostyles.css" rel="alternate stylesheet"  type="text/css">

All that is added is the second line with an alternate stylesheet .. it's that simple!
This is enough for browsers that are aware of alternate stylesheets ( like Mozilla browsers .. and maybe Opera).
A option to change stylesheet appears in the lower left corner of Mozilla Firefox when it encounters a page with alternate stylesheets.

It allows you to choose between Option named after the title attributes.
For bluesnews it's "default" and "alt".

For browsers that don't have this feature bluesnews uses this javascript to handle the switch :
http://www.bluesnews.com/inserts/styleswitcher.js

.. it also sets a cookie to remember the choosen stylesheet.

The script disables all stylesheet but the choosen one .. it does this dynamicly without reloading the page.

The college homepage sets a cookie and then reloads a page that inserts the right stylesheet depending on the cookie.

There is no special browsersupport for this and it requires reloading the page.

I much prefer the bluesnews way of doing it.

mpteach · Post by **mpteach** » Wed Sep 01, 2004 7:23 pm

Ceejay, that's good stuff.

I too like the javascirpt version for stylesheet switching, much more classy ans stylish.

Sooty · Post by **Sooty** » Fri Sep 24, 2004 6:54 am

jmke wrote:here's the list of coolers in test:

* Aerocool HT-101 [H]
* TTIC NPH-101 [H]
* Thermalright SP-94 [H]
* Thermalright SLK948U
* Coolermaster Hyper6 [H]
* Coolermaster Vortex
* Thermalright SB-2
* Thermalright XP120 [H]
* Scythe NCU2000 [H]
* Scythe FCS50 [H]
* Evercool CUW8-715CA
* Swiftech MCX478-V

might miss some, just quickly written down, those with [H] have heatpipes.

I tried contactng thermaltake several times without succes :-/

How about the Bat? ...if it’s not too late to include.

So there's no version of the Sharkoon HSP1 that works with P4?

mpteach · Post by **mpteach** » Fri Sep 24, 2004 7:29 am

CSS are much more powerfull than most people know, expecially when used with a mozilla browser that fully suports css1, IE doesnt support all features completely.

CSS can be used to make menus and several things that previously had to be done in javascript. here are a few links of some crazy stuff.

http://www.meyerweb.com/eric/css/edge/c ... /demo.html

http://www.meyerweb.com/eric/css/edge/index.html

http://www.surguy.net/menu/index.html