The Custom Bracket That Cost Me Three Weekends

I was building a small differential-drive robot. The motors I'd sourced were N20 gear motors — great choice, common, affordable. But the chassis I'd laser-cut had mounting positions that didn't quite match the standard N20 brackets available online. I needed something custom: same basic shape, slightly different hole spacing, one extra M2 mount point.

This should have been a two-hour problem. It turned into three weekends.

Option 1: Order a standard bracket and modify it

The most obvious first attempt. The standard N20 bracket is cheap, maybe ₹20-30 each. I ordered a set, measured carefully, and started filing. The result was functional but ugly, and the filed edges were not exactly strong. On the third test run, the bracket cracked at the weakened point.

Metal filing is actually quite tedious. And the tolerance you can achieve with hand tools is not what you'd want for something load-bearing.

Option 2: Find someone with a 3D printer

I knew someone who had a basic FDM printer — one of the common sub-₹20,000 models that were popular a couple of years ago. He was happy to help. The design took about four hours in FreeCAD (I was learning as I went). We printed the first version, and the hole positions were off by 1.5mm — my measurement error. Back to FreeCAD. Second print was closer but the wall thickness I'd specified was too thin and the part flexed noticeably under load. Third print was right.

Three prints, plus the time to iterate on the design, plus coordinating schedules with my friend — call it a weekend. Not terrible, but also not nothing.

"The actual design time was maybe 6 hours total. The coordination, the waiting, the iterations — that was the weekend." — my notebook from that project

Option 3: Send it to an online service

After the first two attempts, I started looking at online print-on-demand services. Most of them are fine, but the logistics were discouraging. Minimum order values. Unclear pricing until you upload the file. 5-7 day delivery. The standard 0.2mm layer height I needed for functional parts was available, but the page to specify exactly what I wanted — material, infill, post-processing — was confusing enough that I wasn't confident I'd get what I expected.

And I needed two brackets. Just two. That constraint made most services awkward — either there was a minimum quantity, or the per-unit cost at low quantities was more than the project budget warranted.

The thing nobody tells you about custom parts

The friction isn't really in the design, once you've learnt the tools. It's in the gap between 'I have a design' and 'I have the physical part in my hand'.

For standard off-the-shelf hardware, this gap is 2-3 days and known cost. For custom parts, it's a variable — it depends on who you know, what equipment they have, their schedule, their filament stock, the print time, the queue ahead of you.

This is why so many makers end up owning their own 3D printers even when they print maybe once a month. The value of the machine isn't in cost savings. It's in eliminating the gap. You design something at midnight and you have the part by morning.

What changes with accessible local printing

Not everyone should buy a printer. For occasional use — one or two custom parts per project, a few projects a year — the maths doesn't work. The machine costs money, takes up space, requires maintenance, and has a learning curve.

What works better for occasional users is a print service that's close enough to remove the logistics friction. Not 5-day courier — same-day or next-day, where you can describe what you need in plain language, know exactly what you'll get, and have the part in hand before the project momentum stalls.

Three weekends for two motor brackets is too much. It shouldn't be that hard.