![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
Double Impact Block
- Thread starter R1ckCa1n
- Start date
Yeah, same story, different thread. Funny part is the person flaming the most in this thread has one but won't install it..... Go figure.
Erasmus354
[H]F Junkie
- Joined
- Mar 12, 2004
- Messages
- 9,450
He was only flaming an obviously flawed testing methodology 
To question methods is one thing but to due it like a 14 year old is another thing (not you). I never trust reviews these days, I rely on my findings.
I just got my new coolermaster case and will give the block a run with a ddc pump, pa120.2 and 3\8 tubing very soon.
I just got my new coolermaster case and will give the block a run with a ddc pump, pa120.2 and 3\8 tubing very soon.
Uggh, I'm going as fast as I can, testing is a chore. And yes, it is a REAL system I test on. Have made a new discovery however. Some reading plz.
http://www.xtremesystems.org/forums/showthread.php?t=151781
http://www.xtremesystems.org/forums/showthread.php?t=160566
http://www.xtremesystems.org/forums/showthread.php?t=142998
http://www.xtremesystems.org/forums/showthread.php?t=157286
If I could get my hands on the double impact I would certainly test it. I now have some 'nik modded blocks', an enzotech block and I want to test the thermalright block.
waterlogged
Limp Gawd
- Joined
- Oct 11, 2007
- Messages
- 179
So why don't you call the Shark and see if he will send you one for test?
waterlogged
Limp Gawd
- Joined
- Oct 11, 2007
- Messages
- 179
ROFL
urbsnspices
[H]ard|Gawd
- Joined
- Mar 25, 2005
- Messages
- 1,120
I have never been a fan of plastic topped blocks. I guess it is good for showing off your block and checking on your sea-monkeys.
waterlogged
Limp Gawd
- Joined
- Oct 11, 2007
- Messages
- 179
What's a sea monkey? I think I saw a post some where else that the same plastic is used in aircraft windows so I don't see what the problem is. Have you had problems with Aqua Computer plastic?
They must be OK as Sharka sold out their Intel version in a matter of days. I am going to be dropping mine on a Q6600 today (lang ddc, 3/8" tubing, 120.3 rad).
Erasmus354
[H]F Junkie
- Joined
- Mar 12, 2004
- Messages
- 9,450
Sales != performance R1ck, I would think you would know that by now. Especially when there are enough people who watercool on looks and brand and are perfectly fine with losing a bit of performance here or there. Sharka was bound to sell out quick before any substantive testing could be done.
waterlogged
Limp Gawd
- Joined
- Oct 11, 2007
- Messages
- 179
Under what conditions are you looking for test results?
I just was looking at the AC site and I think people are getting confused about what the testing was showing. It is not showing CPU temps! What it is showing is the difference between the CPU temp and the water temp. Obviously the lower the difference the better the cooler is removing heat, right?
AC Forum in English
linky
Arcygenical
Fully [H]
- Joined
- Jun 10, 2005
- Messages
- 25,066
I just Don't understand that graph. It might be because it's 3:30 in the morning, but I see no sense in it, even if it is showing the delta between CPU and water temps.
First off, a lower delta between water and CPU temps mean more heat is being removed from the Loop, but it doesn't say much about the block. Additionally, the values posted vary by a CONSIDERABLE degree. If it takes around 120w of heat to raise the loop temperature by one degree celsius, The graph on that Pdf file *can't* be correct. There's like a 9k differential between the DI and the Gflow... It could be showing the delta between CPU temps and air temp... or it's just a really bad graph.
Secondly, Why is the TDX in second place? I've owned a TDX. It does NOT beat the fuzion.
Third, Kelvin, why Kelvin?
Because Kelvin is used for temperature differences.
It is not that Kelvin you are thinking of
Arcygenical
Fully [H]
- Joined
- Jun 10, 2005
- Messages
- 25,066
Because Kelvin is used for temperature differences.
It is not that Kelvin you are thinking of
I was unaware that there was more than one meaning behind Kelvin
Well I just got done building my new loop and have not finished installing the OS. Prelim temps (taken only from the BIOS as this point) are pretty damn good. Once my restore of all data is finished today I will run it through a couple of tests.
Loop (3/8" tubing with perfect fit barbs)
DDC w/ alphacool top -> AquaComputer 120.3 radiator -> AquaComputer Double Impact -> EK FC8800 block -> nikhsub1 custom clear res
Loop (3/8" tubing with perfect fit barbs)
DDC w/ alphacool top -> AquaComputer 120.3 radiator -> AquaComputer Double Impact -> EK FC8800 block -> nikhsub1 custom clear res
Now the question is can they get a good quantity in stock before the first review site releases the results (Sharka sent blocks to at least one US site for reviews).
waterlogged
Limp Gawd
- Joined
- Oct 11, 2007
- Messages
- 179
Arcygenical
Fully [H]
- Joined
- Jun 10, 2005
- Messages
- 25,066
Well I am about up and running with the new loop. Not sure if I am going to keep it this way (might separate the GPU from the CPU) but all is going well thus far. A quick check of Intel TAT, my Q6600 is showing a idle temp of 38c and have not got monitoring software installed for the GPU. This is within 2c of what my dual core 6600 was measuring with 8mm tubing but since we are talking apples and oranges on the CPU, not sure what this means. I have to redo the loop (need to relocate the res) and will see the difference in temps on a new mount.
Well the loop is up (for now) and according to coretemp, my quad core is giving readings that are interesting....
Core 1 - 38c
Core 2 - 40c
Core 3 - 35c
Core 4 - 34c
I am going to introduce a new res (right now I am using a nikhsub1 clear tube res) this weekend and will do a few remounts to see if the temps stay consistant. I almost wonder if a "bow" to the base would increase performance. Flow appears to take a good hit as the water is not smashing against the opposite side of the res like a fuzion did. The other difference is I also have a EK 8800 block in the loop too so I don't know what kind of hit the flow takes just from that block. I am thinking of placing a second DDC in series to see how that affects the loop.
Core 1 - 38c
Core 2 - 40c
Core 3 - 35c
Core 4 - 34c
I am going to introduce a new res (right now I am using a nikhsub1 clear tube res) this weekend and will do a few remounts to see if the temps stay consistant. I almost wonder if a "bow" to the base would increase performance. Flow appears to take a good hit as the water is not smashing against the opposite side of the res like a fuzion did. The other difference is I also have a EK 8800 block in the loop too so I don't know what kind of hit the flow takes just from that block. I am thinking of placing a second DDC in series to see how that affects the loop.
Arcygenical
Fully [H]
- Joined
- Jun 10, 2005
- Messages
- 25,066
These temps are idle with a stock Q6600 but am getting 54-58 (depending on which core) on load with TAT. Only one mount so not sure if this is the optimal mount or not. I plan to do several this week end when I install the new res. So far so good though...... Runs quite as hell.
Arcygenical
Fully [H]
- Joined
- Jun 10, 2005
- Messages
- 25,066
I'd remount... I'm loading at 58 with my chip at 3400 1.5v...
Nicepants42
Gawd
- Joined
- Aug 30, 2004
- Messages
- 772
Agreed, mine is loading at 55C at 3200 @ 1.41V under a Fusion.
Your EK is known to be fairly restrictive.
That said, the DI block IS restrictive because of its design, far more so than the Fuzion. Look at how its two impingement chambers work: instead of using the loop's initial velocity to smash the water against the HS (think direct impingement: Storm, Fuzion, Apogee, etc), they build up a pressure chamber straight off the hose and then use this pressure to "squeeze" the water through the jets- this is very reminiscent of the Fuzion GFX block. This chamber MUST be restrictive, or else it wouldn't succeed at building any pressure: no pressure, no impingement- the water would just bleed/seep through the jets and keep flowing linearly. One stage of this is restrictive enough, but the DI links two of these in a row. Fortunately, the impact is less pronounced in the second stage due to the loop already having lost a great deal of its power, but it is still affected.
waterlogged
Limp Gawd
- Joined
- Oct 11, 2007
- Messages
- 179
Do you have a DI block?
Arcygenical
Fully [H]
- Joined
- Jun 10, 2005
- Messages
- 25,066
No... I don't think so. But it makes decisions.
Care to explain your last sentence ? Doesn't compute
Arcygenical
Fully [H]
- Joined
- Jun 10, 2005
- Messages
- 25,066
No, I have a Fuzion. But I was raised by two engineers, one of whose degrees is in fluid mechanics. I know that doesn't qualify me per se, but it definitely gives me some insight.
I would be really interested in seeing results with a DI block and a Fuzion GFX in the same loop, but in parallel. They are both high performance (or have the potential to be- I haven't seen any definitive tests), but everyone seems put off by how restrictive they are. Running them side-by-side should offset this significantly.
Erasmus354
[H]F Junkie
- Joined
- Mar 12, 2004
- Messages
- 9,450
No, I have a Fuzion. But I was raised by two engineers, one of whose degrees is in fluid mechanics. I know that doesn't qualify me per se, but it definitely gives me some insight.
I would be really interested in seeing results with a DI block and a Fuzion GFX in the same loop, but in parallel. They are both high performance (or have the potential to be- I haven't seen any definitive tests), but everyone seems put off by how restrictive they are. Running them side-by-side should offset this significantly.
For someone who is trying to qualify himself in topics relating to fluid dynamics you have a very poor understanding of the serial/parallel relation in terms of watercooling.
Yes while putting the components in parallel will lessen the overall restriction to the loop, it will most likely put lower flow rates through each of the two components. The lessened restriction will not be enough to overcome the fact that the flow is now essentially split in half between the two separate branches.
Splitting components into parallel is best served for items that don't have drastic improvements with increased flow. Splitting those parts into parallel will increase the flow for the other components in the loop that are placed in series with it.
I understand your point, but I believe that both of these blocks are much more dependent on pressure than they are on flow rate (within reason, of course). Your 50/50 distribution is also based on the presumption that they are equally restrictive, which I don't think they are.
Edit:
I made my assumption before looking at charts for the blocks. I didn't realize how sensitive the GFX block was to reduced flow.
Chart at the bottom: http://sharkacomputers.com/dtekfuziongfx.html
I doubt the DI block is as sensitive, though I can't find a chart for it.
Edit2:
After thinking about it, I still don't think it's THAT big of a difference. I would expect the DI block to be more restrictive than the GFX block, probably comparable to a Storm G3 based on the jet size relative to the amount of water. IOW, most of the water would still be running through the GPU block. I would be surprised if this turned into more than a 15% temperature delta, which shouldn't mean more than 4C or so. The CPU would still be a fair bit hotter, unfortunately.
It is well known that AC does not build blocks that don't well under low flow........ I would say the block is engineered to work at best under 1gpm.
I'm not so sure about that. As Cathar pointed out over at XS (about this same block, no less!), too little flow will affect the jet speed.
I don't know where the cut-off is, but there is definitely a point where the flow rate is TOO low. Sub-1gpm may be enough, but personally, I'd try to stay in the 1-1.5 range.
Agreed but you must understand that AC equipment is made typically ran with mulitple blocks (ie. 2-6 blocks) per loop all driven off a Aquastream pump. With that in mind, I doupt that AC would care about flow above 1gpm when designing a block. That doesn't mean the block won't perform better at 1-1.5gpm, just got to put yourself in the mind of the designers.
What I find interesting about this block is why two cores are running so much cooler than the other two...... This has been consistant across mulitple mounts.
Did you lap the block or CPU? I think a few guys at XS said their blocks came in concave, so it might not be mating with the IHS fully.