Online 3d software apps are horrid

So i am building a RPi project, and the system is built but i need a housing for it.
I used SketchUp's online modeling app...with great difficulty and made a somewhat rudimentary design.

Working on making the walls and a screw on lid for it. Minor tweaks here and there. If possible, would anyone be able to print me an acual copy once the design is finished? I would of course pay for materials used.

Any help would be greatly appreciated.
I had hoped to have this built before my dad passed, but unfortunately it didnt go as I had hoped. So completing this is a big deal to me.

the site won't let me upload the .skp file. I did try to zip and rar it, but nothing even tho it is less than max allowed size. If i could email it to someone, or if there is some other way to upload it. Please let me know :/

SketchUp is possibly the single worst application one could use to create a model with the intention to print it on a 3D printer. You should have your model in STL format or OBJ format and use a 3D modeling program that has a focus on 3D printing.

Even TinkerCAD is a better program than SketchUp and will generate exactly the type of file needed for 3D printing. It's a bit easier to use than SketchUp for basic shapes.

If you're determined to stay with SketchUp, you can find a method to generate an STL file here:

https://airwolf3d.com/3d-printing-ti...l-3d-printing/

It's especially important to test the model generated by SketchUp for manifold and any other 3D printing failures. There are online 3D model testing sites.

Attachment 11374

Completed the walls and supports. Here is the STL.

I have modeling software somewhere, and Sketchup(while terrible) was the only one i could get to work until i could find my software. TurboCAD.

I've pulled the STL file into Meshmixer (free) and immediately noted there are modules modeled in the box. Unless you want those printed, you'll want to pull them from the file sent to print. Also, the pillar in the center and the inside floor of the box have reversed normals. This means they are "inside out" and will cause print problems. There is also a narrow strip away from the cover and box that has reversed normals as well.

I didn't try to remove the modeled modules inside the box.

I can't advise you of the best way to put the normals back to normal, as I'm not a TurboCAD user.

Attachment 11375

I wasnt using TurboCAD yet.

Here is the remodeled enclosure. I couldnt remove the base without deleting other portion. Ended up completely rebuilding the whole enclosure being careful of deletions, and mindful of measurements.

PC is having issues downloading Meshmixer.

you want use mm NOT inches.
also have a look at Openscad - things like that are doable in minutes and it only outputs fully manifold printable models.

I also don't see any holes in the case for you to plug stuff into the board. is that deliberate ?

How big is it ? if convert directly from mm to inches its 11.7 inches by 6.26 by 2.14
That's pretty big for a pi case.

Enclosure MK2 has fewer problems, all of which seem to be the same type.

Attachment 11376

I'm hopeful that the above image can be clicked to enlarge in order for you to see the red lines indicating fail points. In Meshmixer, the automatic repair pretty much failed, by completely removing the corner pillars and all four walls. On close examination, I can see that there are overlaps of the bottom wall and the right wall, which creates a non-printable area within the corner. It appears at first that the lower wall and the left wall are separated, but they also have an internal overlap.

Are you working from an engineering drawing with dimensions and locations? If so, I would accept that information from you and create this box within OpenSCAD, a script driven descriptor type 3D modeling program. Because it is parametric, it is certain to create a manifold model ready for printing, unless there's an error in the script, of course.

I could attempt the creation of this part by guess-measuring from Meshmixer, but that's a royal pain and I'm not skilled enough in Fusion 360 to use that program to measure it.

If you intend in the future to create 3D models, Fusion 360 would be a good choice, as it is very powerful. Of course, with great power comes great responsibility. In this program, the responsibility is taking on the task to learn a complex piece of software!

@Curious Aardvark, in an earlier post, there are electronic modules (more than one) mounted in the model. I suspect the dimensions you reference may be close, although I paid little attention to that aspect.

Quote:

---

Originally Posted by curious aardvark

you want use mm NOT inches.
also have a look at Openscad - things like that are doable in minutes and it only outputs fully manifold printable models.

I also don't see any holes in the case for you to plug stuff into the board. is that deliberate ?

How big is it ? if convert directly from mm to inches its 11.7 inches by 6.26 by 2.14
That's pretty big for a pi case.

---

L=5 1/2in
W=6 1/4in
H=1 13/16in

all measuring is using ruler and caliper measuring in Inches.
As for holes, i want to build the enclosure first and then do the holes once the erros are ironed out.

trying to open the stl in Openscad but no luck

Quote:

---

Originally Posted by curious aardvark

you want use mm NOT inches.
also have a look at Openscad - things like that are doable in minutes and it only outputs fully manifold printable models.

I also don't see any holes in the case for you to plug stuff into the board. is that deliberate ?

How big is it ? if convert directly from mm to inches its 11.7 inches by 6.26 by 2.14
That's pretty big for a pi case.

---

So in the STL there is 2 items: 1 is the lid, and the second is the enclosure itself. For whatever reason Meshmixer read both pieces as one build.

@fred_dot_u
Im not sure as what all you need so:
L = 5.5in(139.7mm)
W = 6.25in(158.75mm)
H = 1 13/16(46.0375mm)
Wall, Base, Lid = 1/8in(3.175mm) thick

Blue = RPi standoffs
Yellow = Power Distro Standoffs
Orange = 2-CH relay standoffs
Black = Lid
Attachment 11378
^^That is what I am trying to make. Unfortunately, SketchUp doesn't have any way of showing overlaps or malformed planes.

You won't want to import a known defective STL into OpenSCAD. The idea is that I can help by recreating the model in OpenSCAD and provide you with an error free STL, or you can learn to use OpenSCAD to construct the part yourself.

Let's say the lid is 5.5 x 6.25 x 0.25 dimensions.

It's better to use metric in OpenSCAD, converting: 139.7 x 158.75 x 6.35, rounding for ease of use: 140 x 159 x 6. If you require fractional millimeter precision, it's an easy matter to plug the more accurate numbers into the code.

Code:

---

module lid(){
    cube([140, 159, 6]);
}
lid();

---

You don't have to use modules, but it makes things easier in the long run. Also it is a good idea to use parameters defined at the start of the code:

Code:

---

lid_w = 140;
lid_d = 159;
lid_t = 6;

module lid(){
    cube([lid_w, lid_d, lid_t]);
}
lid();

---

I wanted to expand a bit on the parameters part and got carried away. OpenSCAD allows you to pass parameters to modules when you call the module. I colored each panel differently, making it easier to see and identify them. Keep in mind that OpenSCAD doesn't add color to the model and the colors will vanish when the model is rendered (F6), but it does show when the model is previewed (F5).

Code:

---

lid_w = 140;    // also base width
lid_d = 159;    // also base depth
lid_t = 6;      // also wall thickness
wall_h = 25;    // arbitrary guess at wall height
addabit = 0.1;  // fudge factor for clean unions

module lid(color_choice_lid){
    color(color_choice_lid)
    cube([lid_w, lid_d, lid_t]);
}

module left_wall(color_choice_left){
    translate([0, 0, lid_t - addabit])      // elevate wall to be not quite above floor
    color(color_choice_left)
    cube([lid_t, lid_d, wall_h]);
}

module right_wall(color_choice_right){
    translate([lid_w - lid_t, 0, 0])        // use left_wall module, shift it over
    color(color_choice_right)
    left_wall();
}

module front_wall(color_choice_front){
    translate([0, 0, lid_t - addabit])      // same elevation as before
    color(color_choice_front)
    cube([lid_w, lid_t, wall_h]);
}

module back_wall(color_choice_back){
    translate([0, lid_d - lid_t, 0])
    color(color_choice_back)
    front_wall();
}

lid("black");
left_wall("yellow");
right_wall("green");
front_wall("blue");
back_wall("red");

---

Attachment 11379

The great advantage of using parameters is that you can change the value in the assignment area and IF the code is well-written, everything adjusts properly. A good way to test code is to change these figures to ensure that everything else does change appropriately.

I'm not a code wizard, but find this program to be great fun for the brain cells and very useful overall. There's a strong possibility that just about anyone else familiar with OpenSCAD can improve on my thrown-together code. I didn't put a lot of work into it and may have made some trivial errors, but the model generated is manifold and fits the parameters. Advanced coding escapes me.

I'm not sure how I missed the previous message, but it's not a great problem, as one can easily replace the figures in my code with the corrected ones.

The next question at hand is what diameter are the stand-offs (corner rectangles, center cylinders) and the holes within?

Use the lower left corner of the box, the outside point as a reference for the center of the holes and measure x and y from there. Also include height for obvious reasons.

I'll modify the code I just placed to match the information provided and re-post, as it's not going to be particularly large.

Attachment 11384
Here are the measurements and conversions...as best i can make em lol
(They are really strict about file sizes lol)

The drawing is kinda shaky as an engineering drawing, but I'm making the best of it. A well-drawn model will have all dimensions referenced from a single point, typically for this one, it would be the lower left corner. That's not a big deal, as I can do the math and get correct results. Questions to be answered are preceded by stars **.

There are some missing dimensions and some ambiguous ones as well. The answers may be in the drawing, but they are pretty well hidden/obfuscated/located in an unusual manner.

You have a 1/4 square block in the lower left corner. The diagonal dotted line with 3/8 at the end of it is ambiguous. The diagonal of a 1/4" square is 0.3536 inches, not 0.375 inches. There's no need to have a diagonal dimension if the other two are provided. ** Please confirm that the 3/8 dimension is meaningless?

Please also confirm the same for the right side corners, top and bottom. The diagonal of a half inch square is 0.70 inches, not 0.75 as shown in the drawing. ** Another meaningless dimension?

** Does the 3/16 center notation in the lower left corner mean that the hole is to be 3/16 inches diameter? I hope it's not that large, because that leaves a too-thin wall on the standoff, a thirty-second of an inch at the very edges.

The four standoffs in the lower right corner are almost clear. The dimensions of the standoff are in a good location and the y-spacing and x-spacing for the set as well as the y-delta are in the drawing, but it seems to be missing the x-delta, the distance from the edge of the drawing (left edge or left inside, either one). As the rest of the drawing references the inside of the walls, let's be consistent and provide that figure based on that reference.

** The upper left standoffs have most of the necessary information, but in this case, it's missing the x-delta and the y-delta figures. Again, it should be referenced from the bottom and from the left, but if you provide it from the top and from the left, I'll do the math.

** Can you clarify the note to the far right of the drawing, "standoff inner = 1/4, center 3/8 from inner corner" ? To which standoff(s) does this apply? One typically does not measure on a diagonal, so I would not apply it to the dotted diagonal lines in the corners. If this is a reference to holes in the right side standoffs, it's peculiar, as one would simply center the holes in the square cross-section.

I've been running over the drawing for too long tonight and will have to get back to it tomorrow. I'm sure there will be more questions, but you've given me a good bit with which to work.

Quote:

---

Originally Posted by fred_dot_u

** Please confirm that the 3/8 dimension is meaningless?

---

This dimension is the distance from the bottom left inner corner, to the top right of the bottom left standoff. Sort of a double check during my SketchUp work to verify that the measurement for both sides is the same.

Quote:

---

** Another meaningless dimension?

---

Again just a double check for accurate measurements.

Quote:

---

** Does the 3/16 center notation in the lower left corner mean that the hole is to be 3/16 inches diameter? I hope it's not that large, because that leaves a too-thin wall on the standoff, a thirty-second of an inch at the very edges.

---

3/16 is the distance from the bottom left inner corner to the center of the standoff. Another double-check during SketchUp work.

Quote:

---

** The upper left standoffs have most of the necessary information, but in this case, it's missing the x-delta and the y-delta figures. Again, it should be referenced from the bottom and from the left, but if you provide it from the top and from the left, I'll do the math.

---

In Fig. B: the dotted lines just intersect for the final standoff location. Same as Fig. O.

Quote:

---

** Can you clarify the note to the far right of the drawing, "standoff inner = 1/4, center 3/8 from inner corner" ? To which standoff(s) does this apply? One typically does not measure on a diagonal, so I would not apply it to the dotted diagonal lines in the corners. If this is a reference to holes in the right side standoffs, it's peculiar, as one would simply center the holes in the square cross-section.

---

Another double check: from top/bottom respective inner corners to center of the standoffs. SketchUp camera views arent great even when i can control em lol.

It may be double work, but any time i do measurements i measure more than 2 or 3 angles to verify points are correctly plotted.

Wow. I'm not sure how you got SketchUp to tell you that the diagonal distance of a 1/4" square is 3/8". That's not a double-check, that's a contradiction. Either the diagonal distance is 3/8" and the edge of that corner square is 0.2652 which is 0.0152 larger than 1/4" or the diagonal distance is 0.3536", not 0.375" and the corner square has a side of 1/4". Here's the key though. The variation of those numbers is so small that it won't really matter to the construction. I'll stick with the orthogonal figures, as diagonals are not used for measurement in the typical engineering drawing and as you can see, really throw off the references.

Attachment 11385

I'm hopeful the drawing above will enlarge well enough to view. The hole in the standoff is 1/8" and is centered in the square. The two large standoffs have no ambiguities. They are defined for inside and outside diameters and placement from the bottom and placement from the left. ** Please confirm hole diameter for corner standoff. **

Attachment 11386

This one is a bit more undefined. There is a measurement for placement from the bottom, but ** no measurement for placement from the left or right **. Hole diameters are clearly understood. The hole diameter in the square corner standoff is 1/8" and centered.

Attachment 11387

This array is lost in space. All that is needed to anchor it is ** the measurement for distance from left side and the distance from the top (or from left side and distance from bottom). **


The drawings currently do not show height of the standoffs, as that information is clear in the sketch and I kept the Solvespace images to 2D for clarity. It's easy to clutter up a Solvespace drawing with dimensions and constraints. 3D imaging in that program gets even more fun.

If you right click on the images and save them to your computer, they will be larger and easier to view.

Top left standoffs(Fig B):
From top: 7/16(11.30)
from left: 1/2(12.78)

Bottom right standoffs(Fig O):
From right: 1in(25.62)

Corner standoffs match wall height which is 47mm

Bottom left corner standoff diameter is 1/8(3.59)

Conversions are via a digital caliper.

I think those numbers fill in all the blanks. One important observation, though, is that most manufactured electronics today are on boards that are measured in metric. You've provided Imperial units, converted as you said to metric. There's going to be some errors in the conversion, especially when you 3D print this. You may want to check the manufacturer's specifications for hole placement, to see that they are likely measured in millimeters to the whole number and not to decimal equivalents from Imperial units.

Because I'm using Solvespace for the drawing, it toggles easily between metric and imperial, but the OpenSCAD model will be in decimal millimeters, to two decimal places. It's not necessary to stretch it out to three or more, simply because a 3D printer isn't going to give you that level of precision anyway. That sort of logic will probably get you through any rounding errors from imperial to metric or the reverse.

I hope to be able to put some time into the standoff placement in OpenSCAD today and will post if other questions arise.

If your calipers are converting one inch to 25.62 mm, it needs to be replaced or repaired. One inch is 25.4 mm.

I did check the RPi3 spec sheet and the numbers are correct. As for the 25.62: the lighting over my work area was dimmed from a dead bulb and i could'nt see that the caliper wasn't properly lined up haha. My bad.

I would like to thank everyone who has been helping me with this complete this project. I've learned quite a bit more than i thought i knew from my father. Biggest lesson i learned: dont FULLY trust SketchUp lol

I've learned a new feature in OpenSCAD, to make it easier to place items repetitively and perhaps recursively, although I'm not sure both of those terms apply. What's important is that it was easier.

If my understanding of our exchange of dimensions is correct, the result should match your requirements.

The zip file contains three items. One is lid.stl, a very simple rectangular solid, using the dimensions provided. The other is the enclosure mk3.stl file, including the standoffs placed appropriately. The last file is the OpenSCAD source file, with parametric code to allow you to easily adjust any dimensions as required. I did not include the solvespace diagram I used to clarify the dimensions as I felt it was superfluous in this case, as all the information in the solvespace file is included in the parameters of the OpenSCAD file. If you wish to have the solvespace file, let me know. I tend to never delete project files, even when complete, as some blocks of information can be useful later.

Please let me know if this works for you, or if it doesn't, how you'd like to have it changed, or if you want to know where to change it in the OpenSCAD file to recreate it yourself.

If I could put some real effort into studying the software, I would have possibly easily created this in Fusion 360, another great modeling program, but so much more powerful than anything else I've ever used.

This was a fun project.

Quote:

---

I've learned a new feature in OpenSCAD, to make it easier to place items repetitively

---

Care to share ?

Checks scad file.

Ah the 'for' command. Yep that's extremely useful.

That is one seriously complicated script for a box a and a few columns.

Using vectors in the loop was the new thing. I had been searching for list and it was the wrong word.

Not so much complicated as brute force dumb, I think. I really like to use parameters when writing 3D models, as I almost always find that I need to adjust something and something else goes awry from the change, unless it's parametric. Even now, I cannot be sure that the x and y figures are not swapped, but I've not heard back from the OP.

Our makerspace has a programmer who excels in Python and loves OpenSCAD. He writes code, without comments(!) that is completely obscure to me, even though he may try to explain it.

I'm not a programmer and even though I had a class in Fortran many moons ago, it's a struggle to get fancy stuff working and I'll usually just fall back to brute force.

I get the feeling that I could somehow make use of the children() feature, but lose comprehension for that bit of code almost immediately.

It's all fun.