What we are making: an assembled device in its box, and an honest written
account of where the box will fail.
The machine that made it. Four parts worth knowing.
Do it — the fit-up
Does the board go in? Do the standoffs line up with its holes?
Does the lid close, and can you get it off again?
Does the cable pass through its hole with the plug attached, or did you size the hole for bare cable?
Does the sensor see what it needs to see, through its opening?
Assemble the whole thing and power it up. Does it still work in the box?
⚠ Careful
The cable-with-plug problem catches about half of every class. A USB plug is much wider than a USB cable, and a hole sized for the cable means threading the bare end through before soldering — which you cannot do with a moulded plug.
Do it — measure the error
Outsides grow, insides shrink. Both by about the same amount.
drawn printed error
box outer L 90.0 90.4 +0.4
box inner L 84.0 83.6 -0.4
cable hole 8.0 7.6 -0.4
lid lip 83.6 83.9 +0.3
wall thickness 3.0 3.0 0.0
Note what that table says: the error is not random. Every surface has moved outward by
roughly 0.2 mm, which makes outsides bigger and insides smaller. Once you know that number
for your machine, you can design for it rather than discovering it.
The honest review
Same standard as Session 17. Not “it works” — where does it stop working?
Water. Where would it get in first? Point at the place.
Heat. Leave it powered and closed for twenty minutes, then feel the box. Warm, or hot?
Force. Squeeze it. Push the lid sideways. Does anything move that should not?
Service. Time yourself opening it, swapping the board and closing it. Under two minutes?
Sun. You cannot test this today. Say what you expect to happen, and why.
💡 Tip
Pouring water over it is a legitimate test and your teacher may allow it — with the board removed. Do the test that actually answers the question, not the one that is comfortable.
Write the three sentences
What it does. “It holds the ESP32 and the ultrasonic, mounts with two screws, and opens in 40 seconds.”
Where it fails. “The sensor opening faces sideways and rain driven at that angle will get in. The lid lip is 0.3 mm loose.”
What you would change. “A 5 mm hood over the sensor opening, and the lip drawn 0.3 mm larger.”
⚠ Careful
That second sentence goes on your exhibition card. A team who can name exactly how their own enclosure will fail is a team the panel believes about everything else.
End of Unit E
You wrote a specification before you drew anything.
You designed a two-part enclosure with cable entry, mounting and service access.
You know a printed box is not waterproof, and you designed around that instead of ignoring it.
You measured your machine's error and found it was consistent, not random.
And you can say where your own enclosure fails before somebody asks.
Your project deploys in this box. It has two weeks outdoors to prove you right.
Challenge optional — only if you finish early
Work out what your enclosure would cost to make a hundred of.