The new Gemini water blocks

[Hoot]

This thread was created due to the original thread splitting into two areas of interest, namely Q & A about the Gemini blocks and Q & A about numerical modeling of water blocks in general. If you have questions about the Gemini water blocks that were reviewed in this article that you would like to ask or if you have gotten your block and want to contribute to answers about them, then this is the appropriate thread. If you wish to participate in the Numerical Modeling thread, it is here .

The original Gemini thread is still available for reading, but I have locked it to new posts to encourage the split of subject.

Hoot

 

[Tecumseh]

Highlander944, are you going to produce aluminum
versions of the Gemini WBs for people with all
aluminum cooling systems?

Could you send samples to Hoot for testing?

 

[Diggrr]

Nice move Hoot!

I do have a couple of questions in fact...What are the (stock) sizes of tubing to be used with the high flow and low flow blocks? (npt size too, please)

How thick is the copper base, and overall thickness? Since I've heard the 4 holes are to disapear before long, I'd like to research my mounting options. I'm easily handy enough to either make a new lexan plate without the holes, or to remove the 'ears' that have them now, without ruining the nice edge finish.

Thanks ever so much for your time and efforts <salute>!

 

[Malakai]

i have 2 questions,
1) can u put 1/2in barbs on a high flow model?
2)how much would such a small block impede my 500gph flow rate? i have a 500gph danner, and was gonna get a maze2, but that is a very good review(for highlander)
-Malakai

 

[Hoot]

I just don't get barb dimensions. What I call OD, the industry seems to call ID. Here's what I can tell you. Both blocks I have here have barbs that are 3/8" OD and 9/32" ID. I believe the industry refers to these as 3/8" ID barbs. Go figure...

There is nothing to gain by putting higher higher ID (my version) barbs on because the channels in the blocks are already smaller than 9/32. Someone, perhaps Dodge Viper, found some 1/2" OD, 9/32" ID (my version) barbs so you can accomodate 1/2" ID tubing.

As far as how the block will work with a Danner 500gph pump. My first advice would be to take that alleged 500gph pump, put your tubing on it and see how long it takes to fill up a gallon jug. I'll bet a dollar to a donut it does not fill up in 7.2 seconds, even with no head. So, you need to find out what your Danner 500 really delivers. Then you can assess the impact that the blocks will have upon it. I believe the laws of physics (flame invite) say you can only pump approximately 340gph of water through 1/2" ID tubing and thats with unlimited pressure, which most hobby pumps don't even come close to having.

There should be some "truth in pumping" disclosure requirement for these pump manufacturers.

If I had a few hundred dollars to waste, I'd buy the top 10 or so, most popular pumps, rent an accurate flow-meter and do an article on how well they really deliver with both 3/8" ID and 1/2" ID (my version) tubing. Perhaps that would help the trusting souls out there choose more wisely.

Hoot

 

[KingB]

I imagine they call them 3/8 ID barbs so you know what size tubing they should be matched to. That would be my guess.

As for changing them, why? The block was designed for the 3/8 then why would you want to mess with it? The channel diameter inside the block cannot be changed anyway. (ahh.... dumb question, no one on this forum can leave things the way they are... everything has to be modded...sorry I forgot).

 

[Staz]

I just don't get barb dimensions. What I call OD, the industry seems to call ID. Here's what I can tell you. Both blocks I have here have barbs that are 3/8" OD and 9/32" ID. I believe the industry refers to these as 3/8" ID barbs. Go figure...



Hoot

When it coems to hose barbs the ID measurments are the ID of the hose that conects to the barb. While the actual ID measurment should be the same as the ID measurement on the npt side of the bard. (I always knew 3 months writing suply request for engine room 1 while in the navy would come in handy one day)

 

[Hoot]

The standard flow rate unit looks to have a 1/4" thick copper base and 1/4" thick lexan cover, for a total of 1/2". The high flow rate unit looks to have a 1/2" copper base and 1/4" lexan cover, for a total of 3/4". Making a retainer clip to take the place of the four holes would not be much of a challenge. Something like this:

Hoot

 

[Diggrr]

The point is, I now have a 3/8" system, so the fittings it has would work out of the box. That's all I needed. Some systems are getting into strange hose sizes, 10/8mm and such, I just wanted to make sure I wouldn't have to change anything.

Anything on the block thickness? *edit* oops, thanks!


I haven't heard anything from my machinist/friend about the copper and design I gave him, so I'm looking for the best commercial version (which I think Gemini is). He's 3 weeks overdue in getting back to me.

Thanks again.

 

[DodgeViper]

Ok I am finally able to say I have joined the ranks of water-cooling. I am using the Gemini High Volume Spiral water block, Eheim 1250 317GPH pump and a custom Chevy Chevette radiator with 2-120 PanaFlo 86 CFM fans configured in a pulling/pushing format.

I decided to test the pump for flow rate. With a 12” rise, the pump delivered 190 GPH. With the Gemini High Volume water block added to the pump at a 12” rise the flow through the block was 60 GPH. I had changed the fittings on the block to fit my 1/2" tubing system.

Currently I am still burping all the air from the system and here in Arizona is rather warm tonight. My current ambient temperature is 28c and the cpu is at 37c. These numbers are at idled. My cpu is a T*bird 1.4 @ 1575. No volt mod but everything in the bios is set for extreme. Hoping by morning the W/B has had a chance to settle in, and the temps within the home will be down as I do not have the cooling in the home running yet.

Prior to changing over to water this same system running air-cooled the cpu was at 45c with the ambient temperature at 29c.

For some strange reason I think I have trapped air in the radiator. I mixed distilled water with 15 percent of anti-freeze. I only mixed 32oz. And still have 4oz. of mix left. Hoot do you have a clue how much water you used in your system?

The tubing routing on the system goes like this, pump outlet to rad, rad to (T) then to the center of wb, wb back to pump inlet.

 

[Malakai]

the reason i need 1/2in fittings is i already have some of the h2o's components,and i was planning on using a 1/2in system with a maze2. everything but the block is 1/2in. i might get a swifty block though, better flow, ill test the gallon per minute thing as soon as the pump gets here
-Malakai

 

[schoolie]

I've found that 3/8" silicone hose works really well with 1/2" fittings. It takes a little elbow grease to get the hose on, but you don't need clamps and don't have to worry about the hoses popping off. I'm running a Gemini block at the moment. I too have 1/2" hose so I bought a brass hose mender, 1/2" barbs on both ends, and a couple of short pieces of 3/8" hose to make a reducer.

 

[DodgeViper]

Snap shot of my current motherboard monitor. I was able to push the system to 162 fsb x 10x =1620.

 

[Tecumseh]

The standard flow rate unit looks to have a 1/4" thick copper base and 1/4" thick lexan cover, for a total of 1/2". The high flow rate unit looks to have a 1/2" copper base and 1/4" lexan cover, for a total of 3/4". Making a retainer clip to take the place of the four holes would not be much of a challenge. Something like this:

Hoot

Hoot, you confused me. Didn't you report that the base
thickness of the Gemini was 1.25mm in the "Old Gemini
WB" thread? OR was that a type-o?

 

[Hoot]

-snip-
For some strange reason I think I have trapped air in the radiator. I mixed distilled water with 15 percent of anti-freeze. I only mixed 32oz. And still have 4oz. of mix left. Hoot do you have a clue how much water you used in your system?

Hmmm. 32oz sounds about right to me. I probably should have measured and recorded that amount for future reference, but I didn't. Whats weird is the flow reduction you sustained with the high flow rate block. Were you able to measure the internal diameter of the barbs you put in? The ones that came with the Gemini were 9/32". I assume the ones you put in were that ID also. When I tested the four blocks for my article, I coupled my 3/8" ID hose to the kitchen sink spigot and since the Eheim 1048 was rated for 158gph, I set the spigot to fill a 1 gallon jug in 24 seconds or 150gph. I then hooked up each of the four blocks to the hose and timed how long it took to fill the jug with each of them. The MCW-462, Maze-2 and Gemini high flow model did not add any time to how long it took to fill the jug. The Gemini standard flow block took longer (30 seconds). I think my testing process was flawed in that I did not actually use the pump instead of the sink spigot. Since my water cooling system was already set up, I did not want to have to remove the coolant from it and tear it all apart. This may have been a mistake. When I get back from Chicago, next week, perhaps I will do that. I've got enough coolant to do a complete refill.

When I originally set my system up. After I filled it and burped out the major air, I tilted it every which way and thumped it while the pump was running. I got a lot more air out of the radiator. Eventually, it settled down, but I continued to get air out for almost a week.

Hoot

 

[dgunthert]

I'm about ready to watercool my new P4 2.0a system. I already have an Eheim 1048. Now it's time for the wb. Will we be blessed with Socket 478 versions of the Gemini any time soon? I love the simplicity and thinness of the design (not that I know a damn thing about the physics).

 

[Ichelo351]

my first question is, can this waterblock be used with a peltier? and would it be able to keep up with a higher amount of heat? like i would imagine that at a more moderate amount of heat, this wb would do great, but with say a 250 watt peltier, would a larger wb like the swiftech be more able to keep up?

because i am looking into what supplies to get for watercooling, and i think i want the gemini, but at some point i may want to use a peltier, so i must wonder if the gemini is able to be done with a peltier. (and the 250 watt figure is just for a what-if number, i doubt if i would ever use a peltier quite that high powered...)

 

[Aesik]

I coupled my 3/8" ID hose to the kitchen sink spigot and since the Eheim 1048 was rated for 158gph, I set the spigot to fill a 1 gallon jug in 24 seconds or 150gph. I then hooked up each of the four blocks to the hose and timed how long it took to fill the jug with each of them. The MCW-462, Maze-2 and Gemini high flow model did not add any time to how long it took to fill the jug. The Gemini standard flow block took longer (30 seconds). I think my testing process was flawed in that I did not actually use the pump instead of the sink spigot. Since my water cooling system was already set up, I did not want to have to remove the coolant from it and tear it all apart. This may have been a mistake. When I get back from Chicago, next week, perhaps I will do that. I've got enough coolant to do a complete refill.Hoot

This would definitely be a problem. Your faucet is going to have a LOT more head than a pump and the results would not be comparable.

And just another note about head loss. If one system has all 1/2" tubing, but necks down to 3/8" fittings at the block, it will still have less head loss than a system that has 3/8" tubing and 3/8" fittings on the block. Smaller tubing throughout the entire system will contribute more to head loss than larger tubing. Just because one certain aspect of a system has a smaller diameter than the rest, it alone does not determine the total flow rate. Head loss is calculated throughout the ENTIRE system, and flow determined by the ENTIRE system, not just the smallest bottleneck.

 

[lebe0024]

I have the Rio 1100 pump (rated at ~380 GPH). I also will have a heater core from the junk yard. These will be the only two components other than my WB. So, I have two questions:


1) Should I get the High flow model? I bet my GPH won't be at 380 by the time it hits my CPU, but what do you think? I need to know quick, because I'm ordering today.

2) should it go : pump->block->rad or pump->rad->block.

 

[Malakai]

i say pump>rad>block so the coolest water hits the block 1st. prolly not too much difference though.
and as Hoot's review showed, it cant hurt to get the high flow model
-Malakai