Archived report. These are the numbers for the HD20 (RevA), which has been superseded by the HDP30. They are still the only load-cell figures we have for any of our drives. The HDP30 has not been tested yet, so they are kept here rather than retired. Back to the HD20 build page.

HD20 20:1 harmonic drive: output spline torque test

Elliptical wave generator · PETG flex spline · FOC · 100 mm lever arm. Gearbox 1 is the 260704/260705 unit (tightened, then further super-tightened output spline). Gearbox 2 is a second printed unit (white filament) tested fresh off the printer, then again ("notfresh") after further running-in, at both 0.7 A and 1.4 A. The no-gearbox baseline is shared across all runs. Raw load-cell counts scaled ×980.7 to N·m (same factor as prior reports).

58%

best efficiency so far: gearbox 1, super-tightened, 0.7 A

53% → 48%

gearbox 2 efficiency, fresh → notfresh @ 0.7 A (-10.2% steady-state torque)

-18.4%

gearbox 2 (notfresh) vs gearbox 1 (super-tightened) @ 0.7 A

+9.3%

gearbox 2 (notfresh) vs gearbox 1 (tightened) @ 1.4 A

Summary table

Configuration	Steady-state (N·m)	Peak (N·m)	10-90% rise (s)	Duration (s)	Samples
No gearbox, 0.7 A	0.219 ± 0.004	0.224	0.27	30.0	3000
Gearbox 1, tightened spline (260704), 0.7 A	2.372 ± 0.046	2.595	0.03	17.9	1794
Gearbox 1, super-tightened spline (260705), 0.7 A	2.552 ± 0.043	2.833	0.03	18.6	1860
Gearbox 2, fresh, 0.7 A	2.320 ± 0.037	2.595	0.03	21.9	2186
Gearbox 2, notfresh, 0.7 A	2.083 ± 0.037	2.282	0.03	29.1	2914
Gearbox 1, tightened spline (260704), 1.4 A	3.009 ± 0.059	3.308	0.10	20.6	2059
Gearbox 2, notfresh, 1.4 A	3.288 ± 0.062	3.528	0.08	14.9	1497

Steady-state = median of samples above 50% of peak; ± is one standard deviation over that plateau. Torque multiplication and efficiency are each computed against the shared no-gearbox 0.7 A baseline (ideal ratio 20:1).