Hoot's WB Shootout is on the Front Page

[Tecumseh]

Hoot, great review. I like your "Just the facts, ma'am." style.

I'm going crazy because I crave even more facts. Like,
just how does the Gemini Spiral do such a good job?
Is the base plate extra thin? Is there more turbulence
than other blocks. What was up with your DD block? I
have never heard of this mounting problem before.

 

[nikhsub1]

Well DodgeViper has been saying for weeks now, maybe months that this Gemini block kicks booty.

 

[BillA]

Congrats Hoot

an excellent example of useful info being developed without big bucks

and yes, more flow will drop those 462U numbers right down

temps at a different heat load than Hoot's,
but the temp/flow relationship is the same at all loads

be cool

 

[BBigJ]

In Or Out

Hoot, I was dissappointed that you didn't test the for a dependence on the direction of the water flow. You had 3 spiral designs that in theory might do better with inward flow, and since you were taking the trouble to swap wb's on and off...

Otherwise, great story.

 

[Gnufsh]

In the article, Hoot says

Peter sells not only water blocks, but also entire kits for those who do not wish to "roll their own". He has some interesting theories about the amount of water flow needed to do the job and seems happy to discuss them in detail.

Where can I get these kits, and how do I get in contact with him? Great article, BTW.

Thanks,

-Dan

 

[adamtekh]

some links for the blocks would be nice

 

[rivercom9]

You can checkout the Gemini block here. It has links to the kits as well.

http://www.geminicool.com/spec_CPUcooler.html

 

[emtp563]

I did a mini-review with lots of pictures of Gemeni's chipset and ge-force cooler at my site here: http://www.oc-athlonxp.com/graphics_cooling.htm

Plenty of linkage as well.

 

[Highlander944]

GeminiCool Blocks...

Hey guys, this is Peter from GeminiCool. The website for the products is www.geminicool.com ... I have some seeding to do yet.

We are about 1/2 through a production run of over 100 blocks. They should be ready for shipment early next week.

We are new to consumer based ordering. As such, our initial ordering system will be quite simple. We will accept Visa, MC, or Money Orders (I'm looking into PayPal). Orders may be placed by phone or fax. We do have a toll free number, however I want to make sure it is directed to the proper extension before it is posted.

If you can't wait for me to polish the ordering system, feel free to e-mail me at either [email protected] or [email protected] and I will work out some way to get you the block ASAP.

Thanks,

Peter.

 

[KingB]

Just looking at the rad setup on the shootout. It appears to me that fresh air is being pulled from the case exterior, across the heater core in then released into the case itself. Doesn't that setup introduce warm air into the case? Or is that what all those exhaust fans are for?

I am just thinking as there are other components in that setup that are not water cooled and rely on the ambient air inside the case for cooling. From what I can see, there is only 1 other 80mm fan providing fresh air into the case.

Or is th amount of heat introduced intothe case negligable?

 

[rivercom9]

So better temps can be achieved on the Swifty with a higher flow of water? I bought the Swifty from Hoot that he used in the shootout and I would just like to mate it with a proper pump. I intend on using a HC for a rad.

 

[Highlander944]

These are for real.

Hey guys... one of my buddies suggested that I show you that GeminiCool is for real. He implied that, past CPU coolers with a spiral channel were difficult to actually get. So, for the record, we have completed a sample, a first production run, and are in the middle of a second production run. We have about ten (10) of each block still in stock.

Here are some pictures that I hope will help everyone see who we are.

Current Production Run:

Assembly Area:

Our shop:

Peter.

 

[rivercom9]

Ahhh!! Broken links.

 

[BillA]

rivercom9

how much heat you putting into the coolant ?
a big pump will introduce heat also
my 1060 is 50W

probably a 1025 with 1/2 in tubing, and NO elbows

KingB

the 24^C in-case temp was due to cutting back the fan speed to achieve a wb coolant inlet temp of 25^C

there is a move afoot to "standardize" on 25^C as the inlet temp for the testing of all wbs

this will enable head-to head comparisons of different testors' data

now we're looking at heat die simulators to try to equalize the applied loads
- can't be done with CPUs

be cool

 

[rivercom9]

I want my water rig to beable to cool a 130W CPU. How much heat is introduced into the system by an Eheim 1060?

 

[Doc_T]

Peter i see in your second picture there a compressed air hose with a regulator, 2 pieces of clear tubing and a vat of water is this to pressure test the blocks after assembly by placing in water and looking for bubbles? i have used this method to leak check other items

 

[BillA]

rivercom9

in a res ~ 50W, in-line somewhat less
not too sure that there will be a marked improvement with a 1060 over a 1250
(I use a 1060 for testing, not as part of a CPU cooling system)

what you need to pay attention to is the air flow through the rad

more air = more cooling

get a big fan and use a shroud with 1 to 1.5 in. of standoff
then cut the fan speed back when you get tired of listening to it

be cool

 

[Highlander944]

haha one last time.. size only matters a little.

You only need a system flow rate of about 75gph to remove 300Watts of heat and raise the water 1C. That's the physic. The key is to tune your block to your system flow rate. Once that is done, you've got it licked.

Hope that helps.

 

[BillA]

Highlander944

"You only need a system flow rate of about 75gph to remove 300Watts of heat and raise the water 1C. That's the physic."

and what does "the physic" indicate the die temp at equilibrium will be ?
why would anyone want or settle for a coolant temp rise of 1^C ?

the point of higher flow (for wbs that benefit from such) is to lower the equilibrium temp at the die

do observe the graph posted earlier in this thread

be cool

 

[Tecumseh]

haha one last time.. size only matters a little.

You only need a system flow rate of about 75gph to remove 300Watts of heat and raise the water 1C. That's the physic. The key is to tune your block to your system flow rate. Once that is done, you've got it licked.

Hope that helps.

Are you sure about that? I didn't do the calculations yet,
but it might be true only for perfect heat transfer where you
have a lot of turbulence. The rule of thumb would be about
329 watts for your stated conditions, but can you really
achieve that?