The most useful tool in my shop lately isn’t a wrench. It’s the combination of CAD and a 3D printer, because it ends the oldest problem in any garage: the bracket, clip, or holder you need either doesn’t exist or costs $40 plus shipping for ten cents of plastic.
The example I point people to is my firewall. It runs on a repurposed Dell thin client, and I wanted it mounted under the desk, out of the way, not sitting in the dust like every homelab photo you’ve ever seen. No bracket exists for that. Twenty minutes in CAD, a few hours of print time, and the thing mounts like it came from the factory that way.
The workflow that actually works
The process is the same every time, and it’s closer to machining than to the “download and print” stuff people associate with these machines:
- Measure the real part or space with calipers. Twice.
- Model it with parameters, not fixed numbers, so “make the slot 2mm wider” is one edit, not a redraw.
- Print a fast draft in PLA at 0.3mm layers and low infill. The draft exists to be wrong cheaply.
- Test fit, adjust the parameters, print the real one.
Rules of thumb that save the most filament: holes print undersized, so model them 0.2 to 0.4mm over the fastener diameter or drill them after. Anything that mates with another part gets 0.2 to 0.3mm clearance per side. And a printed part’s strength is dominated by perimeter walls and print orientation, not infill percentage. Four walls at 20% infill beats two walls at 80%, and a bracket printed so the layer lines run across the load direction will snap along a layer like it was perforated. Orient the part so the load runs along the layers, not between them.
Material matters in a garage
PLA is fine for the office side of the house and for draft fits, but it starts going soft around 60°C, which a car interior in summer or a spot near an engine will happily exceed. PETG handles the shop fine. For anything under the hood or in sun-baked interiors, ASA is the answer, and it shrugs off UV too. The under-desk firewall bracket is PETG; anything destined for a car gets ASA.
Most of what I design is that same category of unglamorous and perfect: fixtures for the shop, holders for tools, brackets for car projects where the original is unobtainable or made of brittle forty-year-old plastic, and parts destined for CNC where the print is the cheap way to check fit before cutting metal.
Learning CAD is the real unlock, not the printer. The printer is just the output device. Being able to look at a broken part, measure it, and produce a drawing that becomes an object is the same skill as reading a wiring diagram or a torque spec. It moves you from hoping someone sells the fix to making the fix. If you work on cars or run a homelab, that skill pays for the printer several times over.