Completed Salvo Studios S402: ~9L Steel Console Style Case

[Thehack]

I forget: would the flex slot/3-slot card design still accommodate 2.5" drives? I currently am making good use of spots on the S401, maybe wouldn't use them in a future build, so I'm not advocating one way or the other, just curious about their status.

This isn't an original thought but I can't remember where I first saw it (credit where credit is due). Someone once pointed out that there are a few "classes" of air cooler height (and I know there are way more coolers than I can think of on the fly here; these are just a smattering of popular ones). Up to around 40mm covers Intel stock cooler, Noctua NH-L9, and Wraith Stealth (I think?). Up to 50mm covers C7 and Black Ridge (and similar). Not too many coolers come into play at 60mm except the Big Shuriken 2. Then, at 70-80mm, we start getting the Noctua NH-L12 class coolers. So, a proposed cooler height of 59mm is good but still locks down from the highest end down-draft coolers (which is fine - this is just an observation). I would be able to go back to my old RVZ02 configuration of C7 heatsink with NF-A9 (25mm thickness) if memory serves correctly. That could be an effective combo for the proposed S402 width/cooler height.

Thinking of the RVZ02 again reminds me it can only support a 2-slot card and had about the same amount of usable cooler height. I think someone here already established it would be smaller than the Node 202 as well (and much breezier). On a different sort of scale, you could probably knock out both of those cases: smaller than both, matching or better component support (with the proposed 3-slot design), and better ventilation.

I wonder if 3-slot support would also help clarify the difference between the S402 and the Sentry 2.0. I'm cautious here since I don't want to come off as a Sentry-bashing Salvo fan. I am a Salvo fan but I am also extremely pleased to see the dent in mindshare that Sentry has accomplished with respect to premium SFF cases. This is why I think a clarifying difference is helpful - both cases can succeed in their own ways. At present, the S401 has similar component support to the Sentry 2.0 while being slightly larger in most dimensions (correct me if I'm wrong). I think S401 can fit more 2.5" drives and I anticipate the horizontal config is slightly more efficient with air coolers. However, primary component support is similar (~50mm coolers and 2-slot video cards). The S402 will still exist in the same sort of market space. 3-slot support and increased CPU cooler support may increase the S402's desirability when considered in a wider context.

Still, at the end of the day, that number followed by the big L is an absolute measurement. Personally, I come down more on the side of the 3-slot cooler design and resulting increased width. Of course, it will be helpful to see those renders comparing existing S401 dimensions.

Finally, congrats on the sale of the i9 S401 build. I hope it goes to a good home.

Yes. That was my comment. There are kind of magical numbers for coolers. Here they are:

Up to 40mm. This is about up to where the top of the IO cover sits. AR05, NH-L9, ITX30, IS-40.

Up to 48mm. This is based on intel stock cooler specs. C7, Black ridge, AXP-100 a couple unknown ones.

Up to 60mm. This is pretty much the smallest size for 120mm coolers that sits over the RAM. Big Shuriken 2b, Samuel 17, REEVEN Brontes, IS-50/60, AR06, AXP-200, Deepcool Grabiel.

And above that is pretty varied.

60mm is the sweet spot in my opinion. It gives a few decent choices and that is when the cooler can use 120mm fans. It is where your choices really open up. 55mm is a weird spot cause you can't use the 120mm coolers.

 

[vanbeveren]

I’d be very surprised if the s401 doesn’t cool a lot better than the Sentry

 

[batterybull]

I would love to see a four slot option to open up the possibility for bifurcation as well as for the Morpheus II, not to mention bigger cpu coolers. Hopefully with better internal filter compatibility too? Though I understand if this is not the direction you’re going in.

The rounded square and the hex vent patterns both look great!

 

[Talyrius]

Feedback I need: Please reference by number in a response to keep it simple unless you have suggestions.

1. Vent Patterns - Please ignore the square vent pattern on top.
Rounded Square

Circular

Slots

Smaller Hexagons (these are 3/8" diameter where S401 was 1/2")

Others? Show a picture of what you like. I'm not going to steal any custom design from someone and I would honestly prefer a simple pattern that looks functional/practical.

2. USB Front Panel. Type C support or no?
3. 60x25mm upper fan support worth .3L of volume?
4. 2.5/3-slot GPU support?
5. 59mm CPU cooler support?
6. #4 and #5 not important? then would it make sense to keep the dimensions of the S401 and make this a simple revision?

Any feedback is useful for me as I have no way to get this feeback otherwise.

1. Rounded Square.
2. Yes, USB Type-C support is preferred.
3. Nope.
4. I wouldn't use one of these thicker GPU designs myself, but I'm curious to see what else becomes possible with the space the support for them affords. PCIe bifurcation would be amazing.
5. 2.5/3-slot GPU support permits it, so yes.
6. N/A

 

[grsychckn]

So I'm on the plane to my conference this week and because the internet on the plane is 28.8k, I'm waiting to download Discord so I can log in there instead of Remote Desktop back to my home PC. Regardless, I've been thinking about the Flex slot and while I love the idea, I'm not so sure it's going to be the best option for the S402. The main problem is due to the flanges that are bent around the edges so that the vented panel will have something to fasten to. I realized while I was doodling that to use the Flex slot, I'd have to reduce the supported GPU size in order to clear those flanges when installing. What I have decided to try instead is actually a throwback to the S400 with the PCIe angled bracket so the screws mount backwards through the PCIe tabs on the device into the bracket and then the bracket is fastened to the chassis. A couple things here I believe make this design a better choice:

1. The area below the GPU does not have to be static because I can provide different mounting brackets for things like radiators, fans, and disks.
2. The PCIe device(s) will be more secure as the fasteners for the device will be extended farther out onto the chassis (like a fulcrum).
3. If the PCIe mounts get stripped, I only have to replace the bracket and not the chassis.
4. The bend required to make the PCIe tab is hard for my fabricator and then is compounded by requiring him to tap the holes on 0.04" material very near the bend. It's a challenge for him and the easier I make it the less expensive it is.

Pretty good for the first update of the day.

 

[vanbeveren]

Please make sure a new model, especially a bigger one, can accommodate 60x25 fans up top.
There are so many more options - better and quieter ones - in the 25 vs 15 thickness

 

[annasoh323]

Yes. That was my comment. There are kind of magical numbers for coolers. Here they are:

Up to 40mm. This is about up to where the top of the IO cover sits. AR05, NH-L9, ITX30, IS-40.

Up to 48mm. This is based on intel stock cooler specs. C7, Black ridge, AXP-100 a couple unknown ones.

Up to 60mm. This is pretty much the smallest size for 120mm coolers that sits over the RAM. Big Shuriken 2b, Samuel 17, REEVEN Brontes, IS-50/60, AR06, AXP-200, Deepcool Grabiel.

And above that is pretty varied.

60mm is the sweet spot in my opinion. It gives a few decent choices and that is when the cooler can use 120mm fans. It is where your choices really open up. 55mm is a weird spot cause you can't use the 120mm coolers.

Thanks for the extra info @Thehack - I had a feeling it was you. All of this thinking about CPU coolers reminds me I should check for dust... I've been running without the dust filter for a couple of months now...

 

[grsychckn]

I forget: would the flex slot/3-slot card design still accommodate 2.5" drives?

Yes, it would accommodate 3 x 2.5" drives. If I drop the flex slot I can fit 4 x 2.5" drive mounts in again. On a similar note, I'm thinking of making a bracket to mount 2 x 2.5" drives underneath the PSU (because of the added width).

Please make sure a new model, especially a bigger one, can accommodate 60x25 fans up top.
There are so many more options - better and quieter ones - in the 25 vs 15 thickness

That was my primary plan until I realized I could upgrade the fans to 70mm. Unfortunately, there are still few options in the 70mm size, but they are going to flow more air for less noise as a result. Just going from 60mm to 70mm yields ~36% more area for the fans. I'm thinking I'll still keep them limited to 15mm thick though as to support 25mm thick fans the volume of the case has to increase by .3L. This is my backup plan (60x25) if I decide not to support the 70mm fans. I'm working on ordering some now to test.

 

[jakemagee]

Feedback I need: Please reference by number in a response to keep it simple unless you have suggestions.

1. I think the hex, slots, and rounded squares look nice (in that order)... the circles visually appear like they might restrict air flow more than the other designs.
2. If there is space, I would like to see front USB support. And as for USB A or C, while I think A is more common still, I feel like modern cases such as this should be pushing modern features (surely USB C isn't going away anytime soon).
3. The "height" of the case is far less important to me than the "width"... so adding volume for upper fans would be fine with me (as long as it supported 25mm fan widths).
4. I would go with a full 3 slot over a 2.5
5. For the increased "width", I would go ahead and try to support CPU coolers with 120mm fans (60mm or whatever it takes).
6. I prefer more of an S401 revision (2-slot GPU)

 

[Linero]

So, lately I have been doing some research here and there about the ideal pattern for the case's perforation. I found out that the hexagon design is by far the worst, compared to modern designs, in terms of noise levels. It's also not as impressive with airflow as people are claiming for it to be according to a well respected SFF case designer. (Source) Also, here is an article with more information.

A fellow member also linked a similar article here: https://www.pugetsystems.com/labs/articles/Effects-of-Grill-Patterns-on-Fan-Performance-Noise-107/. Firstly, how accurate will these results apply to our proposed circle design compared to their mesh? It's possible that the increased diameter could influence different results. Secondly, mesh has one of the best noise level results. Intake airflow % actually beats the honeycomb counterpart using high fan speeds, which is arguably more important here. It does not lag too far behind with low fan speeds too.

I am not advocating the circles design, yet. I could be horribly off the mark and mistaken here and that I am missing something that will make me look like a total fool, as I am not well versed about such a high level topic. So once again, take all this with a pinch of salt. Nonetheless, I hope this was helpful as always.

 

[Thehack]

So, lately I have been doing some research here and there about the ideal pattern for the case's perforation. I found out that the hexagon design is by far the worst, compared to modern designs, in terms of noise levels. It's also not as impressive with airflow as people are claiming for it to be according to a well respected SFF case designer. (Source) Also, here is an article with more information.

A fellow member also linked a similar article here: https://www.pugetsystems.com/labs/articles/Effects-of-Grill-Patterns-on-Fan-Performance-Noise-107/. Firstly, how accurate will these results apply to our proposed circle design compared to their mesh? It's possible that the increased diameter could influence different results. Secondly, mesh has one of the best noise level results. Intake airflow % actually beats the honeycomb counterpart using high fan speeds, which is arguably more important here. It does not lag too far behind with low fan speeds too.

I am not advocating the circles design, yet. I could be horribly off the mark and mistaken here and that I am missing something that will make me look like a total fool, as I am not well versed about such a high level topic. So once again, take all this with a pinch of salt. Nonetheless, I hope this was helpful as always.

Lazer3d is not a good source of information for this test.

1. He uses acrylic which requires a fairly large separation between honeycomb cutouts. This a steel chassis using punch manufacturering. The separation between cutouts can be very small.

2. The acrylic material he uses is either 3mm or 5mm. This means that the material is about 2.5x thicker than steel. The thicker material increases turbulence and noise. It especially increases noise when you have sharp angles, like honeycomb or triangular cutouts. Circular cutouts do not have this problem as they are uniform. The material thickness increases turbulence when you have small perforations, so if you have a thick material, use larger perforations.

The puget system article is much more accurate and reflective of this chassis. I would also be willing to weigh the puget article more highly since they deal with business to business system integration. Lazer3d is a small operation of custom acrylic consumer chassis.

Let's not overthink it. The greater the perforations % the better. The closer the chassis is to an open case the better it'll perform.

Personally I like the rounded squares. All of these options will do superbly compared to the Dancase or Ghost. I do not think there is large difference between the patterns in terms of performance.

 

[SFF EOL]

So, lately I have been doing some research here and there about the ideal pattern for the case's perforation. I found out that the hexagon design is by far the worst, compared to modern designs, in terms of noise levels. It's also not as impressive with airflow as people are claiming for it to be according to a well respected SFF case designer. (Source) Also, here is an article with more information.

A fellow member also linked a similar article here: https://www.pugetsystems.com/labs/articles/Effects-of-Grill-Patterns-on-Fan-Performance-Noise-107/. Firstly, how accurate will these results apply to our proposed circle design compared to their mesh? It's possible that the increased diameter could influence different results. Secondly, mesh has one of the best noise level results. Intake airflow % actually beats the honeycomb counterpart using high fan speeds, which is arguably more important here. It does not lag too far behind with low fan speeds too.

I am not advocating the circles design, yet. I could be horribly off the mark and mistaken here and that I am missing something that will make me look like a total fool, as I am not well versed about such a high level topic. So once again, take all this with a pinch of salt. Nonetheless, I hope this was helpful as always.

I've also read (somewhere which I can't recall, maybe your links!) that hex wasn't good, which is OK with me because it isn't my favourite anyway.

 

[jakemagee]

Also, I created a Discord server

Is there a requirement to get on the discord server? The link tells me "invite invalid" ...

 

[grsychckn]

Is there a requirement to get on the discord server? The link tells me "invite invalid" ...

No, it is set as public and you should be able to access it without an invite. If you still have problems let me know. There is bad internet at this conference so I don't have a lot of easy access right now but I'll help where I can.

 

[grsychckn]

Update: I was playing around with the fan locations and realized that if I dropped support for the 25mm fans on top I would also have to move the power inlet which then reduces the supported PCIe width below 140mm. For now, I'm keeping the 25mm fan support but increasing the fan diameter to 70mm. I'll try to also include 60mm fan mounts as well.

I'm getting very little time here to work on the S402. When I do, the RDP connection across the country makes it very difficult to get anything done in a short period of time. I may have to schedule some time after I get back to knock some items out.

 

[Linero]

@Thehack Well written and sensible points. I can not disagree with any of them. Rounded squares is a close second after circles for me I guess. I think it is potentially the best middle ground for all of us here too. Either way, I just wanted to peek around at this topic and I have sated my curiosity for now.

@grsychckn It's all good, man. Best of luck and take care

 

[Piewalker]

No, it is set as public and you should be able to access it without an invite. If you still have problems let me know. There is bad internet at this conference so I don't have a lot of easy access right now but I'll help where I can.

Hey @grsychckn, the Discord Server link (https://discord.gg/QRnBu8) isn't working, just as @jakemagee said. Are you sure it's not private? What's the name or address so we can search and join?

 

[robbee]

Lazer3d is not a good source of information for this test.

1. He uses acrylic which requires a fairly large separation between honeycomb cutouts. This a steel chassis using punch manufacturering. The separation between cutouts can be very small.

2. The acrylic material he uses is either 3mm or 5mm. This means that the material is about 2.5x thicker than steel. The thicker material increases turbulence and noise. It especially increases noise when you have sharp angles, like honeycomb or triangular cutouts. Circular cutouts do not have this problem as they are uniform. The material thickness increases turbulence when you have small perforations, so if you have a thick material, use larger perforations.

The puget system article is much more accurate and reflective of this chassis. I would also be willing to weigh the puget article more highly since they deal with business to business system integration. Lazer3d is a small operation of custom acrylic consumer chassis.

Let's not overthink it. The greater the perforations % the better. The closer the chassis is to an open case the better it'll perform.

Personally I like the rounded squares. All of these options will do superbly compared to the Dancase or Ghost. I do not think there is large difference between the patterns in terms of performance.

While you certainly make some good points, steel and aluminium cases also suffer quite a but from turbulence due to fans being close to vent patterns. Even vent patterns without sharp angles, judging from the amount of dancase users that have been talking about it.

Also here's a relevant test that shows that there are differences between the different vent designs: https://www.pugetsystems.com/labs/articles/Effects-of-Grill-Patterns-on-Fan-Performance-Noise-107/

 

[Thehack]

While you certainly make some good points, steel and aluminium cases also suffer quite a but from turbulence due to fans being close to vent patterns. Even vent patterns without sharp angles, judging from the amount of dancase users that have been talking about it.

Also here's a relevant test that shows that there are differences between the different vent designs: https://www.pugetsystems.com/labs/articles/Effects-of-Grill-Patterns-on-Fan-Performance-Noise-107/

Yeah. I never meant to make the point that they don't. Acrylic 3mm panels makes the issue noticeably worse.

Dancase issue is the small hole size. The chassis had aesthetics as a large priority. If performance and silence was a bigger priority you'd have larger ventilation holes. The pitch of S401 holes are likely 3x that of Dancase.

The issue with acrylic is vs steel is that steel you can easily punch bigger and closer spaced holes without worrying about making it fragile. Acrylic is brittle so it is very easy to break it if it is mishandled. And the thicker material creates more turbulence. So the point was that it exacerbate the issues.

You also joined the conversation a bit late so the link you provided has already been discussed at length. The reply you quoted I had already mentioned it.

 

[jirachijirachi]

Is 60mm CPU height possible? That will include Zalman CNPS8900, which is one of the best performers at that height.