So would you be insulted if someone put it on a 8mm loop with a DDC? I am going to buy one from Swifty but don't want to insult you too. Worst case I will run a 10mm loop since it will slip over a 3/8" barb. I am very interested in the performance, if I can get the pump to push hard enough.Cathar said:
![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
Retail G4's coming soon....
- Thread starter CAD OC'er
- Start date
R1ckCa1n said:
Why would I be insulted? I even ran that exact same setup and test here at my end during development. It was an important design goal for me to get it to work well with a 5/16" (8mm) plumbed setup on something like an MCP350.
You're basically proposing to do something that I specifically designed for.
I think maybe you've got the wrong impression of me. I am not a blind-fath "big bore" proponent. Since I've been designing waterblocks and understanding how they work, I have always been a "moderate" proponent when it comes to tubing sizes and flow rates, meaning that I sit in the middle ground. My "playground" and design goals encompass 8mm ID to 1/2" ID, and for some time now (~12 months) I've been stating that 1/2" ID tubing is stupidly large and unnecessary with any well designed waterblock. 1/2" ID tubing is still required for some of the less efficiently designed blocks coming out of the USA.
I regularly test down to 1LPM flow rates and ensure that performance of the blocks that I make even at those flow rates are at least competitive if not superior to the best of other blocks at that level.
Yes, I design for performance, but that also includes understanding and designing for a wide range of tubing sizes and flow rates. 1/2" OD barbs just happens to be where the lowest tubing-based restriction is, and so the best possible performance exists with 1/2" OD barbs - but then again this is true of all waterblocks, not just mine. It is wrong to assume that the blocks that I design ever demand that they be used with big bore tubing.
Sorry for the minor rant R1ckCa1n, I just get a little tired of the misinformation and perception that exists out there that the blocks I make do not work well at low flow rates when I specifically design them to do so.
madmat
2[H]4U
- Joined
- Jun 3, 2005
- Messages
- 2,372
From everything I've seen, from test data on flow vs. performance to reading your thoughts over the last few weeks I think it would be accurate to say that you design your blocks to be friendly to just about any system, they provide exceptional performance at nearly all levels of flow.
Higher flows show more gains but even with blocks designed for "low flow" loops that's been pretty much the case so it's to be expected.
Higher flows show more gains but even with blocks designed for "low flow" loops that's been pretty much the case so it's to be expected.
Cathar said:
That would be great.
Now that you mention it, i have one, and just thought from what i read, that you didn't think this pump was very suitable to the task at hands, (good head pressure but only ~360/lph maximum flow).
At first i was going to use the alphacool acrylic top with the DDC, due to the central thread opening it has (and use 3/8" barbs) but then thought it it would be better to buy a pump that was capable of giving several flow options, hence the 655.
You know, when we see all the people talking about Iwakis and MCP600/650, asking what's the pump you would elect to be the perfect for this setup, and then comparing to the actual pump you use, one gets the impression that a stronger pump is (in this case) the driving force behind good performances, and in retrospect i suppose it has lead to a wrong conclusion in Europe (at least the way i see it) that you only care about high flow rates in order to get the performances you have with your blocks, and using anything inferior either in terms of tubing or pump power is sort of "downgrading" (watercoolplanet sure tried to prove it this way).
Even Swiftech isn't very clear on this aspect (when we read their indications about which tubing and pumps to use with the Storm).
Maybe it was way the "message" was delivered, and nowadays being Swiftech the company behind the marketing of your block, maybe now some european minds will convert and you get to see from this side of the ocean a well deserved recognition.
Thanks for you comments and sugestions Cathar.
madmat said:
...and that has really been what's been irking me about various comments that get thrown around about "low flow" blocks being superior and all that guff.
All a good "low-flow" block really is, is just an efficient block design that has small tubing barbs stuck to it. I could've stuck 1/4" ID tubing barbs on the Storm and called it a "low-flow" block too if I wanted to, because it is the design that is efficient at the flow rates that 1/4" ID tubing supports. Of course performance will be less than it could be when using 1/2" OD barbs, 1/2" ID tubing, and a strong pump, but that is true of every single waterblock. That's just the way that waterblocks work.
The difference is that I offerred 8mm -> 1/2" OD barbs for the blocks that I made, and nobody wanted 8mm OD, 5% wanted 3/8" OD, and 95% wanted 1/2" OD.
I really would have to question my own intelligence if I made the block in such a way then that the largest fittings it could support would allow for at best 8mm ID tubing or whatever, given that no one wanted that tubing size. Sadly, just because the block supports up to 1/2" ID tubing there seems to be the perception that this is only the tubing size that it can be effectively used with.
I do not have the wrong impression of you. If I remember our email exchange when I bought one of your first batches of G4's, you incouraged me to run a 3/8" loop as that was my goal at the time. I was just checkingCathar said:
So if I am getting 1.4 l/m with my existing loop (1046 running above 1048 pressure and a CuplexXT) it might work pretty damn well since the XT seems to be pretty restrictive. I will get the DDC with 8mm fittings just to see how much it helps the overall cooling. It will be a fun test!Cathar said:
Jag said:
Actually the mildly amusing thing about the way in which that was worded is that I was strongly arguing that it should be worded in a way that was more "low flow" friendly. I explicitly said that the block was a low-flow friendly block and that the European market should not be alienated by delivering a message that the block must be used with a strong high-flow pump and large-bore ID tubing.
The reason why that was worded the way it was has more to do the product positioning by Swiftech themselves within their greater product range, rather than what the block is actually capable of with respect to its lower-flow capabilities.
The statement you see at Swiftech is pegged back from a more extreme high-flow focused statement that originally existed, and it was only due to my argument that it even got as "low-flow-friendly" as it is.
Will respond to more of your post in a bit.
Ok...just from a n00b's POV, you are stating that 1/2" tubing is just a wee bit on the crazy side, and 3/8" is OK, but you'd be better served by a different tubing.
Is it your feeling thus that 7/16" tubing, coupled with 1/2" fittings is the way to go? The reason I ask is that I'm still in the middle of composing my system, and I want to insure that I go with what would be optimum, not on the retardedly large side of things.
-Ghent
Is it your feeling thus that 7/16" tubing, coupled with 1/2" fittings is the way to go? The reason I ask is that I'm still in the middle of composing my system, and I want to insure that I go with what would be optimum, not on the retardedly large side of things.
-Ghent
Jag said:
You should achieve ~4-4.5LPM flow rates with the MCP350, 3/8" OD barbs on a Storm, and 3/8" ID tubing and a radiator.
I've recommendd such a setup to more than a few people who bought the blocks, and as far as I'm aware from the feedback I got, all were extremely satisfied, if not rather suprised that it was substantially outperforming their old 1/2" ID setups.
The D5/655 is nice for it's adjustability, but if you're going to run it at its #3 position, then the DDC in stock form would give higher flow rates, and be a lot smaller, as well as being cheaper without having to buy custom tops for the D5, which is why I suggested the DDC.
I think this is perhaps a cultural difference. Somewhat strange but most "English"-based nations are all a pack of rev-heads with an "excess breeds success" attitude to cars, computers, motorbikes, planes, etc. Whatever it is, if you can drive it harder and faster, then they'll do it.
I think that the message gets lost in the fervor. Just because we're all chasing the utmost performance and eking out every last 0.1C from the system doesn't mean that the systems suck given a more moderate approach. It's just a sad reflection of the "if it's not the absolute best, it's junk" mentality, which gets projected too heavily and I would imagine that's the attitude that many Europeans see and are mistaking.
There is a strong following of the more moderate crowd in the "English" based nations that really do understand the European point of view and do cater for it, but having done so, they'll still stick a high-end pump onto the setup with big-bore ID tubing and still chase down that last 0.1C, just because they can.
I do still have serious issues with the way in which WCP conduct their testing, specifically with varied mounting pressures, and only using the results of a single mounting when a bad mount may vary by up to 5C from a good mount. If this is not controlled then who knows what the level of error is?
See above post.
You're welcome. I do try hard to design and cater for the European side of things, even if most of the users of my blocks are all the rev-heads who give the impression (to the Europeans) that the blocks MUST be used with strong pumps. Don't mind them though, they're just happy chasing the last 0.1C, which is something that I also enjoy catering for as well.
Me? I design for the whole gamut of flow rates, but having once done so, I then proceed to use a strong pump and chase the high end performance too, but that's just because I enjoy doing so.
Ghent915 said:
IMO, when chasing the highest performance (that last 0.1C), but without needing to go to the stupidly large 1/2" ID tubing with 1/8" wall thickness, the best tubing combination to use is:
7/16" ID tubing
3/32" wall thickness (i.e. 5/8" OD tubing size)
Stretch the tubing over your 1/2" OD barbs - requires minimal effort and does not even require the dipping of the tubing in hot water.
That is the largest "optimal" performance/space efficient tubing size to use in a set where a Storm block in installed, and there is no real point to using tubing with a larger ID.
cornelious0_0
[H]F Junkie
- Joined
- Apr 6, 2003
- Messages
- 12,783
That G5 looks amazing....some real quality there indeed.
mohammedtaha
2[H]4U
- Joined
- Oct 14, 2004
- Messages
- 3,648
I have a question.
As you decrease the tubing size, isn't the amount of heat that is transferred to the water per second a lot higher? What I'm saying is, as the tubing width decreases, the water, in contact with the copper, heats up faster due to the decreased amount of water change in the block thus transferring more heat than a wider tube per second to the rest of the water in the loopl. This in a sense is what the Storm block does right ? You're trying to remove as much heat as possible using that type of design but the problem is, as the tubes get smaller wouldn't the performance decrease rapidly if the cooling is not up to par and if the pump does not push enough water through.
I would say that the block is excellent as long as you can keep a high flow rate and cool the water. Otherwise it would cause more of a restriction that anything.
Correct me if I'm wrong ..
As you decrease the tubing size, isn't the amount of heat that is transferred to the water per second a lot higher? What I'm saying is, as the tubing width decreases, the water, in contact with the copper, heats up faster due to the decreased amount of water change in the block thus transferring more heat than a wider tube per second to the rest of the water in the loopl. This in a sense is what the Storm block does right ? You're trying to remove as much heat as possible using that type of design but the problem is, as the tubes get smaller wouldn't the performance decrease rapidly if the cooling is not up to par and if the pump does not push enough water through.
I would say that the block is excellent as long as you can keep a high flow rate and cool the water. Otherwise it would cause more of a restriction that anything.
Correct me if I'm wrong ..
mohammedtaha said:
Okay, first we need to get your terminology right. You seem to be confusing heat with temperature.
At full load when doing something consistent, the CPU is emitting a fixed amount of heat energy. The same amount of heat energy enters the water whatever the water flow rate is, or however efficient the waterblock is at transferring the heat energy into the water flow. What you probably meant is that the temperature of the water will increase more as the flow rate is decreased, and you are correct. However let's analyse the scale of that statement.
Let's say that the typical flow rate with a highish-flow block is 7.5LPM, and the typical flow rate with a Storm water block is 5LPM with some arbitrary pump in a full system loop.
Water holds 4186J/kg°C, meaning that it takes 4186 Joules of energy to raise 1kg of water by 1°C. Water, as far as we are concerned, has a density of 1.00, so 1kg of water is effectively 1 litre of water, so basically:
It takes 4186 Joules of energy to raise 1 litre of water by 1°C.
1 Watt = 1 Joule per second.
Now if our flow rate is 7.5 litres per minute, then we have 7.5 / 60 = 0.125 litres per second flowing through the waterblock, or a thermal capacity of the water of 0.125 x 4186 = 523 W/°C.
i.e. At a water flow rate of 7.5 liters per minute it would take 523W of heat load entering the water to raise the water temperature by 1°C as it passes through the water block.
Now if our CPU is emitting 120W of heat energy (very, very hot CPU), then the water is heating up by 120/279 = 0.23°C from inlet to outlet.
Now since the water enters at 0.23°C lower than it leaves, the average water temperature that comes into contact with the metal of the block is the inlet temperature plus half the rise from inlet to outlet. i.e. the temperature delta effect due to the water warming up in the waterblock is actually 0.115°C.
Now if with our Storm waterblock system the flow rate is 5LPM, then the average temperature delta rise from inlet to outlet is 0.173°C.
Therefore the effect of reducing flow rates from 7.5lpm to 5lpm on a hot overclocked full-load 120W CPU, with respect to the water getting warmer and cooling the block less as a result, is a difference of 0.058°C
i.e. it's ~0.06°C, or so small that it's not really going to be noticed.
What about an ultra-low-flow system I hear you ask?
If we reduce our flow rates to 0.75LPM, say, as exists in an ultra-low-flow system, then the difference is 1.035°C for our same 120W CPU from going from 7.5LPM down to 0.75LPM, or something which could be classified as significant.
Basically so long as flow rates are above 2LPM (0.55gpm) then the heat capacity of the water isn't really much of an issue, and even at lower flow rates it's only just starting to become significant. What is more important is that the waterblocks are designed to work really well at those flow rates.
All true, once you understand the significance of the statement as it applies to actual setups. It's not a significant factor for flow rates of 2LPM or higher.
Erasmus354
[H]F Junkie
- Joined
- Mar 12, 2004
- Messages
- 9,450
Thats exactly what I tried to tell these people a couple months ago Cathar I did much the same calculations assuming 2lpm flow rate, and showed that the difference between the inlet and outlet of the block was not very large at all
I did much the same calculations assuming 2lpm flow rate, and showed that the difference between the inlet and outlet of the block was not very large at all Yeah - I saw your post about that Erasmus354.
Also, in reviews at Procooling, the water is entering the waterblock at a fixed temperature, so the water-temperature-rise-per-flow effect is already part of the measured result.
The above calculations are useful if one wanted to establish what the base-line was without the water-temp-rise effect, and then re-apply as for a different coolant type being pumped through the block.
Also, in reviews at Procooling, the water is entering the waterblock at a fixed temperature, so the water-temperature-rise-per-flow effect is already part of the measured result.
The above calculations are useful if one wanted to establish what the base-line was without the water-temp-rise effect, and then re-apply as for a different coolant type being pumped through the block.
mohammedtaha
2[H]4U
- Joined
- Oct 14, 2004
- Messages
- 3,648
Cathar said:
Ok thanks. Good lesson on water cooling.
You can ALWAYS use 3/8" ID tubing on 1/2" OD barbs? Cause my whole (planned) system is gonna have 1/2" OD barbs (see link in sig). I had 1/2" ID tubing planned, but you have swayed me. A while ago I actually had 3/8" tubing, but someone told me to get 1/2" to get the best cooling. You are saying that I can safely get 3/8" tubing with minimal temperature sacrifices, and not just with the Storm block? I can't tell if all what you are saying only applies to the Storm block, because it is a little expensive for my system, so i am going with the DD TDX w/ extra nozzles.
Sry if that's a little repetitive, I just want it to be clear what I am asking.
Sry if that's a little repetitive, I just want it to be clear what I am asking.
DFI Daishi
2[H]4U
- Joined
- Feb 14, 2005
- Messages
- 2,839
heheheheh.....yeah, there are those on here far more knowledgeable than i, when it comes down to the details......Russ said:
i believe that the recurring point is that with a well designed block, although 1/2" ID really does outperform 3/8" overall, it's not a big enough difference to worry about.
oh well, i really didn't know how little of a difference it makes.
Russ said:
3/8" ID tubing starts to become a major source of restriction for most pumps at above 6LPM.
As I said I can't comment on other waterblock's performance, but it would perhaps be useful for you to look at some reviews of the TDX and see what's going on at >6LPM, and whether or not that extra performance would be useful to you, and if you think that it would be, then it would be worthwhile for you to design your system around that premise.
Cathar said:
Thanks for sharing your knowledge.
I'm planning my next system and I was thinking of adding my Northbridge and an SLI system to the cooling loop. I'm going to assume I'll need to make two loops and I was thinking about using the following:
CPU - Storm
NB - Twinplex Pro
http://www.aqua-computer.de/prodimg/nd_products/twinplex_pro_2_500.jpg
SLI - Aquagrati X800, twice
DDC with Alphacool top
Aquatube Rez
DDC / Rez combo ala MCP1000
PA160 or BIP or new 120 Rad you discussed, twice
3/8" OD for all except the Rad which will be 3/8" ID
I'm thinking of having the CPU and NB on one loop and the two vid cards on the other.
I'd love for this to go on one loop but I just love the DDC so much I'd hate to have to move to some other higher flow pump.
I'm a relative newb at WC (I've got 3 WC builds under my belt now) and I've never attempted this many components.
My questions is:
Does this sound reasonable?
Do you have any suggestions?
Why do you suggest 3/8 ID" for the Rad (assuming a Rad that has a 3/8" OD as an option)?
Can something like this be done with one loop and if so, what would the components need to be?
P.S. I know there is no current ATI based SLI but, as I said, this is my next system plan that is at least 4 months out.
Edit: Since no one has answered yet, I wanted to add that I have been considering going with a heatercore instead of a BIP but isn't a heatercore more "High Flow" than what my system is going to be?
NorthernYankee
n00b
- Joined
- Jun 6, 2002
- Messages
- 35
I ordered mine on the 3rd directly from Swiftech...I haven't heard anything from them yet as to when it will ship.
--NY
--NY
mohammedtaha
2[H]4U
- Joined
- Oct 14, 2004
- Messages
- 3,648
I ordered from Frozen. I hope I hear something by the end of the day.
mohammedtaha
2[H]4U
- Joined
- Oct 14, 2004
- Messages
- 3,648
halusha said:I emailed Frozen CPU and asked if they had the STORM blocks in yet and Mark's response was "no not yet".
oh hell Naw ..
Well I got an e-mail from Swiftech today stating that if you have ordered the Storm block directly from their site, it will not be shipped out to you until sometime next week as they are filling orders for retailers before individual purchases. Dangit!
mohammedtaha
2[H]4U
- Joined
- Oct 14, 2004
- Messages
- 3,648
woo hoo ..
My order has been shipped from FrozenCPU .. i'm expecting it next week sometime I hope
My order has been shipped from FrozenCPU .. i'm expecting it next week sometime I hope
LionKinG88
Limp Gawd
- Joined
- Sep 14, 2004
- Messages
- 336
Why haven't they shipped mine
NorthernYankee
n00b
- Joined
- Jun 6, 2002
- Messages
- 35
I just cancelled my direct order and phoned in one to Frozencpu. They said the blocks will be in tomorrow.
--NY
--NY
NorthernYankee
n00b
- Joined
- Jun 6, 2002
- Messages
- 35
Well on th ephone they siad they will be in stock tomorrow...but I just got an email that said it is packed and will be shipped out today...I might actually have it tomorrow. I can't wait
mohammedtaha
2[H]4U
- Joined
- Oct 14, 2004
- Messages
- 3,648
Sc0ttFern said:
The storm block is the same in both the Swiftech and the FrozenCPU page.
The one procooling reviewed is an older revision that was built by Cathar.