I made a B002T002 and this is my feedback

It didn’t turn out quite perfect. I think there are some errors in the model. First some pictures. :slight_smile:

Here it is, straight from the printer. I won’t name the printer since I don’t want to endorse the company behind it. I used regular PLA with 0.3 mm layer height and automatic support structures; quite coarse but I didn’t want to babysit the printer all day (it’s a shared makerspace 3D printer). As you can see there is a raft structure underneath the model which resulted in a prettier bottom side, but caused some problems with corners lifting from the bed.

Here you can see that the model sticks fairly well to the raft, only lifting a little bit. The raft itself however came apart during printing. It actually didn’t stick together very well from the start. When I print the other half I probably wont use a raft.

Now for the problems with the model I found.

Here I’ve put a power and a core module inside the case. Both the LED strip and relay connector are inserted fully but as you can see the locking clips on the relay connectors do not fully engage.

The design thickness of the enclosure walls is 1.6 mm. Mine turned out slightly thicker; about 1.65-1.7 mm. It consists of four parallel layers and no infill so the walls are not very strong.

With the connectors fully inserted (with the boards outside the case) I measured 1.7 mm between the power module board edge and the LED strip connector, so that one is fine. But between the board and the relay connector there is only 1.15 mm. An enclosure that allows the connector to be fully inserted would need to have walls that are thinner than that which would be very brittle. So for the next revision of the power module (or perhaps the upcoming quad relay module :wink:) I would recommend that the connectors are moved closer to the edge so that they stick out farther from the board.

Here is a view of the front with the modules inside. The DC jack is centered in the hole, but the other three connectors are about 1 mm too low. The relay and LED strip connectors can be inserted anyway, but not the USB connector. The distance between the center of the DC and USB connectors with the boards stacked slots is about 7.3 mm, but the distance between the centers of the two slots in the model is about 9 mm. In the STEP model the design distance between the two boards is 13 mm but the distance I measure between my boards is 10 mm.

A diagram of the dimensions I describe:

I fixed up my enclosure by mounting the power module on four washers, raising it by 0.5 mm. I also carved out a few layers from the USB slot. The result looks like this:

Now, I could try to correct these errors in the STEP file. But for reusability in other models it would be better to make these corrections in the source file (which hopefully is even drawn using constructive solid geometry). Would you please consider sharing these source files?

Thanks a lot Anton for your contribution. I sent STP models on your email. Regarding reported bugs - it’s our mistake, we have changed connectors and didn’t change models. Right now we are working on a new system which will better fit our supporters demands, even without a need for 3D printing. I will share a picture of our new concept soon.

Thanks. I got your files and will have a look soon.

Better late than never as they say. I’m working on correcting the CAD models now, based on measurements of the modules I bought in March. When I’m done, where should I publish them? I think they belong in version control so I could maybe make a pull request against the bc-hardware repository? What license applies to these models?

I have a version of the power station now that I’m somewhat confident will work. It is not yet test-printed so take it with a grain of salt.

http://a360.co/2u2egeO

Autodesk Fusion 360 seems to be what everyone use these days. The link above should allow you to download the model in their proprietary format, or export to STEP which will lose some information in the process. Some of the most significant changes I made:

  • Edited the models of the BigClown modules to match measurements of my actual modules.
  • Added buttons to the core module.
  • Increased the case thickness from 1,6 to 2,4 mm.
  • Added simplified (blocks) models of the connectors that plug into the board, to check clearance around the connectors. I used these to reduce the case thickness just around the connectors, so that they should plug in all the way.
  • Simplified the screw holes. They had too fine details before that are impossible to print with an FDM 3D printer and only complicated the model.
  • Reduced the screw hole diameter slightly to prepare for tapping threads in the plastic instead of using self-tapping (wood) screws. I also increased the wall thickness around the screw holes.
  • Removed the connector holes on the back. If/when a quad relay module is released the cut outs can just be put back.

I used 3D Hubs’ knowledge base as inspiration and their rules of thumbs for designing 3D printable objects. Especially their first and second guide for electronics enclosure design.

I’m somewhat ashamed to say that I have gone all-proprietary software with this project. I surveyed the landscape of open source CAD-for-manufacturing modelling software (a completely different category than CAD-for-graphics) that supports STEP import and it appears that the best one is FreeCAD, which seems to still be bad though.

Working with STEP models was not very pleasant. The format retains some information about the shapes of things (unlike STL which is completely useless for making modifications). The design intent is mostly lost however, so what I had to do in most cases was to first simplify the design to mostly rectangular shapes by removing fillets, splitting bodies into several etc. Then I could manipulate those to the shapes I wanted and finally put the details back. Unlike when a model is designed from scratch, Fusion 360 doesn’t retain any design history when working with STEP imports. So when I wanted to change some dimension again later, I had to repeat part of the process of removing details, editing, adding details.

I am currently working on an enclosure for a metering station similar to B001T001. I design it from scratch based on the models of BigClown’s modules that I improved. This enables Fusion 360’s “time machine” edit history which provides a much smoother designing process. This model is available here (still work in progress): http://a360.co/2eRpxg5

I intend to finish the meter station enclosure and possibly one more (I have USB-dongle-like enclosure for a single core module in mind) before I send them off to a professional 3D printing manufacturer in my city to have them printed (the makerspace I went to is closed for the summer).

If you have a A360 account I’ll be happy to invite you to my “BigClown Enclosures” project. I think that will also allow you to “fork” the project. Apologies that this turned out to be somewhat of a commercial for Autodesk – that was not my intention.

So I bought a 3D printer. It’s also made in Czech Republic.

IMG_0482

I finished the meter station enclosure. It came out great on the printer at 200 µm layer height in Prusa PLA.

IMG_0485

There are two design errors in the model, which I haven’t fixed. The first is that the screw standoffs are only attached to the floor and not to the walls. Apparently I used the wrong sketch line. You can see it in this picture:

IMG_0483-1

Still holds together fine. Tapping the holes with a M3 tap before screwing worked great.

The other error is the distance between the two rows of tag headers, which is off by about 0.5 mm. Fixed it by carving off some plastic on one side to widen the holes.

The tags hide the ugly knife scars just fine:

I thought a lot about designing fancy hinged doors on the bottom for easy battery replacement, but ended up just leaving open holes on the bottom instead. The case will be mounted on a wall so the bottom doesn’t need to be fully covered.

Here it is mounted against the outside wall of my apartment. It’s under a roof so it won’t get wet from rain. The side flanges allow me to take the case down easily without undoing the screws.

I’m very happy with the result. The 2.4 mm wall thickness makes the box a lot more sturdy than the old one I made from Bigclown’s models, which had only 1.6 mm wall thickness.

Because it works and I have no need of more meter station enclosures at the moment, I have no plans to fix the errors and print another. I think it is better to have a tested version with known errors published than an untested version that may or may not be fixed.

If someone else wants to fix the errors and print another, go right ahead! The Fusion 360 model is CC BY-SA 4.0 and is available here: http://a360.co/2eRpxg5 . The version I printed is V20. The model at that link has changed since then.

2 Likes

Dear Anton, Thank you very very much for your perfect job! You can contribute all your work to our GitHub if you want. We have new enclosures there too; please take a look here: https://github.com/bigclownlabs/bc-enclosures

Hi Anton,
enclosure looks great! You built even model with components, nice.

Is there any easy way how to export and download only stl model for enclosure “BOTTOM” and "BODY 1"from myhub.autodesk360 to direct open and print in 3D print SW?

Thanks! I might do that!

Not that I know of unfortunately. If you want to test-print the enclosure I can try to fix the issues I found and then publish STL or STEP exports of the bottom and top components separately?

Thanks for reply, I will definitely test your enclosure, when you realase STL models :slight_smile:

Try these.

I have moved the outer headers on the tags module away from the center by 0,1 mm and increased the clearance between the headers and the holes in the top by 0,05 mm, resulting in a edge-to-edge total difference of 0,3 mm. Doesn’t sound like much but it may be just what is needed. Let me know how it works out!

Version history:

  • V20: see above post
  • V21: fix screw standoffs
  • V22: move headers on tags module and cutouts on top piece
  • V23: no change

during the weekend I printed your Model. Tommorow I will post some pictures :slight_smile:
It looks very good and works well. Here is my feedback:

  • the position of holes for Tags is well chosen - Tags are fixed.
  • the high of enclosure and USB position is optimized for old version of Core Module with low profile pin headers, for new Core module, you will need to increase the height by 1.0 mm
  • slot for batteries is helpful :slight_smile:

Some pictures:

Tags fit well:

Battery folder:

Issue with new Core Module with 2.5mm bottom headers:
P2480572

P2480573

Thanks for your feedback! When I first made this I didn’t know that the core module had been updated. I’m not sure if this model can be easily made configurable between the old and new core module. That would probably require re-modeling the core module in Fusion 360 to add parameters for the PCB thickness and header dimensions. During development I used the core module from the B001T001 STEP file as template. I would rather redo this work in OpenSCAD than spend much more time with Fusion 360 but as I wrote earlier, right now I don’t need any more enclosures of this kind.

Here are all the parts I used while making this model. I have changed some dimensions according to my measurements and added some design features I felt was missing (like buttons on the core module):

and my work-in-progress power module enclosure: http://a360.co/2vNkIIy