Fixing Suboptimal Design

In the T48 programmer case post, I mentioned the UV eraser that came with the programmer.

I have one of those cheap drawer-style UV EPROM erasers. Typical poorly made garbage, but it works (though I sure won’t be leaving it unattended while it is on).

Yet, it was still annoying. The drawer is too shallow to accommodate anything but bare standard EPROMs. I find that I usually have my EPROMs in machine-pin sockets to make insertion and extraction from a board when I’m in the edit-build-run loop both easier and less prone to destroying pins. But that would be too tall to fit in the eraser.

Opening the eraser revealed that the original drawer was about half the height it could be as it rode on some little slides molded into the bottom of the case. Plenty of space for taller chips, I just needed a new bottom plate and drawer.

A pretty trivial design, for sure. But I didn’t actually design the replacement. I took about a dozen photos while holding a ruler up to the various relevant details and fed the photos to Codex.

OK -- new project. I have this eprom eraser. XGecu UV EPROM eraser (https://xgecu.myshopify.com/products/uv-light-eprom-eraser-ultraviolet-light-drawer-style-to-erase-eprom-micro-controllers-erasable-ic-timer-can-set-erasure-time) Utter garbage, but it works.

The drawer is unfortunately shallow. But it shouldn't be too hard to print a new base plate and drawer with more depth (I need it to be deeper to be able to erase chips already in pinsaver sockets.

[uploaded 16 photos and gave some basic dimensions in the prompt]

It analyzed the photos and asked me a couple of questions about missing dimensions and/or features I wanted. It then created, unprompted, an outline of the bottom plate with the screw holes as a test fit (which is the first thing I would have done, too).

Printed the test jig. Did a test fit. Gave Codex a few notes on screw hole adjustments, etc… and then Codex generated the final design.

Printed that and it almost worked. There was one flaw where I had failed to catch a particular design issue. A quick chat later and I had a working design.

Final drawer and bottom are printed in PLA+. I would rather have used PETG, but I don’t currently have any on hand. It will be interesting to see how well the PLA+ resists the UV, but I don’t use the eraser often enough to worry about it.

Interestingly, Codex chose OpenSCAD for this model versus using the Blender MCP.

This is a pretty straightforward design. I could easily have done this in Fusion 360 or Blender. It would have taken me a lot longer, for sure.

The model and OpenSCAD project have been posted to Printables.