Reference tables · logged on the Long Bench
Four bench-measured tables you read on the page.
Printer tolerance by fit band, metric and imperial fastener equivalents, wire gauge against current, and cost per part shown as arithmetic. Every figure carries the tool, material batch, ambient temperature and humidity it was logged under. Nothing downloads and nothing is saved — the tables are the resource.
- Published articles
- 581
- Print farm
- 4 machines
- Reference blocks
- 4
- Permalinks changed
- 0
Block 01 · 3D printing
3D printer tolerance by fit band
Clearances that actually printed without binding, grouped by the fit you are after. Each row names the machine, the nozzle, the material batch and the room it ran in. Scroll the table sideways on a narrow screen; it is not shortened here.
| Fit band | Target | Printed result | Printer / nozzle | Batch | Ambient |
|---|---|---|---|---|---|
| Free running | 0.40 mm | 0.38 – 0.42 mm | P1 · 0.4 mm brass | PLA-2401 | 21 °C / 44 % RH |
| Close running | 0.25 mm | 0.23 – 0.28 mm | P2 · 0.4 mm brass | PETG-2403 | 22 °C / 41 % RH |
| Snug / press | 0.10 mm | 0.06 – 0.12 mm (3 of 5 under 0.08) | P3 · 0.4 mm brass | PLA-2401 | 20 °C / 46 % RH |
| Threaded insert | 4.10 mm pilot | 4.06 – 4.11 mm (two needed reaming) | P4 · 0.6 mm | PETG-2407 | 23 °C / 39 % RH |
Numbers came out of the guide library: the tolerance sweep, the insert pilot test and the repeatability set that follows the print farm.
Block 02 · Fasteners
Metric and imperial fastener equivalents
Convert a callout without guessing. The right-hand column flags the sizes people treat as equivalents that will not actually interchange — flagged in the row itself, not in a footnote.
| Metric callout | Pitch | Common imperial near-match | Status |
|---|---|---|---|
| M3 | 0.50 mm | #4-40 UNC | Not interchangeable |
| M4 | 0.70 mm | #8-32 UNC | Not interchangeable |
| M5 | 0.80 mm | #10-32 UNF | Not interchangeable |
| M6 | 1.00 mm | 1/4"-20 UNC | Not interchangeable |
| M3.5 | 0.60 mm | #6-32 UNC | Close, different thread angle |
Head diameters differ
A socket cap M4 and a #8 cap head are not the same across the flats, so a counterbore cut for one will show a proud or sunken head with the other.
Thread angle, not just pitch
Metric coarse and UNC share a 60° form, but pitch differs enough that a few turns in feels tight and then strips. If it goes tight at turn three, stop.
Ream before you force it
Printed bosses tapped for a near-match are the usual failure point. Cut the boss to the metric drill size and tap metric, even if the imperial screw is what you have.
Block 03 · Electronics
Wire gauge against current
A read before the wire is cut. Two columns, because a single run in open air and the same wire bundled with others do not carry the same current. The conditions sit on the same panel as the number they change.
| Gauge | Area | Single run | In a bundle | Conditions logged |
|---|---|---|---|---|
| 22 AWG | 0.33 mm² | 3.0 A | 2.0 A | 25 °C bench, PVC insulation, short runs |
| 18 AWG | 0.82 mm² | 7.0 A | 4.5 A | 25 °C, runs under 300 mm, 2 % drop allowed |
| 16 AWG | 1.31 mm² | 11 A | 7 A | 30 °C at the bench, derated for bundling |
These are one bench's measurements with the listed insulation and ambient. Your enclosure temperature wins over this table. The build-ups are written in the guide library; the failures are in the journal.
Block 04 · Worksheet
Cost per part, as a formula chain
No calculator runs here, so the arithmetic is written out row by row. Substitute your own rates; the shape is the part that matters. Figures are in whatever unit you price filament in.
| Row | What it adds | Working | Per part |
|---|---|---|---|
| 1 | Material | 42 g × 0.024 per g | 1.01 |
| 2 | Machine time | 3.2 h × 0.35 per h | 1.12 |
| 3 | Failure allowance | 12 % of rows 1 + 2 | 0.26 |
| 4 | Finishing | 4 min × 0.30 per min | 1.20 |
| — | Total per good part | rows 1 + 2 + 3 + 4 | 3.59 |
Where the failure allowance came from: 6 of 50 prints in the batch failed the fit check, and the figure is rounded up rather than down. Batch notes are logged in the journal.
Provenance and handoff
Corrections are dated, not deleted.
Dated appendices
When a figure turns out wrong or a batch behaves differently, the correction is appended under the original text with its date. The original stays standing so you can see what changed and when.
Second-hands rule
A table only goes live after a second person reproduces it from the written steps alone, without asking the author anything. If a step needed a spoken clarification, it was rewritten first.
Where the evidence lives
Guides are the evergreen how-to; the journal is dated shop-floor notes from the same bench. The tables in the resource layer are reference, not a procedure — read the guide before you trust a number.