One of the physical things I’m using in my new job, in order to access a wide range of useful software, is a small, fast external hard drive I picked up about nine months ago. I got Debian 12 installed on it but got a bit stuck trying to get Dropbox working. It turns out that I probably just needed one more reboot for everything to click together. With that sorted, I’m now using my laptop as a Debian box, giving me a platform for various technical tools and one that I can experiment on with relative impunity. However, the drive hangs off the side of the laptop on a USB cable so I decided it was time to create a holder for it.
I took some measurements, did some sketching and then set to work in OpenSCAD. I used negative modifiers in Bambu studio when I printed the prototype in order to carve out sections and keep the filament usage and time down. There were a couple of points to adjust and so I made my tweaks and set version 2 to printing. It turns out I’d missed one vital element of a compound dimension and so, although it fitted comfortably over the USB cable, the drive couldn’t slide in on top. I’m now waiting for the third prototype (back to printing a cut down version with just the dimensions I need to assess) to cool down to test.
Could I have avoid this? OpenSCAD isn’t brilliant at letting you easily measure things. Thinking about it though, I could also have created a model of the drive itself and, as long as I’d measured that correctly, I could have positioned my part and the phantom drive next to each other to visually check for crossovers.
What the second version did let me do was test one of the “joiners” you can add via the BOSL2 library. My design was going to have overhangs whichever way I orientated it, so rather than use supports, I decided to print it as two symmetrical halves. I went for snap pins, hollowing out holes in both halves and then also printing two of the small pins. They have a hollow portion in the centre so you can clip them in but they resist coming out. It might be possible to lever the pieces apart with a screwdriver but they are securely joined for my purposes.
Anyway, that third test (a mere 2g of additional filament) has confirmed that I’ve got the fix right so time to get the final version ready to print so I can take it off the printer tomorrow morning.