[Review] PrimoChill Typhoon III

[millertime359]

Yes, there is a billet aluminum top coming that will be available in brushed black and silver anno, as well as polished to a chrome finish.

Can the end user install those onto an existing T3? I would interested in them when they come out if you can.

[Manicdan]

if skinnee yelled at us to read his graphs right, we couldnt argue with it

[HoosTrax]

Can the end user install those onto an existing T3? I would interested in them when they come out if you can.

I second that .

Something like an interchangeable snap-on or stick-on faceplate would be nice.

[Maxspam]

I want a dual 655 version.
Like if you took the ek spinn
And added two 655's to it
That would be freaking AWESOME
OR OR OR
Paint the impellers on the pumps like a swirl or something
haha that would be siiiick

[millertime359]

I want a dual 655 version.

Just get 3 T3 and run six loops

[Vapor]

1. As others have said, adding the flow from each loop is not fair. By halving the amount of pressure drop of the one big serial loop to attain two loops, the restriction is less and flow rates must increase.

It's not halving of pressure drop, it's the harmonic mean divided by 2*sqrt(flowrate through the pump)...in most cases that's closer to thirding the harmonic mean of the pressure drops. In other words, the change in pressure drop when going from the sum of the individual pressure drops to the parallelization of the individual pressure drops is massive. And that is exactly what the "dual loop" bar represents....it's showing that when using the T3 like it's intended, the pump works at higher-than-double the flowrate. That translates into higher flowrates for the two subloops, which are also included in the graph (because, ultimately, flowrate through the components is what drives performance).

Take a D5 PQ curve and then draw (roughly) pressure drop curves that intersect at the serial configuration's GPM (DetroitAC and EK numbers) and then again at the T3's GPM. You'll see what a staggering difference that going parallel (and designing for it) can make in terms of what the pump is doing.

As a pump-top review, it would be unjust to the T3 to not show how the D5 is actually performing when paired with it.

2. I guess I will be the first to mention that the commingling of water within the pumps do not make this a "true" dual loop setup. The cooler water for say a CPU loop will be contaminated by the warmer water of the other loop averaging the water temperature. I really see the term "dual loop" thrown around a lot in this review, but from a temperature perspective, this is not true.

It's more than commingling, each subloop will have the same fluid temps leaving the pump+housing. While having one pump run two distinct loops is enticing, it is, as you pointed out, not possible have them thermally separated. The comparison (on all levels) of this review is to other single-loop configurations, even if it looks like a dual loop config and can perform similarly to a dual loop config (in terms of flowrate). If configured right (which isn't hard to do), the T3 + D5 can outperform almost any other single-pump D5/DDC configuration.


EDIT: here's a fun config....HK 3.0 in one subloop, 2x HWLabs GTX360 rads in the other subloop. With a stock D5, you'll get 1.35-1.4GPM. With a D5+EKv2 or D5+DetroitAC pump top, you'll get 1.45-1.5GPM through the config. With a DDC3.2 + XSPC Res top, you'll get 1.6-1.65GPM. With the T3+D5, you'll get ~2GPM through the HK 3.0 and ~1.85GPM through the radiators.

[sinticfury]

Can anyone help me run my loop? It is going to be a dual loop

Typh>rad1>cpu>nb>Typh
Typh>rad2>gpu>gpu>Typh

[jtf2]

Great review as always, thanks for keeping up the great work started by Martin.

PS - now if I can just find one in stock somewhere ................

i think many of us would like to see martins take on this.
it is amazing all the shilling going on and everyone jumping all over themselves to buy this after only 1 review.
we`ll see in a month or so how everyone is doin with their newfound manna of watercooling,and prove to everyone how great it is.

[Snyxxx]

As a pump-top review, it would be unjust to the T3 to not show how the D5 is actually performing when paired with it. It's more than commingling, each subloop will have the same fluid temps leaving the pump+housing. While having one pump run two distinct loops is enticing, it is, as you pointed out, not possible have them thermally separated. The comparison (on all levels) of this review is to other single-loop configurations, even if it looks like a dual loop config and can perform similarly to a dual loop config (in terms of flowrate). If configured right (which isn't hard to do), the T3 + D5 can outperform almost any other single-pump D5/DDC configuration.

Yes, but some people put a 120.4 on a CPU loop only to maximize the overclocking of the CPU while putting everything else on the other loop with whatever radiator since it does not matter. This is the point of dual loops. I see this review touting this as "dual loop", which it does have two loops, but not to the spirit of the phrase to get TWO different temperatures with the cooler one where it matters.

I have been playing around with various pumps (and in series) including the Aquastream and observing flow rates. I have determined that flow is not always what it is cracked up to be when monitoring temperatures.

I wish skinnee would have performed some Core temperature comparisons to see the bottom line - temperatures.

[shazza]

i think many of us would like to see martins take on this.
it is amazing all the shilling going on and everyone jumping all over themselves to buy this after only 1 review.
we`ll see in a month or so how everyone is doin with their newfound manna of watercooling,and prove to everyone how great it is.

I continue to be amazed. While we may have questions about why skinnee chose to present the data as he did, and it's fair to ask for testing in different configurations ... why in the world would people accuse him of "shilling?" Are we trying to run out one of the few people who is doing us a great service by actually doing some testing? There is nothing in skinnee's history that would indicate he would, IN ANY WAY, purposely slant a review to please a supplier/developer/inventor. Yes, he's excited by the performance of the product; I take that to be because he was surprised, and didn't expect the improved performance. I also find it insulting to say that we need martin's take on this. Of course, we all welcome his views, but why take a dig at skinnee?

It would be nice to see people respond to reviews with legitimate questions (and counter data if they have it), without resorting to personal attacks.

[sinticfury]

Poopi face noone halps moi....Oh hai ther shazza

[Vapor]

Yes, but some people put a 120.4 on a CPU loop only to maximize the overclocking of the CPU while putting everything else on the other loop with whatever radiator since it does not matter. This is the point of dual loops. I see this review touting this as "dual loop", which it does have two loops, but not to the spirit of the phrase to get TWO different temperatures with the cooler one where it matters.

For people who want to run two totally separate loops....this product is not for them, and no one is saying that it is

I continue to be amazed. While we may have questions about why skinnee chose to present the data as he did, and it's fair to ask for testing in different configurations ... why in the world would people accuse him of "shilling?" Are we trying to run out one of the few people who is doing us a great service by actually doing some testing? There is nothing in skinnee's history that would indicate he would, IN ANY WAY, purposely slant a review to please a supplier/developer/inventor. Yes, he's excited by the performance of the product; I take that to be because he was surprised, and didn't expect the improved performance. I also find it insulting to say that we need martin's take on this. Of course, we all welcome his views, but why take a dig at skinnee?

It would be nice to see people respond to reviews with legitimate questions (and counter data if they have it), without resorting to personal attacks.

Well said Shazza

[jtf2]

I continue to be amazed. While we may have questions about why skinnee chose to present the data as he did, and it's fair to ask for testing in different configurations ... why in the world would people accuse him of "shilling?" Are we trying to run out one of the few people who is doing us a great service by actually doing some testing? There is nothing in skinnee's history that would indicate he would, IN ANY WAY, purposely slant a review to please a supplier/developer/inventor. Yes, he's excited by the performance of the product; I take that to be because he was surprised, and didn't expect the improved performance. I also find it insulting to say that we need martin's take on this. Of course, we all welcome his views, but why take a dig at skinnee?

It would be nice to see people respond to reviews with legitimate questions (and counter data if they have it), without resorting to personal attacks.

wow a little touchy eh.run somone out far from it,
shilling not directed his way at all-you took it that way not much for me to say there
all i said was basically get them in peoples hands then we will see whats what.
i mentioned martins name as his name is well respected-as no doubt skinnee`s is too.
seems like you inferred a little too much in what i posted

[Hondacity]

man this is sooo tempting to try out...

Hesmelaugh!!! dude do you have this in your inventory?

[shazza]

wow a little touchy eh.run somone out far from it,
shilling not directed his way at all-you took it that way not much for me to say there
all i said was basically get them in peoples hands then we will see whats what.
i mentioned martins name as his name is well respected-as no doubt skinnee`s is too.
seems like you inferred a little too much in what i posted

While I don't see how else to interpret your original post, I'm happy to apologize if that's not the way you meant it. They are in people's hands, and I expect we'll see more comments/reviews in due time.

My reaction was based not just on this thread, but on the general "attack the messenger" attitude that seems to occur regularly when people (not just skinnee) attempt to post test results.

[Maxspam]

Well technically its a single loop.
You are just running your cpu loop and your gpu loop in parallel

[Utnorris]

If you go read the other thread I already commented on the T3 vs. my dual loop, here is my post:

http://www.xtremesystems.org/forums/...&postcount=311

I found my setup with the T3 to give me virtually the same temps that my previous TRUE dual loop gave me. Why? I believe it is because I will not get any better temps regardless of how much rad I throw at it since my PA120.3 is already handling the heat from the I7 just fine. I like the T3 setup because it made running the tubing easier, cleaner and simpler. It also costs less than a dual pump and dual res setup. Too me, what we need now is a block that can take advantage of the increased flow, such as the Koolance 350. I think then and only then will we see a benefit in terms of temperatures with this setup. This is a revolutionary product because it will make us rethink how we setup our loops and how manufactures design their blocks, or at least it should. We no longer need a RD30 to get incredible flow, not saying this is equal to a RD30, just that it's a lot closer than what we have had up until this point. Try the T3 out yourself, worst case is you sell it for a little less than what you paid for it, but at least then you know and you won't be speculating on the results.

[NaeKuh]

Can anyone help me run my loop? It is going to be a dual loop

Typh>rad1>cpu>nb>Typh
Typh>rad2>gpu>gpu>Typh

Typh > Cpu NB > Typh
Typh > everything else >Typh.

Yes i will be getting one... as soon as skinnee sends back one of my D5's.

However i intend to do it

Typh> GPU GPU GPU >Typh
Typh> Board MCR640 >Typh.

Im sorry but its gonna be a cold and wet day in hell b4 u guys get me off dual DDC-2 for my cpu loop.

[Utnorris]

Typh > Cpu > Typh
Typh > everything else >Typh.

Yes i will be getting one... as soon as skinnee sends back one of my D5's.

However i intend to do it

Typh> GPU GPU >Typh
Typh> Board MCR640 >Typh.

Im sorry but its gonna be a cold and wet day in hell b4 u guys get me off dual DDC-2 for my cpu loop.

Tried it, wasn't impressed. Sure, flow was better, but it still didn't lower the temps till I connected the chiller to it, but that was the chiller not the flow that lowered the temps.

[MomijiTMO]

I believe it is because I will not get any better temps regardless of how much rad I throw at it since my PA120.3 is already handling the heat from the I7 just fine.

OT to the MAX but finally someone speaks the truth. It's absolutely absurd how people are suggesting more than a triple for a i7 loop. Heck, you could throw an i7 on a swiffy + 2C and be done with. What I think people are finding is that i7 is hot and even on water can over heat. The Yorkfields spoiled us rotten and we were able to have sub 1000rpm fans on a triple with room to spare. Now we can't.

I don't see people suggesting more than a triple anymore but I still see people beating the heck out of the dead horse by saying that you will get like stellar temps if you go with a thicker rad over a pencil swiffy. That has already been debunked so get over it.

Now for some ON TOPIC stuf:banana::banana::banana:es:

Yes Shazza, I think the review is good and more importantly it is the first review out there by someone who I have read several reviews of without much complaining.

Going along with Utnorris, yes this could be viewed as a pseudo dual loop because we do not have enough heat to split hairs. Sure load it all on a single 120mm rad and you will see a difference but who is stupid [or curious ] enough to do that? There comes a point where you don't need to nit pick.

If I went from dual loops and suffered a 1-3C penalty by using this top and dropping a pump, I would do it. Sorry but that difference isn't anything to get your knickers in a knot. Who cares? If your cpu was just on the edge of throttling, then you really should go with something more hardcore, like phase.

As it stands, I find this product very intriguing and while I commend whoever built it, I just don't think it is very useful for most folks.

[Snyxxx]

EDIT: here's a fun config....HK 3.0 in one subloop, 2x HWLabs GTX360 rads in the other subloop. With a stock D5, you'll get 1.35-1.4GPM. With a D5+EKv2 or D5+DetroitAC pump top, you'll get 1.45-1.5GPM through the config. With a DDC3.2 + XSPC Res top, you'll get 1.6-1.65GPM. With the T3+D5, you'll get ~2GPM through the HK 3.0 and ~1.85GPM through the radiators.

Never heard of the term "harmonics" used in fluid mechanics before. If you are saying that pressure drop is a function of the square of velocity, then yes.

Did you measure each sub-loop flow with the FM-17 flow meter?

For people who want to run two totally separate loops....this product is not for them, and no one is saying that it is

I guess I am being picky about the usage of "dual loops" and "sub loops" which can easily be confused with "separate loops". Just like the XSPC Dual DDC reservoir, that is still a thermal single loop since the water mixes together at the reservoir.

[Vapor]

Did you measure each sub-loop flow with the FM-17 flow meter?

Don't have a T3, just using common sense and some (fairly) simple math

[NaeKuh]

so whose winning?

[BoxGods]

I guess I am being picky about the usage of "dual loops" and "sub loops" which can easily be confused with "separate loops". Just like the XSPC Dual DDC reservoir, that is still a thermal single loop since the water mixes together at the reservoir.

I will take the "hit" or blame for this one. You are absolutely correct that "Dual Loop" is strictly speaking incorrect. All I can offer in my defense is that my main goal for the T3 was getting new people into water cooling by offering a pump/reservoir that is exceptionally easy to set up and use. One that might just entice some smaller OEM system builders to offer a water cooled machine. "Dual loop" is easier to grasp in a general sense for a novice then serial loop. It honestly was not intended as some sort of marketing hyperbole or to mislead anyone. The reality is that I can't spell "parallel" without a spell checker (which AIM does not have) and when I was discussing the copy with Brian (from PrimoChill) we were on AIM. I have a horrible time with spelling (Valor anyone?)

I will ask Brian about changing the wording to "Parallel" loop but I have ZERO say in that choice as I sold them the design. I imagine changing any printed materials, sending new copy to all the estores listing the T3 and so on make a change problamatic at best.

Don't have a T3, just using common sense and some (fairly) simple math

I would be more then happy to send you one. I had no idea you also tested products for the lab or I would have sent out 2 to start with. Shoot me an email with your shipping info.

[BoxGods]

I just wanted to also say that I am going to be posting less from here on out as there are now more forums with T3 threads then I could ever hope to keep up with. It is after 3AM and here I am posting. The whole forum thing is hard for (some ) people from my generation, as in "old dude" and it's stressing me out lol.

That reminds me. A Year ago I would not have stepped into the forums here at XS on a dare as this place was just BRUTAL. I don't know what happened but now I like coming here a lot as the atmosphere has shifted from destruction to construction. The Admins or whoever have really turned it around and I would encourage any other product designers or company reps to give XS another try--not a walk in the park by any stretch )and really it shouldnt be) but you will get a fair shake.

I have a mod to finish for nVida to display at QuakeCon and then the ION modding contest mod and I am WAY behind. I will check back every few days to see who is winning the latest argument =) and to see if anyone is having the type of problem I can help with. (missing a plug, weird glue, a scratch or w/e).

On the subject of glue. I am sending the guy that was the MOST pissed about the bonding band -- Bei Fei-- test sample #4 of the new clear adhesive to beat on and see if we did better. Sample #5 is going to UTnorris so he can test out the optional Top/bottom port in his heavy weight UFO. PrimoChill is likely going to offer a version of the T3 with these extra port/s at their site only and for a slight premium to cover the extra plugs and labor building them. I will post more details as they become available. It would be great if you guys mention it anytime people post about needing the top ports for an over sized rig or w/e.

Glue test samples 1-3 are in accelerated chemical, UV, and vibration testing and I will know by mid week how they fared. I have already worked out using the new adhesive in post production so if it passes testing it will be in place right away. Possible in time for the next production batch which hits bonding next Friday. You will know as soon as I do.

Almost forgot. For the guys asking about the machined aluminum face plates, yes those will be sold as an add on that can easily be installed by those who already have T3's.

Any serious issues just email me.