Production Aquanaut Basic/Extreme - Ultra Low Profile CPU Block & Pump Mount Combo

[thelaughingman]

@Nouvolo I think this 2nd design is really great. When I see the fill port(s) on the side the immediate comparison that comes to mind is the Eisbaer Solo LT which many are familiar with. As you said this has the advantage of symmetry and more motherboard compatibility so that's even better. Though for some reason, and this is visual only, I prefer the fill port location of the previous design, that Barrow-esque just looks good to my eyes

 

[Nouvolo]

@Nouvolo I think this 2nd design is really great. When I see the fill port(s) on the side the immediate comparison that comes to mind is the Eisbaer Solo LT which many are familiar with. As you said this has the advantage of symmetry and more motherboard compatibility so that's even better. Though for some reason, and this is visual only, I prefer the fill port location of the previous design, that Barrow-esque just looks good to my eyes

I don't think 3 ports on one end is possible. Having checked Barrow's, don't think it supports the EK-Torque well, the outer port holes are right at the edge, and it is not symmetrical also (can hit very close to the ram sticks). and I haven't seen any photos of the Barrow with EK adapters (post some here if anybody can find those). I guess most people probably use Koolance elbow with it also. It should have some limitations also, similar to those I found when designing this update.

 

[youngdocument]

I don't think 3 ports on one end is possible. Having checked Barrow's, don't think it supports the EK-Torque well, the outer port holes are right at the edge, and it is not symmetrical also (can hit very close to the ram sticks). and I haven't seen any photos of the Barrow with EK adapters (post some here if anybody can find those). I guess most people probably use Koolance elbow with it also. It should have some limitations also, similar to those I found when designing this update.

View attachment 1159 View attachment 1160

Can confirm from a previous build that EK Torque 90 degree fittings and regular 10/16 STC fittings scratch and can rub off paint on the pump casing here.

 

[Nouvolo]

Can confirm from a previous build that EK Torque 90 degree fittings and regular 10/16 STC fittings scratch and can rub off paint on the pump casing here.

And Barrow's block seems to be a one piece acrylic, which means it is a reverse flow block. It manages to go a few mm shorter by saving using a thin metal sheet cover on the side (i.e. side ways screws instead of up/down screws), it also needs to go taller for enough room for screw threads. All of this means it is strictly reverse flow design.

 

[thelaughingman]

I don't think 3 ports on one end is possible. Having checked Barrow's, don't think it supports the EK-Torque well, the outer port holes are right at the edge, and it is not symmetrical also (can hit very close to the ram sticks). and I haven't seen any photos of the Barrow with EK adapters (post some here if anybody can find those). I guess most people probably use Koolance elbow with it also. It should have some limitations also, similar to those I found when designing this update.

View attachment 1159 View attachment 1160

yeah, I have no love/hate towards EK Torque, just don't use it all but it surely is very popular and support for them is only a positive thing.

The new design can have 2 open ports 1 on each side right? That way there's always 1 fill-port facing up no matter the orientation for AM4 socket. And buying 1 more stop fitting should not be a dealbreaker to the ones who are looking at this product anyway.

 

[Nouvolo]

yeah, I have no love/hate towards EK Torque, just don't use it all but it surely is very popular and support for them is only a positive thing.

The new design can have 2 open ports 1 on each side right? That way there's always 1 fill-port facing up no matter the orientation for AM4 socket. And buying 1 more stop fitting should not be a dealbreaker to the ones who are looking at this product anyway.

I will probably include the two extra stop fittings in the package. You never know some may complain the product is flaw out of the box.

 

[sos]

[2021-07-13] This iteration has 2 fill ports on either sides (in & out)

I know that it would be to much "awesome" in so compact design but is there a possibility that we can use that fill ports as regular intake/outtake?

 

[Allhopeforhumanity]

I know that it would be to much "awesome" in so compact design but is there a possibility that we can use that fill ports as regular intake/outtake?

You'd be totally bypassing the pump if you did, and would be better off just using a normal water block.

 

[Fitchew]

[2021-07-13] Alternative Design Aquanaut (with fill port)

This iteration has 2 fill ports on either sides (in & out). Fill port should always face up to avoid spilling (rotate the rig when filling). This config is quite symmetrical and should better fit the motherboards.

Let me know what you think.

View attachment 1153 View attachment 1154
View attachment 1156 View attachment 1158

This is pretty clever design. I like it!

 

[duynguyenle]

[2021-07-13] Alternative Design Aquanaut (with fill port)

This iteration has 2 fill ports on either sides (in & out). Fill port should always face up to avoid spilling (rotate the rig when filling). This config is quite symmetrical and should better fit the motherboards.

Let me know what you think.

One thing you must keep in mind is that for AMD CPUs, the fin direction have quite a large impact on cooling performance. The finstack should be running vertical (running parallel to the 'long side' of the AM4 mounting plate. In your latest render, it looks like the fins are running laterally (sideways aka parallel to the short side of the AM4 mounting pattern), which I think is not the most optimal (though depending on the fluid routing, you might actually have no choice in this matter.

Both the Optimus AM4 and the TechN AM4 waterblocks (I think two of the highest performing AM4 blocks on the market at the moment) use this arrangement, and Optimus heavily push this design in their advertising (for Intel I don't think it matter as much, because the die is symmetrical)

 

[evd]

So what is the measured real world impact of this fin direction thought?

 

[duynguyenle]

So what is the measured real world impact of this fin direction thought?

I don't find any tests of anyone doing A-B testing regarding fin orientation on the same exact block, EK's promotional materials simply says it doesn't matter, but 1usmus has tested all the high performance blocks (real-world): the TechN, Heatkiller and Optimus (verticalfins) and the EK Magnitude (tranverse fins) and the results are summarised here:

Review: Liquid Cooling with a TechN waterblock (Page 5)

While TechN is new to the waterblock market, it has already developed waterblocks not only for Intel and AMD processors, but also for graphics cards (particularly the Nvidia RTX 2080Ti). Usually it is...

www.guru3d.com

If you got an EK or Optimus block and want to do direct testing back to back with two different orientations, that would be very valuable data to us here. Too many variables to robustly say whether the vertical fin blocks are better due to the fin orientation itself, or just due to the blocks being inherently better designed, the only conclusion that can be determined based on this data is that all 3 blocks with vertical fin orientation performed better than the tranverse one, by up to about 2 degrees (TechN vs. EK Magnitude)

 

[Nouvolo]

I am careful not to sound like making excuse for Aquanaut (or future versions) performance. But there is not much difference between branded and aliexpress water block performances, forward or reverse flow, or whatever. (I think there was a video from LTT some years ago about that).

Update:

AliExpress Water Cooling DOESN'T Suck?!​

Personally, I consider 2-3°C as close to margin of error, anything over 5°C may be considered significant, but there is more to it, read further below:

One important information that many forgot to note when reviewing water blocks (Aq and others) is the ambient temperature. Especially at this time of the year, normally 23°C vs a summer of 40°C (in Western Pacific). That's over 15°C delta. One's room maybe particularly warm and cozy. Even seasoned YTubers often omit such info.

And things get even more complicated with the latest AMD Ryzen CPU and its PBO(1 or 2). You get all kinds of micro spike in the CPU clocks even at idle, but nothing shows up in the task list. Many complains high idle temps, and it's so hard to pinpoint what is the exact cause.

Not to mention many other factors e.g. CPU vs CPU+GPU loop, rad size, etc.

Without elaborating on these relevant details, it's impossible to respond to reviews and complaints about Aq.

 

[duynguyenle]

I am careful not to sound like making excuse for Aquanaut (or future versions) performance. But there is not much difference between branded and aliexpress water block performances, forward or reverse flow, or whatever. (I think there was a video from LTT some years ago about that).

Update:

AliExpress Water Cooling DOESN'T Suck?!​

Personally, I consider 2-3°C as close to margin of error, anything over 5°C may be considered significant, but there is more to it, read further below:

One important information that many forgot to note when reviewing water blocks (Aq and others) is the ambient temperature. Especially at this time of the year, normally 23°C vs a summer of 40°C (in Western Pacific). That's over 15°C delta. One's room maybe particularly warm and cozy. Even seasoned YTubers often omit such info.

And things get even more complicated with the latest AMD Ryzen CPU and its PBO(1 or 2). You get all kinds of micro spike in the CPU clocks even at idle, but nothing shows up in the task list. Many complains high idle temps, and it's so hard to pinpoint what is the exact cause.

Not to mention many other factors e.g. CPU vs CPU+GPU loop, rad size, etc.

Without elaborating on these relevant details, it's impossible to respond to reviews and complaints about Aq.

No complaints from me! Given the packaging and volume constraints, the first Aquanaut is a good implementation of the concept, with all the tradeoffs being neccessary to get the pump down to that size. All I'm suggesting is that while you've still got the design in early stages, it would be good to consider all possible variables so you can get the best balance of compromises for the final product.

 

[MozzieT]

One thing you must keep in mind is that for AMD CPUs, the fin direction have quite a large impact on cooling performance. The finstack should be running vertical (running parallel to the 'long side' of the AM4 mounting plate. In your latest render, it looks like the fins are running laterally (sideways aka parallel to the short side of the AM4 mounting pattern), which I think is not the most optimal (though depending on the fluid routing, you might actually have no choice in this matter.

Both the Optimus AM4 and the TechN AM4 waterblocks (I think two of the highest performing AM4 blocks on the market at the moment) use this arrangement, and Optimus heavily push this design in their advertising (for Intel I don't think it matter as much, because the die is symmetrical)

I would imagine the coldplate could simply be rotated 90 degrees? seems doable on the existing design, unsure if the new design would retain that.

 

[Nouvolo]

I would imagine the coldplate could simply be rotated 90 degrees? seems doable on the existing design, unsure if the new design would retain that.

Don't think its possible to rotate. Aq's rectangular shape only has reflections symmetry but not rotational symmetry. i.e. fin stack direction left/right (or up/down) interchangeable, but cannot rotate from left/right to up/down. The fin stack should point towards the main ports (non fill ports) in current design.

 

[MozzieT]

Don't think its possible to rotate. Aq's rectangular shape only has reflections symmetry but not rotational symmetry. i.e. fin stack direction left/right (or up/down) interchangeable, but cannot rotate from left/right to up/down. The fin stack should point towards the main ports (non fill ports) in current design.

My mistake. The images look oh so close to being perfectly square and same goes for the mounting holes, but measuring carefully it is ever so slightly rectangular. Is this something that could be tweaked in future designs? However small the difference in real world performance, it seems like an easy tweak to make if you are going to be getting the new batch specially made?

I may have missed some of the context in previous pages but is there a targeted ETA for a new version? (assuming all goes well with MOQ etc.)

 

[Nouvolo]

No complaints from me! Given the packaging and volume constraints, the first Aquanaut is a good implementation of the concept, with all the tradeoffs being neccessary to get the pump down to that size. All I'm suggesting is that while you've still got the design in early stages, it would be good to consider all possible variables so you can get the best balance of compromises for the final product.

No definite time frame yet, but should be in 2021 (this year). Still many logistics to solve...

 

[eedev]

Looks who's speaking about Aquanaut (briefly) !

 

[Nouvolo]

My mistake. The images look oh so close to being perfectly square and same goes for the mounting holes, but measuring carefully it is ever so slightly rectangular. Is this something that could be tweaked in future designs? However small the difference in real world performance, it seems like an easy tweak to make if you are going to be getting the new batch specially made?

I may have missed some of the context in previous pages but is there a targeted ETA for a new version? (assuming all goes well with MOQ etc.)

Just been calculating the BOM and other costs. The new version will cost about USD120 + shipping, let me know what you think.