procedure · Living document — the current source of truth. Episodes are snapshots; this stays current.

Bench Procedure — Measuring Motor Current (Neato D10 salvage)

Goal: get each salvaged motor’s stall current so the final motor drivers can be sized correctly (drivers are sized to stall, not running). Part of the Neato D10 brain-transplant project — see neato-d10-handoff.md.

Last updated: 2026-08-04

This doc was revised after the teardown photos. Two things changed: the blower turned out to have its current printed on it and to need no driver at all, and the primary method switched from live current measurement to winding resistance, which is safer and more accurate on the meter in hand.

What still needs measuring

Motor	Status
Blower/vacuum (EVERFLOW `F121225BU`)	✅ Done — 2.0 A printed on the label. Brushless PWM fan, needs no driver. Do not measure it (see below).
Roller brush (`905-0460-RoHS`)	⬜ Measure
Drive wheel L (`260-0016`)	⬜ Measure
Drive wheel R (`260-0016`)	⬜ Measure

Why the blower is exempt: it’s brushless with an integrated driver IC. Probing it in Ω mode measures driver electronics, not a winding — the number means nothing. And its rated current is already on the label.

Why resistance instead of live current

A motor’s stall/inrush spike lasts tens of milliseconds. The meter here is an MS8233A: 2000 counts, ~2–3 samples/sec. It will not catch that transient. Chasing it produces a number that’s wrong in an unknown direction.

Winding resistance gives the same answer deterministically:

stall current ≈ 14.4 V ÷ R

At stall the rotor isn’t turning, so there’s no back-EMF, and the winding is just a resistor across the supply. This is the textbook stall calculation, and it needs no power applied at all.

Meter notes (MS8233A specifically)

The 10 A jack IS fused on this unit — the panel reads MAX 10A FUSED, MAX 30 sec every 15 min (confirmed from the meter photo, 2026-08-04). That’s a safety margin, not a licence: a fuse can be blown, missing, or the wrong type, and it still only protects against a sustained overload, not the initial short-circuit spark across a 6200 mAh Li-ion pack. Still worth avoiding current mode when resistance gives the same answer — but the earlier “genuine unfused short” framing was wrong for this meter.
Lowest Ω range is 200 → 0.1 Ω resolution. Motor windings land around 1–5 Ω, so expect ~5–10% error. Fine for sizing a driver; round up generously, don’t treat it as precise.
Resistance mode uses the meter’s own internal battery. No external supply, no L293D, nothing connected to the motor.

Procedure — winding resistance

Setup

Red lead in the mAVΩ jack (NOT the 10 A jack). Black in COM.
Dial → Ω, 200 range.
Short the probes together. Note the reading — that’s lead resistance, typically 0.1–0.4 Ω. Subtract it from every measurement.

Per motor 4. Motor completely disconnected — unplugged from the harness, not wired to anything. 5. Probe across the two motor terminals:

Wheel motors: the two chunky solder posts flanking the encoder disc, where the can’s brush terminals pass through the encoder board. NOT the harness wires — those run through encoder circuitry and will give a meaningless reading.
Brush motor: its two terminals directly.

Note the reading. Rotate the shaft a few degrees by hand and measure again. Repeat ~6 times through a full rotation — commutator position changes the reading significantly.
Take the LOWEST value, subtract lead resistance → that’s R.
14.4 ÷ R = stall current.

Worked example: lowest reading 3.4 Ω, leads 0.4 Ω → R = 3.0 Ω → stall ≈ 4.8 A.

Optional — live running current

Only worth doing if you want the running figure for power budgeting. It is not required for driver sizing.

Red lead → 10 A jack, dial → 10 A DC, set before connecting anything.
In series — break the circuit so current flows through the meter: [+14.4V] → [red probe | METER | black probe] → [motor +] … [motor −] → [GND]
Wheels only may run via the Elegoo L293D (≤600 mA/ch). The brush motor: direct from supply through the meter, never through the L293D under load.
Read fast, disconnect. Remember the 30-second limit on that jack.
Move the red lead back to mAVΩ afterwards. Leaving it in the 10 A jack and then probing a voltage is the classic way to short a supply through the meter.

Turn numbers into parts

Rule of thumb: driver continuous rating ≥ measured stall current, with margin.

If stall came out…	Driver
Wheels ≤ ~1 A each	TB6612FNG (dual — one board covers both) ✅ current plan
Wheels ~1–3 A each	TB6612 marginal; consider `DRV8871` ×2
Wheels ≥ ~3 A each	`DRV8871` ×2 or equivalent — TB6612 is out
Brush ≤ ~1.5–2 A	MOSFET module, or a BTS7960 for headroom
Brush higher	BTS7960 (rated far above this); just mind wiring and heatsinking
Blower	No driver. PWM direct from ESP32 GPIO at ~25 kHz.

Once measured, update the BOM in neato-d10-brain-transplant.md and order.

While you’re in there — two free data points

Both use continuity/Ω mode, both take a minute:

Count the wheel harness wires. 6 = 2 motor + 4 encoder (quadrature, direction-aware — good). 5 = 2 motor + 3 encoder (single channel, direction-blind — matters for odometry quality and for slam_toolbox). Then use continuity to find which two wires reach the motor solder posts, and label them now while it’s obvious.
Count the blower’s wires. 4 confirms the brushless-PWM conclusion (black GND / red +14.4 V / yellow tach / blue PWM). 2 would mean it’s a plain brushed motor after all, and a MOSFET or BTS7960 goes back on the list.

Safety notes

Resistance measurements are zero-risk — nothing is powered. Do these first, and you may not need to power anything at all.
Never measure current in parallel / across the supply — series only.
Never leave the red lead in the 10 A jack when you go back to measuring voltage.
The raw 14.4 V pack can dump enormous current into a short. Keep leads tidy, connections deliberate.
Secure any motor before powering it — an unmounted motor can jump when it kicks.

Record your results here

Lead resistance (probes shorted): 0.1 Ω   ← subtract from all readings below

                  R lowest (Ω)   minus leads   → stall (14.4/R)
Drive wheel L  :  6.8            6.7           2.1 A    (2026-08-04; stable cluster 6.8–9 Ω via alligator clips, lowest stable taken)
Drive wheel R  :  6.0            5.9           2.4 A    (2026-08-04; steady 6.0–6.7 Ω via clips, lowest stable taken)
Roller brush   :  ______         ______        ______ A
Blower/vacuum  :  n/a — 2.0 A from label, brushless, no driver needed

Wheel harness wire count : ______  (6 = quadrature, 5 = single channel)
Blower wire count        : ______  (4 = PWM fan, 2 = brushed)

Notes: