Production Winter One -- 15.6L SFF case, 3090 Support, 3-slot GPUs, dual 280mm radiators, CFD Optimized Design

ignsvn

By Toutatis!
Moderator
Apr 4, 2016
1,283
1,086
awesome. I’m going to keep an open mind when it comes to what switch to use, because I’m still a bit new to the world of mechanical keyboards :)

Happy to help. Those switches are - if I recall correctly - intentionally heavy to prevent accidental press, which I think will be better for a PC on/off switch.

And if you want to make them hot swap-able, you can use this hot swap sockets:


Edit:
Some 3D printed keycaps https://www.thingiverse.com/glitchpudding/collections/cherry-mx
 

WinterCharm

Airflow Optimizer
Original poster
Jan 19, 2019
325
1,337
Happy to help. Those switches are - if I recall correctly - intentionally heavy to prevent accidental press, which I think will be better for a PC on/off switch.

And if you want to make them hot swap-able, you can use this hot swap sockets:


Edit:
Some 3D printed keycaps https://www.thingiverse.com/glitchpudding/collections/cherry-mx

Amazing. I'm bookmarking this because it should be very useful for the future... :)

What’s your estimated time line ?

Estimated timeline:

1. Finish V0.4 and V0.5 CAD drawings by end of October (lots of finishing stuff like exact sizes of slots and screw holes and so on), CFD for airflow + liquid cooling configurations, exploring side panel perforation)
2. Finish Semi-Passive Kit (SPK) CPU and GPU heat sink design + CFD by end of November.
3. Consult with local machine shop and modify particular parts for manufacturability, build CNC files. (V0.7)
4. First Physical Prototype (via local machine shop) by Jan 2020
5. Get SPK prototype made (I need to find someone who can make my custom heat sink designs with integrated heat pipes for this) by Jan 2020.
6. Do physical testing in normal, and semi-passive, as well as passive modes, with real hardware and analyze data.

(As necessary) modify designs and repeat steps 3-6.

Once design are finalized, I'm hoping to send out some prototypes to maybe Optimum Tech, LTT, BitWit, Jayz, and Science Studio, so people can see what builds in this thing will look like, and I'll post a build guide / video of my own.

Crowdfunding will open around the same time (perhaps with the same machine shop & heat sink maker that I used for the prototypes (I'm trying to choose people who are willing to scale up). If we meet our funding goals, we will focus on delivering those orders. Once crowdfunding is started, My *only* goal will be to ensure that cases are made and shipped out to everyone, until all orders are fulfilled.

After all crowdfunding orders are fulfilled, It'll be waiting to hear feedback from users, and look at doing consistent low volume production, so we have cases on hand for those who want to order.

Do you think you could support a bifurcated riser so I can get 2 x 1080ti s and use an apogee 2 ?

What would bifurcated riser support require? Do I need to space the PCIE slots on the case a bit differently? (including a third one?) Or just include a second mounting plate for a bifurcated riser? Depending on what this requires I may or may not be able to support it.
 
Last edited:

TheArkratos

Caliper Novice
Mar 26, 2019
32
35
What would bifurcated riser support require? Do I need to space the PCIE slots on the case a bit differently? (including a third one?) Or just include a second mounting plate for a bifurcated riser? Depending on what this requires I may or may not be able to support it.

there are a bunch of bifurcation risers out there, most common are this from amerirack or any that C_Payne makes. You can ask questions over at this thread or on Hardforum. But basically just include a mounting plate for the riser you choose to support and you are good.
 
  • Like
Reactions: WinterCharm

osin17

Cable Smoosher
Sep 10, 2019
11
8
Really interesting case. I hope
* every corsair 240 mm aio will fit perfectly inside. Every 240mm AIO in fact
* every gpu 240 will fit inside also. AORUS, EVGA...
 
  • Like
Reactions: WinterCharm

WinterCharm

Airflow Optimizer
Original poster
Jan 19, 2019
325
1,337
As long as the hwl 240 gts fits I'm happy

They should fit comfortably :) I've been working on tweaking the dimensions *just a tiny bit* and I may have managed to get this thing *under* 15L, so there's an update coming in the next week or so...

I'm very close to done with the final designs of the machined case itself, and then we can continue to tweak Heat Sink design a bit more. The thing that's been giving me the most trouble is figuring out "safe" dimensions for the Semi-Passive Kit, as socket / die placement on Motherboards and GPU's is quite varied. It's going to take a while to work that stuff out, before I can say "these are the final design dimensions" for the SPK.

The current outer shell design will be perforated, and is done in a way to balance airflow so that the case can be run with fans In "all push" or "all pull" or a balanced in/out configuration, and in a way that airflow follows the path we desire.

I have continued to run CFD + Thermal analysis on various prototypes of the Semi-Passive Kit (SPK), and I'm narrowing down the design, to work with Keep Out Zones of the two latest sockets (AM4 and Z390).

Ultimately, this is the trouble with "high performance" air cooling in a small case -- it's not that it cannot be done, it's that piping on Liquid Cooling setups is flexible so that the heat can be carried from any socket >> any Rad. Whereas, when Air Cooling, especially semi passively, getting the heat pipe layout to a point where it can work with most sockets and motherboards, while also having enough surface area and fin spacing for airflow. In this case, a smaller interior volume actually *helps* because it directs airflow in a very particular way. But at the same time, smaller case dimensions throw constraints on your heat sink fin height, which is one way to add fin area when spacing has to be increased. And then you run into big sinks having to use aluminum fins (with Copper mounting block + heat pipes ofc for performance reasons) in order to achieve a cooler weight that won't stress/strain the PCB (even with a backplate). (A pure copper heat sink of this size would be like 1kg)
 
Last edited:

WinterCharm

Airflow Optimizer
Original poster
Jan 19, 2019
325
1,337
Really interesting case. I hope
* every corsair 240 mm aio will fit perfectly inside. Every 240mm AIO in fact
* every gpu 240 will fit inside also. AORUS, EVGA...

I'm going to double check and confirm with you that this is achieved, because I want this case to cater to "dual" AIO setups :)

Edit:

* every 240mm corsair AIO should fit just fine, so long as they're under 30mm thick, and you use slim fans :)
* Aorus water force and EVGA hybrid cards should work just fine, too.
 
Last edited:

WinterCharm

Airflow Optimizer
Original poster
Jan 19, 2019
325
1,337
2020 Update! :D

Happy New Year! Since October, I’ve been hard at work on my CFD-driven airflow focused SFFPC case. The first shots I showed you were simple concepts exploring the inner layout, and possible construction methods. The design has since evolved and matured. **I’m now ready to share a little bit about v0.5**, which is the latest iteration.

Now, let's get into the main design!

--------

Renders





--------

Case Concept

The case was designed to hit a balance between 5 main goals:

1. Design driven by functionality and an Understanding of Fluid Dynamics.
2. *Viable* variety of options for cooling high end components. (Liquid *and* Air cooling, including semi-passive, and passive configurations)
3. Compatibility with off-the-shelf parts (ITX motherboards, low profile CPU coolers, full length 2.n slot GPUs, SFX power supplies, 240mm radiators, AIOs, etc.)
4. Premium Look and Feel (Design driven by functionality does not *have* to be ugly)
5. Sub 15L so it’s squarely an SFF case.

--------

What niche does this case sit in?

Silent Winter sits on the *opposite* side of the cooling vs compactness tradeoffs we usually see in the SFF market (cases typically forego fan mounts and venting to shave down to sub-10L). Silent Winter is a **CFD-Driven, airflow optimized** take on the premium SFF case, with **maximum compatibility** in mind. Even if it makes the case slightly bigger (14L vs sub-10L), the benefits of better airflow are hard to argue with: More air lets you handle hotter components, gives you more OC headroom, or allows for lower fan RPMs, and therefore a quieter system.

More details about the Semi-Passive Kit will be coming soon, but they’re currently pending another round of CFD studies.

--------

Silent Winter v0.5 Case Specifications: (subject to change)

Physical Dimensions:
  • Length: 30.5 cm
  • Height: 30.5 cm
  • Width: 15.4 cm
  • Volume: 14.33 L or 14,325 cm^3

Parts Compatibility:

Code:
    Motherboard: 
         Mini ITX with dual rad & quad fan
         Mini DTX with Single rad & quad fan
         {!}(Air Cooling Requires 3 fan headers on Motherboard)
         {!}(Liquid cooling Requires 2 fan headers + 1 pump header on Motherboard)
         {!}(Semi-passive Mode Requires 2 fan headers on Motherboard)

    CPU Sink: 60 mm max height

    GPU:   295 mm x 48mm x 150 mm max (~2.5 slot)
           200 mm x 54mm x 150 mm max {!}(Short GPU REQUIRED with pump/res)   
           (Note: 1S = 18mm, 2S = 36mm, 3S = 54mm)

    Radiators:  Dual 240mm Radiators: (290mm x 140mm x 30mm) dimensions
              
    Fans:  Quad 140mm x 25mm fans (In anticipation of NFA14x25)
           Quad 120mm x 25mm fans
           Quad 120mm x 15mm slim fans {!}(with radiators)

    Pump/Res: 130mm x 130mm x 55mm (NEED Short GPU) {!}(Check GPU & Pump/Res dimensions carefully!)
            (If you decide to not use a pump-block combo on the CPU/GPU)
      
            EK-Quantum Kinetic FLT 120 DDC (120mm x 105mm x 51.2mm)
            EK-XTOP SPC-60 PWM (62mm x 62mm x 51mm)
            Swiftech MCRES MICRO REV2 RESERVOIR (101mm x 103mm x 38mm)
            Swiftech MCP35X Micropump (59mm x 92mm x 62mm)
            Swiftech MCP35B Micropump (59mm x 92mm x 62mm)
            Swiftech MCP355 Micropump (44mm x 87mm x 61mm)
            Alphacool DC-LT Pump w/ FrozenCPU Res (49mm x 49mm x 65mm)
            Lightobject DC24 Straight-through brushless pump (116mm X 50mm X 50mm)       

    Power Supply: SFX (100 x 125 x 63.5 mm) or SFX-L (125 x 125 x 63.5 - not recommended, but okay)

    Power Button: Kailh Box Navy with custom round aluminum keycap (changeable for modding)

    SATA SSDs: Room for 3 x 2.5" SSDs (100mm x 70mm x 6.8mm) {!}(SHORT GPU required)

---------

Layout, and Structure

With v05, WinterCase had fundamental changes to the structure. In previous designs (v03 and before) the top and side panels attached to an internal frame. In v04 and v05 this internal frame was removed, and now the *panels* themselves are the main structural component. This was done to maximize *internal* volume.

The inner layout is a sandwich layout (popularized by the Dan A4) with the PCIE riser bridging the GPU / Motherboard. While this is typical of a Sub10L case, Traditionally, cases above 10L have a “standard” PC layout (such as the NCase M1). However, there are problems with this design in maintaining a *linear airflow path* (which is crucial for passive cooling). The “traditional” PC layout creates dead zones in airflow when passively cooling a case.

The linear airflow path and unique vent pattern emphasizes low air resistance in this case, allowing for natural convection, and maximizing airflow at *low* fan RPMs in semipassive or active cooling setups. This design enables an *entirely silent* passively cooled mid-range system (3700X + 2060 Super) or a *whisper quiet* high end system (3950X + 2080Ti) using 4 case fans and huge heat sinks, giving you the reliability of air cooling with the performance of water cooling. In Active Air-cooled mode, you can maximize part compatibility for a reasonably quiet and very flexible system. In Active Liquid Cooled mode, **dual 240mm slim (30mm or thinner) radiators** ensure gobs of overclocking headroom even on high end components (9900k OC + 2080Ti OC) for the hardcore enthusiast.

----------

Semi-Passive Kit (SUBJECT TO CHANGE)

Passive Cooling Projection: (Assuming 5CFM natural convection) [2] [3]

* CPU: up to 100W (~85ºC max)
* GPU: up to 150W (~85ºC max)

Semi-passive Cooling Projection: (Assuming 30 CFM fan airflow) [2] [3]

* CPU: up to 170W (~85ºC max)
* GPU: up to 335W (~85ºC max)

[2] TDP's here are rough preliminary calculations based on expected heatsink volume and area, and 23ºC / 73ºF ambient air. These are obviously subject to change, as the designs are not final. I cannot stress this enough.

[3] Numbers are based of REAL power consumption, NOT TDP. TDP is largely a marketing term, and it's highly advisable to base cooling off Package Power Consumption (Courtesy of Anandtech) as almost all power consumed by a chip (aside from rounding errors) is dissipated as heat. Neither Intel nor AMD provide accurate TDPs for their chips... (See chart for listed TDP vs Real power consumption). Keep this in mind when considering hardware to use in this build.

-----------

Closing Comments

This is a **late design prototype**. It’s about 80% complete. Most of the slots / cutouts / standoffs / dimensions are finalized. Some things remaining include:

* Baseplate - will attach to the rectangular section on the bottom (so the case doesn't topple)
* Carrying Handle - (raised rectangular section up top will house this, should sit flush)
* Tempered Glass Side Panels (optional, but widely requested)
* Final Screw Placements / Assembly
* Airflow related tweaks (currently running CFD on this)
* Heat Sink Design tweaks (pending more CFD)

Once I finish the detailed design, I plan to make a physical prototype and validate my numbers, make some revisions, and lining up production. Then, comes crowdfunding and actually starting production. I *hope* to get these in people’s hands by August 2020.

Thank you for taking the time to read over this and check out my design. Feel free to ask me questions in the comments here. **I want to know your thoughts about this design, and where you think I could improve things.** <3

------------

Timeline

1. Finish Semi-Passive Kit (SPK) CPU and GPU heat sink design + CFD (by end of Jan 2020)
2. Update and finalize first CAD for first physical prototype (v06)
3. Create First Physical Prototype (via local machine shop) by (Early Feb 2020)
4. Do physical testing in normal, semi-passive, and passive modes, with real hardware and analyze data, update CFD models. (Feb 2020)
5. (As necessary) modify designs and repeat steps 3-6. (Feb 2020)
6. Update for manufacturability (v07)
8. Finalize Designs and prepare for crowdfunding. (hopefully by March 2020)
9. Crowdfunding + Manufacturing (March-April of 2020)
10. Delivery of case (hopefully by August 2020)
 
  • Like
Reactions: DmanX and an_Age

runninghack

Minimal Tinkerer
New User
Nov 21, 2019
4
3
Are you willing to share some results of your CFD studies? Even we're don't know much (if any) about it, it would still be helpful for us to understand what exactly have been considered.
 
  • Like
Reactions: WinterCharm

an_Age

Average Stuffer
Nov 21, 2018
67
46
Very cool concept and I'm definitely going to follow the production. I can't help but think the production schedule is a bit optimistic though. It all seems complicated and a big risk for delays?
 

WinterCharm

Airflow Optimizer
Original poster
Jan 19, 2019
325
1,337
Very cool concept and I'm definitely going to follow the production. I can't help but think the production schedule is a bit optimistic though. It all seems complicated and a big risk for delays?

Yes, right now the production schedule is a tad optimistic, but I'll adjust It as things progress. I'd rather set difficult goals and strive toward them with lots of effort, than set "safe" goals and get too complacent...
 
Last edited:

WinterCharm

Airflow Optimizer
Original poster
Jan 19, 2019
325
1,337
Are you willing to share some results of your CFD studies? Even we're don't know much (if any) about it, it would still be helpful for us to understand what exactly have been considered.

Yes, I will be sharing some of the results of the CFD studies later on :). I'm still running the current set of studies on the heat sink design. They should complete in a week's time. (my poor 4790k has been toiling away at it for a while now).

If you're curious I'll jump into a bunch of detail on how these studies have informed the design thus far: (wall of text for all you enthusiasts) :D

The first thing I learned during the CFD studies informed the top and bottom panel design. To elaborate: whenever there is flow of a fluid past a surface (like air past aluminum), there is drag because of the boundary layer, leading to some amount of stagnant air. This results in a pressure drop which reduces flow velocity. So the reason you see such massive openings on the top and bottom is because in a passive cooling configuration, very slow convection can be significantly "stopped up" by a series of small holes. For example, a case which has an array of small holes in the top panel (with, let's say 50% open area) has 13.5 METERS of total perimeter (adding all the holes). That is a lot of surface against which there will be stagnated flow *in a passive cooling setup*. It's fine in active cooling because you use a fan to overcome the resistance, but you still pay with a pressure drop (blow on one side of the holes, and you'll feel significantly less force on the other side. Repeat this with a single big hole that achieves 50% OA and you'll incur less pressure drop due to a smaller boundary perimeter.

So, the reason for those huge slots on the end plates is to minimize air resistance for all flow configurations. In passive cooling, it lets warm air escape with minimal resistance, and prevents stagnation that would effectively hotbox the PC. In active cooling configurations, the lower resistance leads to less fan noise, as you can achieve good flow at significantly lower fan RPMs (fan airspeed velocity is highest at the furthest ⅓ of the fan blade, so having a rectangle covering the hub leads to minimal losses). This highly effective airflow configuration when combined with the semi passive heatsink which has widely spaced fins for easy airflow (for similar reasons), cuts down significantly on air resistance, boosting cooling capacity, but keep noise *very* low.

Airflow also drove the choice of going with a Sandwich layout. One of the things you learn in fluid dynamics is that changing the direction of fluid flow by 90º leads to a *roughly* 30% pressure drop. Lower pressure = less airflow. To compensate, you need to generate higher "initial" pressure (meaning higher fan RPM / noise). While this case is big enough to contain a "traditional" PC layout (T shaped, no riser) the heat sink geometry required for passive cooling would be significantly more challenging (See the 2019 Mac Pro Thermal Design, and layout to get an idea of what this would look like). In an SFF case, due to the Power Supply being in the way, a 90º Turn (and the accompanying pressure drop) would be unavoidable, leading to cooling inefficiencies. By contrast, a Sandwich layout (pioneered by the Dan A4), with all components vertical, would preserve a linear airflow path in a small form factor, and allow for a sane design and easy mounting of large heatsinks to both the CPU and GPU.

Hopefully that illuminates how CFD and fluid flow have informed the overall design and airflow considerations made in this case :)

--------

Also on the topic of heat sink design, Passive heatsinks are a unique beast and the design constraints and conditions are a bit different when compared to "active" (fan-on) heatsinks. This is worth exploring in more detail:

The closer together your fins are, the higher resistance you will have to pushing air through the fin stack. To overcome the resistance to flow, you have to place fans directly on the heat sink (like we see in every CPU cooler), and if fin density goes up enough, you need fans with higher static pressure (like on radiators with dense fins in a liquid cooling loop). Passive heatsinks rely on natural convection, with only the air density difference driving flow. If you pulled the fans off a "normal" heatsink (that has fins close together), it would hurt cooling performance significantly.

To support passive airflow, Your inter-fin distance has to *widen* to prevent stagnation, especially as the heat sink gets larger/longer and the path air has to travel increases (like in a vertical fin stack). In semi-passive mode (with case fans + a heat sink some distance away from the fan), this constraint still applies as even the best fans at lower rpms and far away from the middle of the heat sink are, at best, only pushing 4 inches H2O column of pressure, and the pressure losses are real for any heat sink that's long, or far away from a fan. (Of course, one way around this is massively boost fan RPM! if you use 10k RPM server blower fans, then you can get away with "traditional" heatsinks like they do in 1U and 2U server chassis, with linear airflow front>>back, but now have a screaming banshee in a box...)

So, what if you took an off the shelf "fan-on" heat sink like the NH-U12S, and remove every 3 out of 4 fins, and slap that into a CPU? You have lots of fin spacing, so air will flow nicely through it. However, you've just removed a bunch of mass, and walked into some overheating issues as the *overall mass of the system is too low*. Therefore you must increase the thermal mass of the fins by making them thicker and larger to keep chip temps under control. This is the defining difference between a semi-passive / passive, vs an active heat sink.

Here’s a perfect example of everything I've talked about - a passive/semi passive vs active heat sink designed by a company regarded as one of the best in the business: Noctua.

Noctua NH D14 Active cooler - notice how thin and closely spaced the fins are

Noctua Passive CPU cooler - notice how thick and widely spaced the fins are. It can dissipate 180W in “semi passive” mode.

*both* are valid heat sink designs, but with different applications and tradeoffs in mind, and both are designed by Noctua - a company that *really* knows their stuff when it comes to cooling computers. The stark differences in design are due to the considerations I’ve outlined above.

In the WinterCase, the SPK heat sinks will actually extend from top to bottom leaving space for the 4 140x25mm fans. While the heat sinks handle passive cooling, they can also be used for extremely quiet semi-passive cooling (driven by the case fans). With wide fin spacing, you can achieve lots of flow *and run the fans extremely slow/quiet* and still move lots of air through the heat sink, enabling very quiet high end builds. While it sounds too good to be true, there is a tradeoff - passive heatsinks take up significantly more space than their densely packed active counterparts.
 
Last edited:

Necere

Shrink Way Wielder
NCASE
Feb 22, 2015
1,677
3,168
I don't see any details on construction, can I ask what methods you're thinking of using? It kind of looks from the renders that the exterior shell is one piece and meant to be extruded, but the size of the extrusion + all those CNC'd holes would likely end up being cost prohibitive. You'd probably need to break it into separate parts for front/back/sides, at a minimum, and/or use sheet metal instead. Though with sheet metal it's unlikely you could get a one-piece wraparound shell, either, unless your manufacturer has the specific tooling for that kind of bend operation.
 

WinterCharm

Airflow Optimizer
Original poster
Jan 19, 2019
325
1,337
I don't see any details on construction, can I ask what methods you're thinking of using? It kind of looks from the renders that the exterior shell is one piece and meant to be extruded, but the size of the extrusion + all those CNC'd holes would likely end up being cost prohibitive. You'd probably need to break it into separate parts for front/back/sides, at a minimum, and/or use sheet metal instead. Though with sheet metal it's unlikely you could get a one-piece wraparound shell, either, unless your manufacturer has the specific tooling for that kind of bend operation.

The outer shell is going to be 4 pieces, (front + rear + side panels) that I plan to have separately extruded and then machined. The render has bad lighting (my mistake) which blew out some of the color and details so the seams got lost.

Structurally, the front + rear panel + central spine, and top / bottom "end plates" are the main load bearing components.

Currently, I'm thinking of doing Aluminum 6061 for the entire case (including the spine). It's currently 13 parts to be machined (3 small, 4 medium, 5 large, and 1 tiny power button) and 4 of those 13 parts are 2D / flat. This, plus the PCIE riser, and heat sinks are all the parts. I'm planning to contract the heat sinks to a separate company that specializes in those.

I haven't lined up a manufacturer yet, made any physical prototypes, or had to make alterations for manufacturability (yet - I fully expect to). If you've got any advice for a newbie such as myself, who's willing to put the time and effort in to get this right, It would be greatly appreciated. You've been doing this for a while, and I'm sure you have a lot of experience and words of wisdom for anyone venturing into this space.
 
Last edited:

Necere

Shrink Way Wielder
NCASE
Feb 22, 2015
1,677
3,168
I don't have too much knowledge about doing extrusion/CNC cases myself, but I would point you towards the Ghost, Mjolnir, and Circle One as those are the closest to what you're doing. If you're committed to this method of construction, I wouldn't expect it to come out any less costly than those cases.

One general tip for CNC is that more holes = more machine time = higher cost, so anything you can do to reduce the total number of machined holes will help to keep costs under control.
 
  • Like
Reactions: WinterCharm

an_Age

Average Stuffer
Nov 21, 2018
67
46
Yes, right now the production schedule is a tad optimistic, but I'll adjust It as things progress. I'd rather set difficult goals and strive toward them with lots of effort, than set "safe" goals and get too complacent...

That's fair! But I still think safe goals and beating them is even better.

Regardless, best of luck. I'm looking forward to seeing what comes out in the end.
 
  • Like
Reactions: WinterCharm

gwertheim

King of Cable Management
Gold Supporter
Nov 27, 2017
938
1,553
The outer shell is going to be 4 pieces, (front + rear + side panels) that I plan to have separately extruded and then machined. The render has bad lighting (my mistake) which blew out some of the color and details so the seams got lost.

Structurally, the front + rear panel + central spine, and top / bottom "end plates" are the main load bearing components.

Currently, I'm thinking of doing Aluminum 6061 for the entire case (including the spine). It's currently 13 parts to be machined (3 small, 4 medium, 5 large, and 1 tiny power button) and 4 of those 13 parts are 2D / flat. This, plus the PCIE riser, and heat sinks are all the parts. I'm planning to contract the heat sinks to a separate company that specializes in those.

I haven't lined up a manufacturer yet, made any physical prototypes, or had to make alterations for manufacturability (yet - I fully expect to). If you've got any advice for a newbie such as myself, who's willing to put the time and effort in to get this right, It would be greatly appreciated. You've been doing this for a while, and I'm sure you have a lot of experience and words of wisdom for anyone venturing into this space.

What I would do is find a company who does rapid prototyping in your area and see if you can get the case printed in plastic or something cheap and sturdy. Then you can see it and even do a test fit of some parts and further refine your design. Then work with someone to refine the product even further until your satisfied that you have a product that you are happy with
 

WinterCharm

Airflow Optimizer
Original poster
Jan 19, 2019
325
1,337
What I would do is find a company who does rapid prototyping in your area and see if you can get the case printed in plastic or something cheap and sturdy. Then you can see it and even do a test fit of some parts and further refine your design. Then work with someone to refine the product even further until your satisfied that you have a product that you are happy with

I shall explore this option for sure, thanks for the tip! :)