Can anyone rate the primochill Typhoon III?

[BoxGods]

is this the GOD of all watercooling? I dont understand you guys are way too technical for me

No worries man, I don't understand it all either. I look at it like this: You don't need to understand every single thing about pheromones, reproduction, genetics, natural selection, and biology to know kissing a beautiful woman is a great way to spend your time =)

[Vapor]

I don't agree. OK, maybe a more accurate way to say it, but your very first statement WAY back in this thread is the easiest way to understand it. The difference is WHERE the split (for want of a better term) happens. In the Typhoon III it happens in the pump itself.

Yeah, I probably got too technical....I've been plodding through numbers for a couple days now and it's gotten seemless for me

I totally agree that where the split occurs is the key. It's the difference between merely reducing the restriction of a set of components and actually enhancing the flow through a set of components.

[millertime359]

A long time to never I am afraid. Comes down to cost for the most part. The tooling for the fittings would pay for a nice car.

Yea, I understand that. I started using that 7/16 and started to like it. Seeing the ghost fittings seem to work best on the T3, I was hoping they would start supplying them.

[JOCKTHEGLIDE]

No worries man, I don't understand it all either. I look at it like this: You don't need to understand every single thing about pheromones, reproduction, genetics, natural selection, and biology to know kissing a beautiful woman is a great way to spend your time =)

it seems like 13 pages of something I see one guy comlaining about tape on it then I see others talking about two loops.....

[millertime359]

it seems like 13 pages of something I see one guy comlaining about tape on it then I see others talking about two loops.....

Seems you got the majority of what was going on

[JOCKTHEGLIDE]

Seems you got the majority of what was going on

[nypunkgeek]

What kind of manufacturing are you using for this. I have a hookup with a plastic injection molding company if that is a type of manufacturing you are looking for. PM me if you would like more info. They specialize in short run things and are very good at keeping costs really low.

[skinnee]

That is why I am confused here, I read all the talk of parrellel loops being insignificant and essentially that is all this is, is parallel loops. Initially I was thinking of getting rid of my DDC 3.2 and getting this res and a d5 but in all actuality now I think I would be worse off if I did that.

It just seems confusing as if the T3 is some revolutionary new thing but in all reality the breakthrough here is likely parrallel loops not the T3. Seems like same thing is going to happen if you just have a pump top with two outlets or a pump top with a y on the outlet?

The T3 is a major step for our loops, what product has been released in the last couple years that provided this much of an advancement? I can't think of anything that comes this close.

The whole concept is confusing, mostly because we've had the idea that parallel is bad, mom using your full including middle name bad. Look at the several threads that were major technical discussions when parallel GPU block configs started popping up...people went bat crazy.

I won't begin to say I know the numbers behind the design or the formulas like Vapor does, he has the big brain around skinnee labs...me, just the squirrel sized gray matter. Vapor and Geno both have as firm an understanding on the numbers as anyone. You didn't see my frantic run up the stairs when the first numbers popped up on testing...it was like my hair was on fire. My thoughts were along the lines of I really fubar'd something, this can't be right.

Anyways loking forward to the results, It's all kind of confusing at first but I think I am starting to get the jist of it now Hopefully Skinnee will have temps to go along with his numbers.

Cheers!

These numbers are pretty insane. Question though, will it make a real world measurable difference in temps of the components being cooled? Not hating, just curious?

I wish there was time before the review deadline, but I'm already on thin ice for taking as long as I have.

However, I feel its not needed. Look at the testing Vapor has done on flow dependence for blocks, some blocks scale very well with increased flow while others are not. Its no secret that when flow increases through a radiator (to a point) the cooling performance will increase, as to what scale factor, that has not been determined.

A long time to never I am afraid. Comes down to cost for the most part. The tooling for the fittings would pay for a nice car. Added to that is the other matched fitting parts that are coming so that there is a FULL line of big radius T's, Y's, Elbow's, flow wheels, Bulkhead pass thru's etc.







The tooling for just one size is I would guess North of 25K

DarthBeavis has been telling people, Geno's designs are solid. Another thing you guys need to remember is Geno is one of us...he's a modder first so if there is a part he is designing such as the fittings above, you know the design will be sound.

Now when can I get a hold of some of those T's and Y's? I need some more T's!!!

[BoxGods]

I haven't had time for much modding lately =(

I do have to get a build done for the nVidia booth at Quake Con though and then there is the nVidia ION mod contest I am participating in. If your interested in submitting a design for an ION based mod...hit the link

Speaking of T's, there will be two. A high flow true T and a through T that you would use for things like fill port lines where you only need a branch to top off with etc.

[Vapor]

Which T is the red one above? Cause that looks like a Spread Eagle Y to me

I might be in the market for some fill port type Ts soon....and being transparent would be a huge plus.

[Utnorris]

Ok, so since we are all still waiting for Skinnee to release his official review I thought I would share my observations now that I have it setup and running with my system up and running also. Now, first this is by no means scientific and may or may not reflect your results or even Skinnee's. That being said, here is a list of my components:

Hardware:

Rampage Extreme II
I7 920 @ 4.2Ghz with 1.28v Vcore
6Gb Dominator GT 2000Mhz @ 8-8-8-24 2t Timings
3 x GTX280 Asus Top
Intel X25-M 80Gb SSD

Cooling:

CPU Sub-loop
Heat Killer Cu 3.0
Thermochill PA120.3 with 3 x San Ace Fans at 5v

GPU Sub-loop
3 x EVGA Hydro Copper Full Cover water blocks in parrallel
Thermochill PA120.3 with 3 x San Ace Fans at 5v

RES/Pump:
Typhoon III with D5 Vario

3/8" ID, 1/2" OD Masterkleer tubing with Koolance QD's and BP Compression fittings.

Ok, so now we have the hardware listed, here is what I have observed so far and again this is in no way scientific.

The Good:

1 - Much better flow than my two MCP355's with XSPC tops, specifically on the CPU loop.

2 - Cleaner setup due to one res/pump combo.

3 - Temps on my GPU loop have gone down by a couple degrees.

4 - Looks sweet even with the stripe across the res.

5 - Less expensive than running two separate loops with separate res' and pumps.

The not so good:

1 - CPU temps didn't go down. This is probably due to the limitations of water cooling. I think I have hit my max without either doing a push/pull configuration or shrouds, both of which would be a tight fit.

2 - It's a B**ch to bleed. This is mainly due to me having such a large case and a lot of heavy equipment in the case that makes it hard to turn on it's back. Also, I think the remaining air bubbles are having baby bubbles since I can't seem to bleed it completely, close, but I still have a couple of bubbles.

3 - Fat base fittings require a new o-ring to make a proper seal.

So in conclusion, while my CPU temps did not improve my GPU temps did. This is probably due to the I7 being a very hot processor, however my GPU temps did improve probably due to the increase in Rad space being shared by the CPU loop that the CPU was not using. While there were a few things that I was disappointed in such as the white stripe, after I got it setup those things were easily forgotten by the overall performance of the unit. Remember, I am getting at least the same results and in the case of my GPU's, better results than I did with to separate loops. Up until now I would have called BS on this, but seeing is believing. I think the Typhoon III sets new standards in what we can expect from water cooling especially from a single loop divided into to two loops. Before doing this I would have stated that a dual loop would be the best option where isolating the CPU by itself would be ideal. After setting this up I am now convinced that is not the case and this single pump/res combo is equal too, if not slightly better than two individual loops. If I had to put a rating on the Typhoon III it would be:

Performance: 10 out of 10
Cosmetics: 7 out of 10 due to the white stripe and it not having LED holes.
Ease of setup: This is one is complicated since it cleans up the loop, but I found it a PITA to bleed compared to other pump/res combos I have used such as the XSPC res top, however I would still give it a 8.5 out of 10 in this category.

Overall: 9 out of 10 in my book because I value performance more than anything else and it definitely lives up to the performance expectation considering it is a single loop at heart and easily keeps up with a separate dual loop configuration.

So there you have it. Again, this is not scientific, nor will it necessarily reflect results from anyone else. I hope that this helps those that were sitting on the fence decide as I am very happy with my setup.

[nypunkgeek]

I found it a PITA to bleed compared

But once it is bled it looks so good.

[Utnorris]

I don't disagree. By the way, your build looks Sweet!!!

[bentleya]

Geno, ON another forum i have some one who has few problems. He has a Tj07 and it dos'nt fit, do you think you could head over to WCUK?

[Utnorris]

Here is mine, maybe this will give you an idea of why say bleeding was a pain for me. My system is nowhere near being done, probably never will be.

[BoxGods]

Which T is the red one above? Cause that looks like a Spread Eagle Y to me

I might be in the market for some fill port type Ts soon....and being transparent would be a huge plus.

Call the leg with no tube on it 9 o'clock. If the waters main path were from the 9 o'clock leg to the 3 o'clock leg with the 6 leg being just a fill port leg...well not so good obviously, Now if the 6 leg was feeding the 3 and 0 o'clock leg then you get very clean lower restriction flow. The better choice for a fill port type T would ba a pass through or straight T.

Ok, so since we are all still waiting for Skinnee to release his official review I thought I would share my observations now that I have it setup and running with my system up and running also. Now, first this is by no means scientific and may or may not reflect your results or even Skinnee's. That being said, here is a list of my components:

Hardware:

Rampage Extreme II
I7 920 @ 4.2Ghz with 1.28v Vcore
6Gb Dominator GT 2000Mhz @ 8-8-8-24 2t Timings
3 x GTX280 Asus Top
Intel X25-M 80Gb SSD

Cooling:

CPU Sub-loop
Heat Killer Cu 3.0
Thermochill PA120.3 with 3 x San Ace Fans at 5v

GPU Sub-loop
3 x EVGA Hydro Copper Full Cover water blocks in parrallel
Thermochill PA120.3 with 3 x San Ace Fans at 5v

RES/Pump:
Typhoon III with D5 Vario

3/8" ID, 1/2" OD Masterkleer tubing with Koolance QD's and BP Compression fittings.

Ok, so now we have the hardware listed, here is what I have observed so far and again this is in no way scientific.

The Good:

1 - Much better flow than my two MCP355's with XSPC tops, specifically on the CPU loop.

2 - Cleaner setup due to one res/pump combo.

3 - Temps on my GPU loop have gone down by a couple degrees.

4 - Looks sweet even with the stripe across the res.

5 - Less expensive than running two separate loops with separate res' and pumps.

The not so good:

1 - CPU temps didn't go down. This is probably due to the limitations of water cooling. I think I have hit my max without either doing a push/pull configuration or shrouds, both of which would be a tight fit.

2 - It's a B**ch to bleed. This is mainly due to me having such a large case and a lot of heavy equipment in the case that makes it hard to turn on it's back. Also, I think the remaining air bubbles are having baby bubbles since I can't seem to bleed it completely, close, but I still have a couple of bubbles.

3 - Fat base fittings require a new o-ring to make a proper seal.

So in conclusion, while my CPU temps did not improve my GPU temps did. This is probably due to the I7 being a very hot processor, however my GPU temps did improve probably due to the increase in Rad space being shared by the CPU loop that the CPU was not using. While there were a few things that I was disappointed in such as the white stripe, after I got it setup those things were easily forgotten by the overall performance of the unit. Remember, I am getting at least the same results and in the case of my GPU's, better results than I did with to separate loops. Up until now I would have called BS on this, but seeing is believing. I think the Typhoon III sets new standards in what we can expect from water cooling especially from a single loop divided into to two loops. Before doing this I would have stated that a dual loop would be the best option where isolating the CPU by itself would be ideal. After setting this up I am now convinced that is not the case and this single pump/res combo is equal too, if not slightly better than two individual loops. If I had to put a rating on the Typhoon III it would be:

Performance: 10 out of 10
Cosmetics: 7 out of 10 due to the white stripe and it not having LED holes.
Ease of setup: This is one is complicated since it cleans up the loop, but I found it a PITA to bleed compared to other pump/res combos I have used such as the XSPC res top, however I would still give it a 8.5 out of 10 in this category.

Overall: 9 out of 10 in my book because I value performance more than anything else and it definitely lives up to the performance expectation considering it is a single loop at heart and easily keeps up with a separate dual loop configuration.

So there you have it. Again, this is not scientific, nor will it necessarily reflect results from anyone else. I hope that this helps those that were sitting on the fence decide as I am very happy with my setup.

Thank you very much for taking the time to post.

Sorry that bleeding the loops is harder on your big rig. My goal was for it to be a VERY easy to use setup for novices and casual water coolers with typical cases. If it is easier then more guys will take make the jump into liquid cooling--more people liquid cooling is good for us for obvious reasons. Having never built a case that large/heavy I have no experience that direction. I do have a lot in a normal case size though and front fill ports will always be easier IMHO.

The fatter O ring is for the bitspower fittings you are using right? Just wanted to make it clear the supplied compression fittings work fine =)

Do you know the size of the O ring you used so I can add the information to the directions?

Again, I know you took a good deal of your time to write/share you impressions, thanks. Rigs killer btw...now I want to do a fat cuber lol.

But once it is bled it looks so good.

NICE!!!

Wait till you see the billet aluminum face plate machined from a solid 1/2" thick plate.

[Vapor]

Call the leg with no tube on it 9 o'clock. If the waters main path were from the 9 o'clock leg to the 3 o'clock leg with the 6 leg being just a fill port leg...well not so good obviously, Now if the 6 leg was feeding the 3 and 0 o'clock leg then you get very clean lower restriction flow. The better choice for a fill port type T would ba a pass through or straight T.

Yeah, I get that's not the fill port T, but is that the "high flow true T" or a third kind? Just never seen one like that before, hence I called it a Spread Eagle Y

[BoxGods]

Geno, ON another forum i have some one who has few problems. He has a Tj07 and it dos'nt fit, do you think you could head over to WCUK?

Not sure why he is having a problem...T3 is 100% to the ATX specification. Everyone is harping on the no side Chanel thing because the other new bay rez pump has them...and because bending the little tabs out of the way in say a Cosmos is SO hard lol.

I am going to try to help but I think the fact that WCUP puts out a competing product makes their forums...call it a no win situation for me.

[BoxGods]

Yeah, I get that's not the fill port T, but is that the "high flow true T" or a third kind? Just never seen one like that before, hence I called it a Spread Eagle Y

I am betting you just like saying "spread eagle"

The straight tee just doesn't have that central splitter that gives you the Y impression your getting I think.

[Vapor]

Yeah, I've just never really thought of using a T to split...pretty interesting idea, I can think of a few uses already

While I do like saying spread eagle, I first saw it and thought of a Y with its legs broken outward, but that was too long of a name

[bentleya]

Not sure why he is having a problem...T3 is 100% to the ATX specification. Everyone is harping on the no side Chanel thing because the other new bay rez pump has them...and because bending the little tabs out of the way in say a Cosmos is SO hard lol.

I am going to try to help but I think the fact that WCUP puts out a competing product makes their forums...call it a no win situation for me.

Thanks for taking a look Geno, the competing product, is in a different category with different way of doing things, which needs a few revisions. But anyone on there who knows what they are doing which is a few, but few know nothing who do posts on theses forums. Would buy the better product in this case the T3 .

But having to wait a long time and 3 lots of shipping changes ETC, a fair few of them are very emiconal about it.

[alacheesu]

I get what you're saying...but that's not the case.

If you took a loop with a CPU block, a radiator, and other stuff, it would have a pressure drop curve of y= AX^3 + BX^2 + CX + D (well, D=0). If you doubled the components in serial, you would have a pressure drop curve of y= 2AX^3 + 2BX^2 + 2CX + 2D. Basically, you'd double the resistance....

But if you took those same components and split the (doubled up) serial loop into two parallel loops....the overall restriction is much lower than even AX^3 + BX^2 + CX + D (how much so depends on the Y split and all associated with that).

I've been trying to figure this out before. How do you find the restriction over the parallel loops (from before the first Y to after the second), assuming no restriction in the Y's/tubing/etc? If X is flow rate, and f(X) equals your first "y" from above, would it be 2*f(1/2*X)?

[millertime359]

The fatter O ring is for the bitspower fittings you are using right? Just wanted to make it clear the supplied compression fittings work fine =)

Do you know the size of the O ring you used so I can add the information to the directions?

Only problem is you are limited to 2 sizes of tubing and he wasn't using one of those. I may be redoing my loop and seeing I already have compression fittings for the 7/16" 5/8", I hate to get rid of them. When I get this, I will be in the same boat as Utnorris in not being able to use the ghost fittings.

[Utnorris]

The o-ring size is #9, thickness is 3/32". Yeah if I did have so much already invested in fittings, specifically QD's, I would have switched tubing out and used the fittings that came with it. Speaking of the enclosed fittings, I do find it interesting that if you buy the res from Jab-tech you get a total of 8 fittings included, but if you buy it from Frozencpu or Primochill directly they only give you 4 fittings, at least that is what their website states.

As far as the bleeding goes, I was trying to emphasize that it was hard due to the size of my case and most will not have this issue. My GPU water blocks alone weigh about 6lbs each, so my case with components is much heavier than the usual case. For most folks, bleeding should not be an issue.

And again, I want state that I am impressed with the res and the performance and am happy with the results. For me, performance trumps the minor issues I ran into, but I did want to mention them, even though they were minor.

[Manicdan]

i hope this dosnt sound dumb, but can someone (skinnny) test the flowrate of the pump in 3 configurations:

1, with absolutely nothing attached, the pump by itself (and the flowrate meter)
2, with the T3, both loops combined, measuring total of both loops
3, with a Y on the inlet and outlet of the pump, measuring between the two Ys to mimic a parallel setup

whats shocking is how simple this is, its really nothing more than a pump with 2 inlets and outlets, and that we havnt seen pumps sold built to handle a parallel system. which btw is proven to really make a difference. i plan to get one for just a CPU, but i will still be able to run the cpu alone, and the radiators on the second loop. this kind of technology can make almost any setup work better for cheaper, and look better too. also it solves another issue for me, my custom case (which i still havnt taken pictures of) is 3 stacked towers, and theres only 6" of room where the pump sits, but with the way they built the D5, its really not a comfortable fit for the pump unless i lay it on its side. so it will help with fitting issues too.

once the tape issue is concluded, ill get one, and i promise to take photos of my case. (the worst part is i dont have a camera, but my sis needed me to fix hers, so its fixed and sitting right here in front of me, and im still too lazy to take it out, lol)