Putting It All Together
Constructing the Loop
In putting together the components of the Alphacool NexXxos Cool Answer 360 kit, there are several decisions that need to be made with some performance-related trade-offs to be considered. By default, Alphacool recommends constructing the loop as follows:
The trade-offs inherent with the above coolant path is temperature-related. Specifically, you are dumping heat from the pump directly into the coolant and the CPU block as a result. The amount of heat transfer here is minimal, but it is there. An alternate approach (and one I prefer when building loops) is the following layout:
The above coolant path avoids the temperature-related challenges identified in the first path because the coolant is actively cooled by the radiator prior to it entering the CPU block. However, this new coolant path has trade-offs of its own, more along the lines of restriction. Going through any radiator causes flow restriction, leading to reduced coolant throughput of any down-stream items, which includes the CPU block in this path. This could lead to slightly lower performance because the less coolant passes through the block over time in comparison with that in the original path. However, this flow reduction is minimal and is outweighed by the lower coolant temperature flowing into the CPU block.
As shown, the Alphacool kit was constructed with the coolant path setup as follows:
Prior to building the coolant loop, all parts were placed in the test bench and tubing carefully measured before cutting to size. The three fans included in the kit were installed to the radiator in a push configuration, where the fans suck air in and push it through the radiator. The alternate configuration, the pull configuration, orients the fans to pull fresh air through the radiator and expel it through the fan. There is minimal performance difference between the two configurations, the choice coming down more to personal preference and radiator fit and orientation in the case itself. The third fan configuration would be a push/pull configuration where fans are mounted to both sides of the radiator, pulling fresh air into the radiator as well as pulling it through and expelling air out. However, this type of configuration would require a total of six fans.
Tubing is fixed to the inlet (lower) and outlet (upper) rear ports of the reservoir using the kit's compression fittings. The front port is left intact to act as a fill / drain port. In testing, it was found that the reservoir needed to be filled to a minimum of 75% (using the front measuring line) so that the pump inlet inside the reservoir remained fully submerged and was not able to suck in air while powered.
The three fans included in the kit were fixed to the top of the radiator, oriented to pull air into the fan and push it through the body and fins of the radiator. The kit's hex headed screws were used to fix the fans in place. Compression fittings were installed in the upper inlet / outlet ports as shown. The fans were oriented on the radiator so that the Alphacool corporate logo on the lower side side of the radiator shows upright when flat on its back.
The kit's compression fittings were also used to attach the tubing to the NexXxos XP3 Light CPU block.
Intel Z170 board mount
The mount kit takes up virtually no space when installed to the Intel Z170 motherboard shown. The long hex-headed bolts feed through the bottom of the board, held in place with a white plastic washer and nut to secure it in place. While this mounting mechanism requires board removal to install (depending on the case used), it dramatically minimizes the mounting kit's footprint on the board's top surface. The plastic washers protect the board's surface from damage resulting from tightening the nut as well as acting to electrically isolate the block and its hardware from the board.
When mounted to a Z170-style motherboard, the NexXxos XP3 Light CPU block fits perfectly into the allocated space with no tight areas along any sides of the block. You should have absolutely no problem mounting this block on any LGA115X-style board. The minimalistic style of the mounting bracket (as well as its thin design) ensures that it will not get caught on any close proximity components, with its height more than adequate to ensure this as well. Because of the block's rectangular design, it can only be mounted in a vertical orientation (as shown) with the inlet / outlet ports parallel to the memory slots.
The block mounts to the CPU though hex-headed bolts that feed through the mount holes on the underside of the board. A plastic washer sits in between the bolt head and underside of the board to protect the board's surface from the bolt head. The plastic washer also acts to electrically isolate the block and its hardware from the board.
Intel X99 board mount
The mount kit uprights screw directly into the threaded mount points in the socket assembly on the top of the Intel X99 board. Like the Z170-based mounting mechanism, the pole-based system takes up little space in the board's CPU socket area.
The NexXxos XP3 Light CPU block fits well to the Intel X99 motherboard used for testing with no contact issues between itself and the memory or the VRM heat sink. Further, all DIMM slots (both in the upper and lower sets) remain full usable with the block mounted. The block mounts to the board in a similar fashion to that seen on the Intel Z170 board with the only difference being the upright mounting method. The uprights thread directly into the Intel X99 CPU socket, instead of requiring a separate underside baseplate. Because of the block's rectangular design, it can only be mounted in a vertical orientation (as shown) with the inlet / outlet ports parallel to the memory slots.
Talk about not understanding
Talk about not understanding your market. A 2×5.25″ bay reservoir is a little bit passe these days no?
It would seem so, most new
It would seem so, most new enthusiast cases don’t even have 5.25 bays anymore.
maury, would you please
maury, would you please consider doing a review of a unit using quick connect couplings?
they would be the only thing that might, depending on your review, get me back to water cooling, which i gave up five years ago because of the hassle and the noise ironically
i have been using noctua’s since, but really am curious whether it would be relatively easy and hassle free with the quick connects?
i suspect there is a flow problem with them, but maybe some company has sorted out that issue
I use quick connects on my
I use quick connects on my test benched when doing the stand alone block reviews and have found that they do not impact performance. It really depends on which quick connects you use and what fittings / tubing you connect to them. I’ve had really good luck with the Alphacool HF quick connects and the Koolance quick connects as well. They do get pricey though…
maybe you can do a video of
maybe you can do a video of building a water cooling loop with quick connects for novices?
Perhaps passe a bit. However
Perhaps passe a bit. However keep in mind that kit with old blocks and old bay res are very old by now. It belong to the old Alphacool before they basically sack everybody and hired new people with new ideas.
I would agree that old XP3 block is very high restriction block and it can’t compete with top of the line performers like EK Supremacy, BP Summit or WC Heatkiller. But I find absolutely astounding that 1200 rpm e-loops can’t deal with heat on 60mm 9FPI (!)rad. That does not compute. I’m using exclusively Alphacool rads U and X-flows and on one PC I’m running Scythe 500 rpm fans in push only on 60mm 480 with single AC D5 and I have no issues whatsoever. e-Loops are superb fans even at sub 800rpm they are perfect for low FPI rads. I don’t know what’s wrong with the setup but it simply isn’t possible for 1200 rpm bionic fans to fail at removing heat from 9 FPI rad.
Heck I’m using e-Loops in push @750rpms on 16FPI X-Flows and they are perfect in every way. Either pump is crap or its crap. 😀
I think the performance issue
I think the performance issue seen in testing was more with the low flow of the pump in combination with the thickness of the radiator, made for a "perfect storm" of sorts…
Strange, that looks like the
Strange, that looks like the standard D5 VPP655 pump. Did you get a broken pump or something? Looking at your picture of the back of the res and the D5, it sure looks like their standard VPP655, a 1500 l/h pump. It should go to like 4800 rpm at the highest setting.
Datasheet for the D5 VPP655: https://www.alphacool.com/AtsdDataSheet/getDataSheet/articleId/10620
Product page:
https://www.alphacool.com/shop/sets-und-systems/alphacool-nexxxos/16245/alphacool-nexxxos-cool-answer-360-d5/ut-set
Edit: Looks like when I link directly to their English site, it changes to their German one in an incognito window. Needs to set a cookie or something by clicking the UK flag upper right corner.
The odd thing is that their
The odd thing is that their page lists two different max flow rates, depending on which page you look at. The Description page shows a flow of 350 l/h, while the tech spec page lists it at 1400 l/h. In any case, something was severely impacting the performance of the kit.
Found on this page on alphacool site:
https://www.alphacool.com/shop/pumps/alphacool-pumps/10620/alphacool-vpp655-single-edition
The 350 figure must be a typo
The 350 figure must be a typo in the description, they have miss-spelled words too here and there. Also mixing English/German in a chaotic manner? Sheesh, when will they fix their site… Anyway, the “Technische Details” tab say 1500, same in the datasheet.
This datasheet has both the VPP655 and the VPP644:
http://www.aquatuning.de/download/Alphacool-Laing-VPP655%20-%20TPP644-Datasheet.pdf
This one seems to be for the PWM controlled version:
http://www.aquatuning.de/download/Alphacool-Laing-VPP655%20-PWM-Datasheet.pdf
I don’t doubt that, performance was surely sub-par. That 90 degree bend at the back of the res that you mentioned on the podcast probably doesn’t help. And that XP3 block looks like a highly restrictive design, probably needs the pump at max rpm. Might work better with one of those high pressure DDCs like Swiftech MCP35X or MCP50X. I have a MCP35X and it can power through pretty much everything. Drawback is the DDC whine at high RPM, I’d take a D5 over that any day. All DDC pumps I’ve used have had that high pitched whine when over ~3000 RPM. A D5 is heaven in comparison, even at max RPM I can’t hear it outside the case.
Thanks for doing some custom loop stuff reviews, I appreciate it 🙂
Thanks for the additional
Thanks for the additional information Pholostan. I've updated some information in the review to better elaborate on the kit's performance mysteries as well as updating the pump flow rate numbers…