1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

Scratch Build – In Progress BBJ-G

Discussion in 'Project Logs' started by Frode Bergeton Nilsen, 19 Jan 2021.

  1. Frode Bergeton Nilsen

    Frode Bergeton Nilsen Minimodder

    Joined:
    15 Jan 2021
    Posts:
    46
    Likes Received:
    54
    Backbruner Jr. Glass (BBJG)
    3D-printed projects are just different. I hope to find a home on this platform, to share my work.

    [​IMG]
    Air intake bottom and top. Exhaust in the rear. 3x360 rads. Dual pump. Support any GPU, MB, and PSU up to 200mm. IO at the top, with a generous channel top to bottom for external cabling. Rubber feets at any corner, that is designed to support tilting. Designed ground up for water cooling, particularly filling and draining.

    This is originally a mesh design, that is converted to glass.

    [​IMG]
    Hardly anything about this design is standard. It might look simple, but given the print size limits, and print physics, the design is probably one of the more complex in here. Here is an example of a cross section of the top.

    [​IMG]
    This is the third iteration. This design actually works, and makes a really solid construction. I started out, not even knowing if this was feasible, as I could not find any work out there, using any technique that was applicable for this kind of construction.

    This is the only glass design of mine, as glass makes no sense for the bigger ones. If people rather would have me go all mesh, which is far better performing than this design, let me know. This is a 3x360=1080 design. I also got a 1440, a 2180 and a 2880 design. All in cases that one man can carry. My plan was to get the print farm running, finish this, then do the 2880, then a air cooled version. The BBJG is the most ordinary of my designs.

    Please note that Fusion 360 is brought to its knees given the complexity of this design. I have seen people having really nice MBs in their designs, but I simply have no idea how to get hold of that. I have not bothered to design fans for 360. If someone has anything that can be shared, like a fan, GFX and MB, my computer will be happy to chew on it for a few hours, so I get a better 360 render. Adding a single nut, currently takes minutes to compute.

    If you like my work, and would like me in your workforce. Please let me know. I opened my first cad application ever, and printed my first piece ever, april 2020. I've never taken a single graphic design class, I am an infrastructure designer. If you like this design, and would like to print it in your printfarm, let me know. If you want to join in, and try to generate a community, in which designers can jointly make designs to print, and would like to work me on stuff like this, let me know. At least, we now have a framework to build on.
     
  2. Cheapskate

    Cheapskate Insane? or just stupid?

    Joined:
    13 May 2007
    Posts:
    12,411
    Likes Received:
    1,968
    Looks neat! Am I seeing printed fasteners? I like the look of big screws and bolts.
    If you are selling here, be warned; Most of this community is more interested in doing things themselves. You might do well with print services, though.
     
  3. Defyant Mods

    Defyant Mods Multimodder

    Joined:
    14 Sep 2004
    Posts:
    1,305
    Likes Received:
    780
    The more 3D printed cases i see the happier i am :rock: Your right at home here :thumb:
     
  4. Frode Bergeton Nilsen

    Frode Bergeton Nilsen Minimodder

    Joined:
    15 Jan 2021
    Posts:
    46
    Likes Received:
    54
    This is a really Labours build, that really should not be printed by non-professionals.

    I would love to collaborate with people in this forum, and that would naturally be completely free. Just be aware how much work this is, both to design and to produce.

    Bolts are printed in PETG. M14 2mm thread. I could not get smaller bolts to work properly, using a 0.6 nozzle. I failed to find anyone using bolts in this embedded manner, thus as with most anything with this build, had to figure out everything from scratch, with no other work at all, to base mine on.
    Thank you.

    I looked around a bit before posting, and your work was part of the reason of me joining in. Great work.
     
    Defyant Mods likes this.
  5. Frode Bergeton Nilsen

    Frode Bergeton Nilsen Minimodder

    Joined:
    15 Jan 2021
    Posts:
    46
    Likes Received:
    54
    The Corner BBJ section
    This section is used for both the 4mm mesh BBJ and the glass one. It is used in the bottom right rear and top left rear corner. It's a pain to both design and print.

    [​IMG]

    Filament
    I need the filament to not give in to physical pressure, and deform as a result of physical stress. That simply kills PLA, as PLA is malleable. PLA is not suited for anything really, other than prototyping stuff, that are expected to be thrown away. I need to be able to print, without generating a gass issue. That kills ABS. At this point, I simply do not have the gear required, to deal with the gasses produced by ABS. ASA produces less gass, but, enough gasses, that it to is unfeasible and unhealthy for me as a printer. That pretty much dictate the use of PETG, as the only viable filament.

    PETG cannot be easily painted. It probably is possible, I simply do not know how, nor does anyone else it seems. There is like a billion people failing at it, yet no successes. Not that I am aware.

    Prusa states that it can be sanded and work on, which is my experience as well. I will use this property for the female M14 thread.

    This leaves me, with a plastic that I cannot really finish or polish the surface on. I need to design with surface finish in mind, the surface it has, coming straight of the printer.

    Object surfaces
    There are six surfaces of a cube. If printing a cube, the surface facing the print plate, or down, will have a print plate finish. The sides of the cube, will have a equal side finish, while the top will have a third, top finish. A printed case is thus a mess of different surfaces with different structure.

    It is impossible to print great looking circles, in the vertical direction. Those imperfections, is a part of 3D printing, and as the finish of the printer is final, well, it is final. That is just the way it is. The only way to remedy this, is to attempt to hide joints that needs vertical printing, as to suppress the issue. Design to avoid it, as well.

    Also, I cannot print a ball, or a sphere. Edges can be canfered, but that leaves a pretty squary expression. It is really hard to design for curves, as they print really ugly vertically. Particularly with a bigger nozzle.

    Mesh looks ugly if printed as a top surface. Mesh turns in structural weak, if printed as a top surface with support. Depending on the mesh shape, it may not print vertically at all, like a honeycomb pattern or straight angled rectangles. Circles can be printed, but leaves a lot of tricky areas to fill, in between the circles, and thus, prints turn out ugly.

    Support
    I cannot print into thin air. I have to print on something. I try to stick to 45° angle as a max, as far as possible. PETG bridge poorly, quickly requiring support. The surface printed on top of supports, are simply ugly, and slightly drooping, thus ruining the physics of the part. As a result, I need to design for the least amount of support possible. I hardly use it at all. This requirement is a pain to work with. In general, I use support for the fastener of bolts, and some places for the internal cable channel, as it is hidden, and thus can be ugly.

    Vertical printing
    I use a printer with a moving print bed. On top of that, PETG is sticky, and as the plastic cools, it tends to expand and assert pressure upwards, into the nozzle. As a result, it is tricky to reliably print objects like longer rods, in an vertical orientation. This limits the use of cylinders even further, as they cannot be printed nicely horizontally. It also means that I need to design any part as not to need to be printed vertically, unless it is fairly wide, as to support itself properly.

    Warping
    Parts that are thick in the vertical direction, typically warps in its corners facing the print bed. Angus made a piece on it, stating that he had no real remedy for it. By keeping my external surfaces thin in general, I basically work around issue. For the eight corner pieces, they all warp in the corners. That is a deliberate choice, of structural integrity, over finish. The top corner feet helps hide the issue.

    Integrated horisontal mesh surface
    The integration of the mesh part, adds considerable strength in the x-y plane, or the horizontal plane. It also adds 6mm thickness in the y direction (vertically), which I needed for structural integrity, given how Swiss cheese this piece is.

    [​IMG]

    Nozzle size
    Prints of this size, is a crash course on nozzle size. If I were to print this object with the same infill, and as thick perimeters as I do, it takes 1 day 20 hours and 15 minutes with a .4mm nozzle. Note that the red colored sections are in top print finish.

    [​IMG]

    Printing it with safe settings and a .6 nozzle, it takes only 21 hours and 17 minutes.

    [​IMG]

    0.6mm nozzle effects both print finish, and what can be printed. There are four rear corners that prints in about 85 hours this way, instead of 177 hours with a .4mm nozzle. That is just the four rear corner pieces.

    Education needed
    The issue of just releasing the source to the public, is that people will print this with hardly any infill and use PLA. People will end up with boxes that melts (50°C) and loses structural integrity over time. On top of that, the reduction in perimeter and and infill people will insist on to reduce print time, combined with pushing print speeds weakening layer adhesion, and we have a recipe for a 15Kg box disintegrating while being lifted. People will simply get hurt, as will follow of stupidity. It is difficult for me, to stress this point enough. A water cooled PC in a disintegrating chassis, makes people sing "We didn't start the fire", blaming the sun for melting their PLA printed case.

    Just reducing perimeter, hours will be shaved of the time to print, but the end part will lack the strength I designed it to have. As in the sample beneath.

    [​IMG]

    Mass
    The piece is about 440g. Thus I get about two of these, out of a single spool of 1Kg filament. 750g spools thus makes little sense. The bigger nozzle is far more forgiving on the filament. Cheap PETG filament works just fine.

    Print support
    For this part, I only use print support on the print bed, and only for the countersinking of the bolts. This results in a slight droop, so I need to shave of a layer of plastic on the bolts to compensate for it.

    This is the amount of wasted plastic, for support.
    [​IMG]

    Front crossbar
    The crossbar on the side, connects on three out of four surfaces. There is some distance between the holes, adding strength. Looking closer, you should be able to see, that the main fasteners are on the inside. If moving the fasteners to the center side of the box, the physics simply do not add up. There is a hard limit of about 247mm, as the max length for my Prusa.

    [​IMG]

    Rear crossbars
    The crossbar in the rear, has far tighter spacing between the bolts. The bigger ones, needs to stay in place, for the main structure to remain intact. Every hole is limiting any other hole. They are all interconnected.

    [​IMG]

    Vertical crossbar
    The same tight spacing of the holes. The rigidity of the top and bottom makes for fairly rigid plane structures, and the rads in the rear add a lot of rigidity as well. This turned out to be enough. Note that both the inner and outer parts, needs rigidity on their own, and adding 10mm+10mm parts, do not equate a 20mm thick piece. Having one angled part, rather than two separate flat ones, offers significantly more strength.

    [​IMG]

    Cable channel
    The case is designed to support cable spaghetti, as in stuffing it behind mesh. This turned out to be a serious challenge. I decided to keep the outer mesh an integral part of the corners, thus they cannot be removed. Due to the fact that I cannot print plastic in mid air, making a box for cables, requires multiple parts. They need to be possible to mount, thus the need for both a braket and a cover. What I ended up with, is a design in which the pieces needs to be mounted before the rads go in. Cables being accessed from the outside, removing the middle outer cover.

    I made a removable bracket, for the more cable modding savvy, that might not want this cover at all. I have not made parts for that, but that is really easy to do.

    [​IMG]

    The cover itself can then be tailored as needed. The bolts will end up underneath the rads and the tubes, as in impossible to access once the build is complete.

    [​IMG]

    Note
    I am not sure if people care for all these details. To give anything printed any sex appeal, all these issues simply needs to be ironed out. If people find this post a difficult read, I have no idea how to cover the IO section on the top, as it is far more complex than this. I will do my best, to make stuff as accessible as possible for those not being printing nerds.

    Not comprehending the details and complexity is fine. The thing is, that most people start asking for things, covered in the above sections, describing what is possible, and impossible. What they ask for, is typically impossible. Like "I want to print this in PLA with my Chinese 3D-printer that cannot print PETG", and "I will just plaster it and give it a coat of paint". If you live in an alternate universe, in which PLA is not malleable and do not lose structural integrity at about 50°C, then I am simply not in that universe. I simply cannot stress this point enough.

    Next is printing this part, and get a second printer up running.
     
    IBMer, Cheapskate and AV Case Mods like this.
  6. Cheapskate

    Cheapskate Insane? or just stupid?

    Joined:
    13 May 2007
    Posts:
    12,411
    Likes Received:
    1,968
    Relying solely on the printer and not tapping/drilling smaller holes for screws is making it overly complicated, but that's a lot of the charm of the design. :D
    The super-fine threads of smaller metric screws doesn't work well for plastics either.
     
    Defyant Mods likes this.
  7. Frode Bergeton Nilsen

    Frode Bergeton Nilsen Minimodder

    Joined:
    15 Jan 2021
    Posts:
    46
    Likes Received:
    54
    The build volume, is the main challenge designing this. Build volume, is the maximum depth, width, and height of a printed object. The screws are big, relative to the build volume, resulting in a Swiss cheese. In particular, the screw heads needs to be big, to accompany a wider flat drive, due to the weak torque characteristics of PETG. That being said, once you master the technique, this isn't that complicated, rather straight forward. It just takes like forever to model.

    I have never seen this soluting for cable management anywhere. The two smaller holes in the last image above, those are directly beneath the 11mm space between the fans. The water tubes runs in an arch around the cables. There is plenty of room to join and extend cables in the messy sandwich box, and due to the density of the mesh, black cables are basicly invisible in that box. Yet it doubles as ventilation. The single bigger hole needed to be where it is at, due to the channel for the external cables, and hole placements. cables will run from that, and into the left or right cable channel, behind the MB. For a shrewed modder, this cable channel might be unnecessary. The inner mesh can be removed, or a center piece without holes can be made.

    The build turns in a lot slicker, if there is an inner mesh, as it makes for a less messy interior design, with cleaner lines. This is only a 1080 rad design, but that is 9 fan cables, and if running RGB fans, that is a plenty of cables. There is a cable channel in both the bottom and the top. In my early designs, I ran the cables straight into the PSU section. That was just a mess. In this design, fan cables are kept separate from other cables, for ease of maintenance and assembly. This is a derivative of my Sideburner design, which has 24 fans, and 4 fan controllers. Custom cables for all that, costs a fortune.

    For this design, using the motherboard fan control, makes sense, if using the right fans. Cables can easily be pulled into the cable channel, the cable spaghetti can be kept inside the channel, and single cables can run straight to the fan headers.

    [​IMG]

    The smaller mech piece, has a version with a smaller cutout for running cables. The two embedded m14 tapped holes, are just regular 54mm separated. These holes are in all four rear corner pieces, and can be used to mount fan controllers, pumps, RGB controllers, or SSDs. Since i have a standardized design, a single bracket can be used for all these spots. Designing a bracket for a new device is fairly straight forward. For this design, the spot behind the PSU, is the most elegant for a RGB controller, while an SATA SSD, would probably make the most sense behind the MB, to the left from the front, as for simple and neat cabling.

    There is a modulearity to all the parts, and design for reuse is a design target. A 27mm base, and a 4mm, 6mm, and 10mm surface covering the base. Screws are 13,5mm from the edges, and basically a multipler of 27 is applied almost everywhere. This leaves a repetitive pattern, that in the end, is orderly and calm. Symmetry, lines, and surfaces, are all designed to fit the expression I am seeking.

    [​IMG]
    54mm spacing m14 holes are marked up in the image above. Note that the two boxes running in line, the two hole is the center, also are 54mm apart. This is both a functional feature, as to support modules, and more important, an expressional feature of the expression I am seeking. As a designer, I use symetry. In the image above, a lot of patterns are repeated or mirrored. A key trait of mine. Mesh i centered and mirrored. Mesh patterns are carried on from one part to another. There are clear lines throughout the design. Even when I do need to deviate, there are lines and structures that I do play on, and that is by purpose.

    [​IMG]

    In the image above, the long line follows from the structure, into the crossbeam. The big X, is hidden from view, once the covers are in place. The hole next to the small x, needs the line neighbouring it to be shifted, and it is by 10mm, which is the same padding on the opposite side, as in the long line. The short line marks the center of the frame, and the holes marked with circles, are mirrored by the center line. This is also why, the holes on the inside is not 54mm apart, but again, they are hidden from view. The width of the PSU section is dictated by the ATX standard. The front is 10mm+27mm+6mm. A 27mm part, padded by 4, 6, or 10 mm. The EKWB rad is 130mm +-1mm. Thus, it needs 131mm width. Two holes, the first at 13,5mm, and then 54mm spacing. Not that hard really. A technique carefully designed, to support my personal expression. The width of the Sideburner design is square as for width and dept, defined by rad width. Those dimensions are carried over to the rear section, for the backburner. As in modular design.

    In the last image above, it would be natural to place the top of the crossbar in the front, down on the print plate. But that will result in a different surface than the corner parts. Thus printing it the other way, or up-side-down, will produce an uglier part, but a consistent surface. To support that, the corossbar part, needs to be signed with that orientation in mind, as not to need support material to print reliably.

    Looking at the last image above, note that the top surface of the mesh is visible. It does not print gracefully. The top surface of printed mesh, the side opposite of the printbed, is ugly. Adding that second top mesh, hides the flaw.

    Also, for height there is an opening on the rad side of 30mm-120mm-120mm-120mm-30mm. Again, the top could have been shallower. I design for reuse of parts, repetitive patterns, and symmetry. I seek a calm expression, with clearly repetitive patterns, with clear lines and shapes. I seek an expression, in which complex parts, like a rad assembly for example, feels like one part, times three. The goal, is to reflect simplicity, black boxing the complexity into fewer and bigger parts. I started out trying to design for compactness, but found it stupid in the end. This design is easy to work with, a compact design is not. Also, looking at this box, every section is easily recognized for its role and functionality. In particular, the corner feet, and filling and draining this thing. There is a lot of form following function here, but also plenty of form generating functions.

    That pretty much covers the basis. I've ran out of .6mm nozzles, so there is another delay. I only got one.
     
  8. Frode Bergeton Nilsen

    Frode Bergeton Nilsen Minimodder

    Joined:
    15 Jan 2021
    Posts:
    46
    Likes Received:
    54
    The part is now in process.

    As always, I needed to demoisturize the filament, which is done by a few hours in the owen. Next is checking the nozzle and heat block for leaks, and clean both of them. Leaks between the heatbreak and the nozzle happens on a regular basis, more on that later. It just will happen, and I will rather cover it then. Since I use a coated nozzle, cleaning is done using regular household kitchen paper. cleaning must be done hot, that is, the nozzle is at 250°C, when it is cleaned. The plastic need to be liquid.

    Image of printhead, with nozzle,and heating block with rubber sock:
    [​IMG]

    There are some modification needed, for reliable results. The Prusa heatbreak needs to go, as it requires too high temp for the filament in use. I also use a sock on the heater element, as it safes many a part during printing, when a leaks occur, and prints in general turns in nicer. Also, cleaning is a lot easier. The nozzle is a E3D 0.6mm one.

    The heatbed needs to be cleaned. I get best results cleaning it first with acetone, then with isopropanol, in that order.

    Image of cleaning agents:
    [​IMG]

    The Prusa starts out with the skirt loop, which is just a single strip of plastic around the print. Then it prints the support material. Then perimeters, which is the outer edges of the part. I use 4 layers both vertically. Finally, the perimeters is filled by infill. Using a bigger nozzle, allows for 0.3mm height for each layer, as opposed to 0.15mm or 0.2mm for a regular nozzle. The heigher layer height is more forgiving for uneven surfaces or filament challenges. Still, it happens on a regular basis, that the first layer does not stick to the surface, and thus the print is a failure.

    Please note that the following images are macro shots, of a print bed in motion. The images are of the actual printing prosess, in a poor lit room.

    Image of support and skirt:
    [​IMG]

    Then perimeter is added:
    [​IMG]

    Last stage is adding the infill. Note that there will be glitches, or areas that cannot be filled perfectly:
    [​IMG]

    The filament is forced onto the print plate, and the textured pattern of the plate, hides most of the individual lines of plastic. I find it particularly nice for mesh patterns. Looking closer, you will see a jagged pattern in the plastic, which is a result of using a stepping motor to move nozzle.

    For this print, I use 4mmx4mm square holes, separated by 2mm of filament. I print them at a 45° angle. This results in the printer first printing the perimiter, which is a 4x4mm box, then a single line of filament of infill, between the boxes. The infill line, is the one that nails the finish, as it ensures that the lines of the squares blend, on the surface side. The surface side is facing the print plate.

    Image of mesh square perimeters:
    [​IMG]

    Up close, this is a pretty rough process. Things are not perfect, and imperfections are plenty. But given the scale, this really does not matter.

    Image of mesh squares with infill:
    [​IMG]

    This technique, leaves a first layer, with a lot of small perimeter parts. The printer needs to retract the filament in between printing these, as not to spill filament all over. This is not a perfect science. Stringing will occur, as even if the printer pulls the filament back, it cannot retract melted filament. Some stringing is expected.

    Image of stray string, before it is baked into the printed object:
    [​IMG]

    Prusa has just recently updated its software and firmware. An ironing feature is added, in which top layers are ironed for a smoother surface. This is done by running the nozzle across the top surface a billion times, adding very little filament. It estimated about 5 hours added print time for this piece. Also, Prusa warns about added reliability issues. It might work, but it adds 25% of print time and irreliability. I am not trying it out for this piece, this time around. Other slicers has pioneered this technique, and to be honest, I was unimpressed with the results, particularly given the added print time.

    Edit:Replaced oversized image.
     
    Last edited: 3 Feb 2021
    Cheapskate likes this.
  9. Frode Bergeton Nilsen

    Frode Bergeton Nilsen Minimodder

    Joined:
    15 Jan 2021
    Posts:
    46
    Likes Received:
    54
    Edit:
    Tried to edit the above post. This image need some rescaling:
    [​IMG]

    The one in the post violates the guidance size for the forum. Just delete this post, but please fix the img url, as not to break the forum engine
     
    Last edited: 3 Feb 2021
  10. Frode Bergeton Nilsen

    Frode Bergeton Nilsen Minimodder

    Joined:
    15 Jan 2021
    Posts:
    46
    Likes Received:
    54
    I started out using a default Prusa MK3S printer, and even Prusa filament. I simply could not get it to print. Layers would not adhere, and if there was a hick-up in the print, it would grow, and not fix itself. For the structure to be strong enough, I worked for months to get consistent and strong results.

    Once again, the images are macro shots of a moving printbed, in poor light.

    First of all, infill needs to make sense. Layers need to adhere. I use a simple pattern, as it prints reliably, with consistent strong and strong adhesion. If there is a hick-up, it self corrects, in a few layers. I simply could not get great results with a .4mm nozzle. It is not just a question of percentage, but more a question of what can be printed reliably, and give the expected strength consistently.

    Image of 40% infill pattern:
    [​IMG]

    Second image of infill pattern:
    [​IMG]

    I use an unusual thick perimeter. There are many reasons for that. The main being strength. The other one is that thinner walls tend to reflect the inner structure on the outside surface. In particular, looking at the embedded nuts, anyone should be able to see, that perimeter is far stronger than this infill pattern. 4 layers, is not really that thick. Printing this at the default 2 layers, is simply not strong enough in my opinion. Particularly when facturing in imperfections. Also note how clean the side of the printed part is. The height of a layer, is 0.3mm.

    Image of perimeter of through hole and outer edge, bounded by infill:
    [​IMG]

    I also manged to grab a somewhat sharp image of the top of the printed mesh. It is stringy. Print lines are clearly visible. There are plenty of air pockets. The surface is simply ugly. This is why all my designs either prints mesh fasing the print bed, or being printed vertically. Rectangles or squares at an angle, is one of very few mesh patterns, that prints both horizontally and vertically. I consider this result to be a really nice and consistent print, as in a best case scenario.

    Image of top side of the printed mesh:
    [​IMG]

    There is like a ton of hard constraints, when 3D printing a case like this one. Getting past them constraints, and managing to finalize a given expression, is really hard. The end result is also pretty far from what people are accustomed to, particularly for surface finish. I would think most people can agree on that mesh looks ugly on the top side.
     
  11. Frode Bergeton Nilsen

    Frode Bergeton Nilsen Minimodder

    Joined:
    15 Jan 2021
    Posts:
    46
    Likes Received:
    54
    The print turned out fine. No leaks, small amount of stringing.

    Image of printed part, straight of printer:
    [​IMG]

    Next is removing supports, stringing, and tapping.

    Image of tapping bit:
    [​IMG]

    Image of tapping tool:
    [​IMG]

    Image of tapping tool and part
    [​IMG]

    Image of closeup of tapping the part (note the curved tip):
    [​IMG]

    I also forgot to factor in the physics of the tapping tool, when modeling this part. It is a bit tight, but worked out using tweezers.

    Image of tapping tool and hard to work with holes:
    [​IMG]

    Image of how this was solved, to get enough torque:
    [​IMG]

    I do not use two of the holes, and will be removing them in future prints. They are covered by the brackets.

    Since PETG can be worked with, tapping works just fine.

    Image of tapped hole:
    [​IMG]

    I also snapped some shots during printing this part. I always leave a section that is unthreaded, as to ensure that all threads are up to spec, and accessible for the tapper. The printed threads are not that nice, which is why I need to tap.

    Image of hole with threads, during printing:
    [​IMG]

    Surface is different as described earlier.

    Image of unfeatured top surface:
    [​IMG]

    Fetures like holes, introduces challenges, as how to print the layer. It will result in patterns, that will reflect light differently on the top surface.

    Image of featured top surface:
    [​IMG]

    The print bed is textured. Thus, the side facing the print bed, has a nice texture, particularly the mesh.

    Image of print bed finish:
    [​IMG]

    Image of print bed facing side and mesh:
    [​IMG]

    The side finish, is more revealing. The bridging part on top of holes, can result in some uneven lines. The smallest imperfection will easily show. The finish is generally nice, but hardly ever perfect.

    Image of side finish:
    [​IMG]

    I took some more images of the top side of the mesh, this time with better light and smaller aperture, and not to forget, a still subject.

    Image of top finish of mesh:
    [​IMG]

    Given the height of this part, warping is expected. This time around, the print liftet of the textured plate, and did not drag it of its magnets.

    Image showing warping of corner:
    [​IMG]

    Curves do not print gracefully vertically. There is a slight droop in the top section of a circle, and at an angle, the print is jagged. Much like trying to build a circle vertically with Lego bricks. I just live with it. People do not really notice, unless this is pointed out. My printed designs are a bit squary, but this is reason for that.

    Image showing the slight droop and circle cutoff on top of vertical printed hole:
    [​IMG]

    Images showing the jagged edge resulting from layering, on holes printed vertically:
    [​IMG]

    Although this build my appear bulky, there are some areas of concern. In some areas, there is only 6mm x 6mm, at worst, only times 4. Like the top part, with holes on 4 sides, or in the upcoming crossbars.

    Image showing example of 6mm x 6mm corner (taken on moving bed, live printing):
    [​IMG]

    This is also the bigger part of this build. The printer cannot really print that big parts, and those printers that can, either costs a fortune, or have their issues, particularly for PETG. The small print size is the main challenge, trying to print something as big and robust, as this.

    That is pretty much it, as for printing tech. Thought I rather just get printing parts, sharing them here, and assemble the thing. Please do ask along the way, if you got questions.
     
  12. Frode Bergeton Nilsen

    Frode Bergeton Nilsen Minimodder

    Joined:
    15 Jan 2021
    Posts:
    46
    Likes Received:
    54
    The front crossbar is done. It only took 4hours and 41minutes. That is excluding the screws.

    Image of the crossbar:
    [​IMG]

    For consistency, it has the same surface as the corner part. The result is not anywhere near the typical result in this forum, but it is pretty calm, and about what to expect of a printed part. In all fairness, a regular steel chassis, is far more messy and uglier than this, internally.

    Image of the top surface of the mounted crossbar:
    [​IMG]

    Image of progress thus far:
    [​IMG]

    I also had my PC die on me. It turns out the MB is a dead fish. That slowed the progress down for the day. The brackets will not be done today.
     
    IBMer and Cheapskate like this.
  13. Frode Bergeton Nilsen

    Frode Bergeton Nilsen Minimodder

    Joined:
    15 Jan 2021
    Posts:
    46
    Likes Received:
    54
    Two brackets, a cover, and a crossbar is finished, along with some screws.

    First off, was the brackets. The curve is ugly and jagged, as expected, on the top surface. It is basically hidden, so I will leave it as such.

    Image of the two brackets:
    [​IMG]

    As said earlier, things needs to be split up into multiple parts. The brackets has one sentral screw for mounting, and it cannot spin, as it is flush with the mesh part. It is straight forward to mount.

    Image of mounted brackets:
    [​IMG]

    Then there is the top mesh part cover part. It serves as a key ventilation vent, as an inlet for cold air.
    [​IMG]

    When mounted, it makes a plane surface. There is the issue of drooping when printing using support, which is why the screws are not perfectly flush. Still, the use of repetitive patterns and few but clear lines, results in a relaxed end result. Despite the differences in texture. The vented section is print-surface-finish, but it will be a module at that. The rest of the plain, will be print-top-surface-finish.

    Another image showing the print surface differences.
    [​IMG]

    I also got another crossbar done. It is pushed about as far as my printer allows.

    Image of crossbar finished on the printer bed:
    [​IMG]

    The crossbar cannot be mounted, without removing the mesh cover. Please note that the surface is print-bed-surface, but that it does not matter, as it will be hidden by the rads.

    Image of mounted crossbar, with mesh removed:
    [​IMG]

    The bottom and top of this build, is the exact same parts. For a printed part, I really like the finish coming of the Prusa, using the textured print plate. Given that this is a Swiss cheese, the nuts blend in nicely. It is not this blending polished glossy finish of a lot of the work in this forum, but for what it is, it is really nice. The main benefit of 3D-printing, is the ability to build features into the parts. I have never seen any design for cable routing like this before. The entire concept is new. As is the idea of using a cable channel, as an air vent. Unless you flash a light at it, the space between the mesh is hardly lit at all, and the mesh is thick, thus it is hard to see what is inside.

    Image of the outer top/bottom surface:
    [​IMG]
     
  14. Frode Bergeton Nilsen

    Frode Bergeton Nilsen Minimodder

    Joined:
    15 Jan 2021
    Posts:
    46
    Likes Received:
    54
    Two more parts are done. The crossbar for the MB, and the middle internal mesh part. It is easier to explain these parts, when the building in the case. Do note that the middle channel of the MB structure, is a channel for external cables, running straight through the entire case. It is really easy to work with. And again, I have never seen anyone doing anything like this.

    The crossbar printed in 20 hours and 57 minutes. The mesh part in 5 hours and 23 minutes. Screws are on top of that.

    Image of the build so far, from the rear:
    [​IMG]

    Image of the build so far, from the front:
    [​IMG]
     
  15. Frode Bergeton Nilsen

    Frode Bergeton Nilsen Minimodder

    Joined:
    15 Jan 2021
    Posts:
    46
    Likes Received:
    54
    I am still waiting for the delivery of the V6 heatbreaks, the original ones that work, and not the problematic ones that Prusa uses. Also, I still only got one .6mm nozzle, thus the slow progress. I am also low on socks.

    The mirror corner part is done. It only took 20hours and 57minutes to print, for this part alone.

    Image of the inverse corner part:
    [​IMG]

    The internal middle mesh part, needs to be installed, after mounting both the crossbars. The supports on the middle mesh parts, interfere with the nut for the crossbar mount. Also, it is only the two nuts, at the end of crossbar, that joins the outer parts. Designing this thing, I had no idea what to expect as for strength for such a joint, but it turns in pretty solid. Also, do note how weak the crossbar is, structurally, due to the introduction of the external cable channel. The covers mounted too it, significantly strengthen it.

    Image showing the Motherboard crossbar, and the joints with the corner parts:
    [​IMG]

    The internal mesh part, needs some extra support, as not to bend. The support, blocks the screws for mounting the crossbars. The result of this design, are joints bewteen the meshes, that are nicely aligned.

    Image of the middle internal mesh part, from the bottom:
    [​IMG]

    I also discovered that warping reared its ugly head again. The outer crossbar had warped, but only in one end. Just show how unpredictable this is.

    Image of crossbar corner warping:
    [​IMG]

    Image of crossbar hardly warping at all. Same part:
    [​IMG]
     
  16. Defyant Mods

    Defyant Mods Multimodder

    Joined:
    14 Sep 2004
    Posts:
    1,305
    Likes Received:
    780
    Amazing work so so far :rock::rock: it's really is a fantastic design :thumb: and it's great to see the Prusa printer being pushed to it limits.

    One of the very first things i tried to print when i got my printer was a 220 x 220 Hex mesh grill.... with very similar results as yours.... warping was my biggest prob. Eventually i re printed and used a brim slowed the initial layer down got my retraction just right setup Z-hop (this cased a bit of stringing) dropped my print speed and layer height. This gave me a much better print :thumb: (but took longer) Even a month later i'm still dialling in for that perfect print :D

    For anything that's important now i prints on rafts... or at least use a brim that stops the warping of far ends. I'll then trim that edge perfect on the router :)

    Again this is an amazing working log:rock:
     
  17. Frode Bergeton Nilsen

    Frode Bergeton Nilsen Minimodder

    Joined:
    15 Jan 2021
    Posts:
    46
    Likes Received:
    54
    Thanks. Really appreciate it. I have to say, looking at some of the amazing other blogs in here, with their insane artistic work, this feels like a boring project compared. Hopefully people get it in the end, but hey, these other blogs are insanely great. If people like my work in this forum, I really appreciate it.

    I've been pushing it into the night, to start of a 15 hour print. This it how far I have gotten thus far:

    Image of progress at this point, from the front:
    [​IMG]

    Image of progress, from the rear:
    [​IMG]

    Image of side view of cable channel:
    [​IMG]

    Image of progress, bottom view:
    [​IMG]

    To access the cable channel, the middle cover is removed. The side of channel is not used. The channel is pretty generous in size, and thus should be pretty easy to work with, despite the limited access to it. Also note the the opening for the external cable channel. The design goal, is to stuff the fan or RGB cable spaghetti into these channels, making the rest of the cabling less cluttered and easier to work with. Keeping the external cabling separate as well, just makes everything so much easier to deal with. This is a design, that ground up deals with all cabling issues, including the external ones. It took like forever to get to this point, and again, I have not seen anyone using these concepts anywhere at all. Also note how chick the mesh is. That allows for air to pass through, for the part to more stiff and durable, but also prevent transparency. That can be seen clearly in the angled view below.

    Closeup image of the cable channel:
    [​IMG]

    I seriously need to call it a day.
     
  18. Frode Bergeton Nilsen

    Frode Bergeton Nilsen Minimodder

    Joined:
    15 Jan 2021
    Posts:
    46
    Likes Received:
    54
    Quick progress update.

    The 15 hour print for the front corner is done.

    Image of front corner part:
    [​IMG]

    Image of progress thus far, from the rear:
    [​IMG]

    Progress from the front:
    [​IMG]

    Note that 10mm is added to dept in the front, to support taller GPUs.
     
  19. Frode Bergeton Nilsen

    Frode Bergeton Nilsen Minimodder

    Joined:
    15 Jan 2021
    Posts:
    46
    Likes Received:
    54
    Today was a bit of a setback. I had to fix my daily driver, which involved building a new computer from scratch. Not realizing the clock was off, Fusion 360 refused to logg in. The clock was off, due to a brand new MB. Seems like I got on top of it. Also, I printed some screws last night, and there was a crash during printing, as there frequently is, when printing the screws. As a result, I needed to re-tighten the nozzle, and clean the print head. Being as delayed as I were, I did not document it. There is plenty of users struggling with leaks, using the E3D V6. There are plenty of people claiming to have a fix, but it is pretty obvious that a lot of are applying every alleged fix there is, and still the thing leaks on a regular basis. Like for me, after 5-6 days of constant printing. My understanding is that the nozzle is hammered into leaking, when there is a crash during printing. Looking at this project, people might wish for a bigger printbed. That will result in like 3-4 days of print time, and it only takes a few crashes, to ruin the print.

    Anyway. The printer is back at it, and I have a working computer. Update will follow tomorrow. Pics lacking today, as the need top push things back up running again. This is how it has constantly been for me, since I started printing. I have learned to expect something to break all the time, as that is the norm. I also had some nuts having a crash two days ago. That resulted in a layer shift, that is, the printer shifts the x, y, or both planes.

    Image of nut with shifted threads, lines illustrating the shift:
    [​IMG]
     
  20. Cheapskate

    Cheapskate Insane? or just stupid?

    Joined:
    13 May 2007
    Posts:
    12,411
    Likes Received:
    1,968
    Ah, Yes. Windows 10 reduces the control on the clock to a, "let windows automatically adjust the time and date," checkbox, then can't do that because the security settings won't connect to anything with a different time. :lol:
    I hear they plan to delete the legacy control panel...
     

Share This Page