Pico Z-Axis Control for the ComMarker B4 60W MOPA

Machine: ComMarker B4 60W MOPA fiber laser · Status: concept awaiting measurements · Nothing here is a measured specification yet.

The idea in one paragraph

Keep the B4's built-in Z motor, lead screw and hand knob. Count how far that screw actually turns with a 600 PPR quadrature encoder coupled to the screw / top-knob shaft, read it with a Raspberry Pi Pico, and show the height on an OLED. The Pico can then drive the existing + / − buttons through isolated relay contacts to go to a stored height — so focus, and the small defocus offsets used for MOPA colour marking, become repeatable instead of hunted for by eye every time.

Mockup diagram: ComMarker B4 column with laser head over base plate, top hand knob, 600 PPR rotary encoder on the screw shaft, Raspberry Pi Pico box, OLED showing Z 142.35 mm, home sensor on the column, plus/minus button panel, and a workpiece on the base with caliper-part-enter-height workflow — Concept mockup (generated diagram, not a photo and not to scale): encoder on the screw shaft, Pico + OLED readout, home sensor, relay-paralleled +/− buttons, and the caliper-a-part workflow. The screw lead is **unknown** until it is measured.

1. The problem

The B4 already has a built-in motor and lead screw with +/− buttons, plus a top hand knob for fine focus. What it does not have is any position readout, any go-to height, or any way to return to the same focus — or the same deliberate defocus — twice. Every part change means re-finding focus by eye, and a MOPA colour that worked last week depends on remembering exactly how far out of focus the head was.

2. The idea

Count the screw, don't replace it. A 600 PPR quadrature encoder is coupled to the lead-screw / top-knob shaft, on the screw side of the motor coupling, so moves made with the hand knob count too — not just motor moves.
Pico reads it in hardware. A Raspberry Pi Pico (RP2040) decodes quadrature with its PIO blocks: 600 PPR × 4 edges = 2,400 counts per revolution. No counts should be lost to a busy main loop.
Original buttons keep working. The Pico parallels the existing +/− buttons through isolated relay contacts. Pressing the machine's own buttons still works exactly as before; the Pico just gets to “press” them too.
OLED height readout (for example Z 142.35 mm), with the current lens / preset shown alongside.
Position stored in flash, so the height survives power-off — and homing to a sensor on the column to re-establish an absolute reference.
Backlash compensation: final approach to a target is always made from the same direction, so nut backlash lands on the same side every time.
Stall detection: if a move is commanded and the encoder does not change, stop — something is jammed, at a limit, or the coupling is slipping.

3. The workflow Hank wants

Caliper a part → enter its height → the head goes to focus.

Caliper the part and measure its thickness.
Enter that height on the Pico.
The Pico moves the head to focus height − part thickness, approaching from the compensated direction.

A preset bank per lens (110 and 200) stores Focus and small ± defocus offsets. Those offsets are exactly what MOPA colour marking plays with, so a colour recipe becomes “this lens, this preset” instead of a remembered knob position — with the actual offset values to be proven by colour tests, not assumed.

4. Honest accuracy note

Counting screw rotations gives very fine resolution — it does not give guaranteed accuracy. The encoder cannot see backlash, nut wear, or a slipping coupling. So, before any number on this page is trusted:

Measure the lead empirically: 10 full knob turns versus the actual height change on a caliper / dial indicator. The lead (mm/rev) is unknown and unmeasured until that is done — it is not stated anywhere on this page.
Measure backlash the same way, and set the compensation from the measurement.
Verify against a dial indicator over real moves, including after direction changes.
Zero at true focus per lens. Focal length is per-machine and per-lens; the stored zero has to be found on this machine, for each lens.

5. Status / next steps

Concept only — no parts ordered, nothing fitted. Waiting on Hank's measurements and photos:

Shaft diameter, and a photo with the crank / hand knob removed (decides the encoder coupler).
The 10-turn lead measurement — or threads per 50 mm plus a direct travel test.
Photos of the +/− button panel (decides how the relay contacts parallel it).
Lens list to build presets for: 110 / 200.

Hank uses LightBurn; the Pico Z stays standalone for now — no LightBurn integration is planned or promised.

6. Rough BOM

Part	Notes	Estimate
Raspberry Pi Pico (RP2040)	PIO quadrature decoding, flash position store	est.
600 PPR quadrature encoder	2,400 counts/rev after ×4 decoding	est.
Shaft coupler	Size waits on the shaft-diameter measurement	est.
OLED display	Height readout, e.g. `Z 142.35 mm`	est.
2-channel relay module	Isolated contacts paralleling the +/− buttons	est.
Home sensor	Column-mounted homing reference	est.
Jog control	Entering part height / picking presets	est.
Separate 5 V supply	Never powered from the laser PSU	est.
Total (rough, US)		~$65–100

Estimate only, for orientation — not a quoted price list. The related earlier workup that adds a separate stepper, Pi 5 and overhead camera is at /projectaddons/; this project is the smaller “count the screw that's already there” version.

← Back to Laser View home · Mockup image (PNG) · Project add-ons · Projects