Neje CNC Laser: 7 Proven Secrets for Flawless Engraving Every Time
Diode engraving goes wrong for the same handful of reasons: focus drifts, the surface is wrong, power and speed fight each other. This page walks through seven settings and habits that make a Neje CNC laser repeat the same result on the tenth part as on the first. It is written for engineers and buyers who need marks that survive inspection, not just a photo of one good coupon.

In this article
- 1
- 2
- 3
- 4
- 5
- 6
- 7
- 8
- 9
Why a Neje CNC laser drifts off spec
A diode module is not a sealed industrial head. The lens sits a few millimetres from the work, the beam is a rectangle rather than a round spot, and the whole carriage rides on belts. Every one of those facts shows up in the finished mark. Two runs that look identical on the screen can differ by 0.3 mm in focus height after a bump or a chip of debris.
The second reason is thermal. A diode stack loses output as it heats, so a job that starts at full power finishes 10–15% weaker unless you compensate. On stainless and anodized aluminium this shows as a light first pass and a dark last pass on the same plate. On wood it shows as inconsistent char depth across a 300 mm part.
Neither problem is solved by buying a bigger module. They are solved by controlling three variables: distance from lens to surface, energy per unit area, and how fast heat leaves the cut zone. The seven sections below treat each one in turn. Treat them as a process sheet, not a list of tips.
Secret 1: set focus with a ramp test, not a gauge block
The acrylic ramp test is the only focus method that accounts for your specific lens and your specific material. Cut a strip of 3 mm cast acrylic at roughly 10°, lay it under the head, and run a single line at low power. The narrowest point of the burned line marks the true focal distance for that lens and that standoff.
Do this once per lens, then write the number on the machine. A typical 5.5 W module with a fixed-focus lens lands between 20 mm and 45 mm from the nozzle tip depending on the lens barrel. Recheck after any crash, belt change or lens swap.
For engraved plates, focus on the top surface, not the bed. A 2 mm sheet of 304 stainless changes the number. The same applies to curved parts: on a cylindrical surface, expect the mark to widen at the edges, and either rotate the part or accept the falloff deliberately.
A short note on spot shape. Diode output is a rectangle, roughly 0.08 × 0.1 mm at focus on a 5.5 W head. That means line spacing must be tighter than the long axis, or you get visible ridges in filled areas. 0.05–0.07 mm line interval is a practical starting range for solid fills.
Secret 2: engineer the surface before the beam arrives
Laser marking reacts with the surface you give it. Oil, oxide, a brushed grain, a protective film: each changes absorption and each changes the colour you get. Wipe stainless and aluminium with isopropyl alcohol at 99% or higher and let it flash off. On brass and copper, oxide grows back within hours, so mark the same day you degrease.
On anodized aluminium, the mark is a bleaching of the dye layer, not a cut. Thicker anodic coatings need more energy and produce a softer grey. Hardcoat anodizing at 25 μm and above may barely mark at all with a diode module, and that is a boundary worth knowing before you promise a customer a black logo on a hardcoat panel.
Bare mild steel will engrave, but the mark is shallow and rusts. A thin coat of layout fluid or a dedicated marking spray gives a darker, more even result on 1018 and 1045. On 304 and 316 stainless, a light pass in the 20–30% power range produces a colour change rather than a groove; higher power just makes a rough grey pit.
Plastics need their own decision. ABS and POM cut cleanly but release fumes that coat the lens. PMMA engraves to a frosted white and is the easiest plastic for a diode. PVC should not be lasered at all; the chlorine attacks the optics and the operator.
Secret 3: balance power and speed as energy per area
Most operators adjust power and speed by feel. A cleaner approach is to fix the energy delivered per unit area and change only one variable at a time. For a given material, there is a band where the mark is fully formed but the heat-affected zone stays narrow.
A practical starting point on 304 stainless with a 5.5 W optical output is 60–70% power at 600–800 mm/min for a colour mark, and 80–90% at 300–400 mm/min when you actually want depth. Move outside that band and you either get a faint mark or a wide, discoloured halo.
The same logic applies to wood and leather. On 3 mm birch ply, 40% power at 1,200 mm/min gives a clean dark line; 80% at 1,200 mm/min burns through the top veneer. Lower the speed rather than raising the power when you want more contrast, because power drives the halo and speed drives the edge quality.
Number of passes is a third lever, and it is usually the safest one. Three passes at 50% power produce a cleaner mark than one pass at 100%, with less char and less warp on thin material. It costs cycle time, so decide which you value.
Secret 4: use air assist to shape the mark, not just to cool
Air assist does three things: it clears the plume so the beam reaches the surface, it cools the mark so the edge stays crisp, and it pushes debris away from the cut. On engraving, the first two matter most.
Run 10–20 L/min of clean, dry air for engraving. Higher flow is not better. Above roughly 30 L/min on thin material you start to see uneven marking because the air stream cools one side of the pass more than the other.
Position the nozzle so the jet trails the beam direction, blowing into the already-engraved area. That pushes smoke and particles away from the unmarked surface instead of across it. On a Neje CNC laser with an adjustable nozzle, 2–3 mm of standoff and a 15° trailing angle works well for flat plates.
For anodized aluminium, reduce flow to the low end, around 8–12 L/min. Too much air bleaches the mark unevenly. For wood and ply, run higher and accept the extra noise; the plume is the main cause of grey, muddy engraving.
Secret 5: treat the lens like a spindle insert
A dirty lens is the diode equivalent of a chipped insert. It does not fail loudly; it just makes every downstream result worse. A thin film of condensed material can cut delivered power by roughly 30% and create hot spots in the beam profile.
Inspect the protective window and the collimating lens weekly under oblique light. Look for a haze ring rather than a visible speck. Clean with lens-grade swabs and 99% isopropyl alcohol only. Never use tissue, shop towels or compressed air from an open line; both scratch coatings and both carry oil.
Replace the protective window on a schedule, not on failure. If you mark coated steel or plastics daily, a 4–6 week interval is reasonable. The window costs far less than a scrapped batch.
Keep a log of lens changes with the date and the job count. When marks suddenly go light across every material, the lens is the first suspect, and the log tells you in ten seconds whether it is due.
Secret 6: software settings that actually change the mark
Diode firmware is 8-bit, so power resolves into 256 steps. That matters when you are trying to hold a grey level. Below about 5% power, many modules do not fire reliably at all, which is why very light fills come out patchy.
Set the line interval from the spot size rather than from a preset. For a 0.08 × 0.1 mm spot, 0.05 mm interval gives about 50% overlap and a solid fill. Wider than 0.1 mm and you will see stripes on any angled surface.
Overscan matters more than most people expect. Add 2–5% overscan so the head is at full speed before it reaches the first pixel. Without it, the left and right edges of a filled rectangle are darker than the centre.
Finally, match the transmission mode to the job. Grayscale mode gives smoother photographic results; threshold mode gives sharper text and logos. Do not mix them on one panel if the two areas need to look consistent under the same light.
Secret 7: keep a parameter library and test on coupons
A parameter library is just a spreadsheet with material, thickness, power, speed, line interval, passes, focus distance, air flow and the date. That is enough to reproduce a job six months later when the operator has changed and the material batch has not.
Test on coupons cut from the same sheet, not on scrap of a similar grade. Stainless from two suppliers can mark differently at the same settings because of surface finish and alloy chemistry. Keep one coupon per batch with the date written on it.
Requalify when anything changes: a new lens, a new module, a new material supplier. Run a three-point test at low, medium and high energy and record the result visually. Ten minutes of testing prevents a full plate of rejects.
This is the same discipline we apply to machined parts. Every job at GreatLight runs against a documented setup, and the same logic keeps a diode laser honest. Engraving a serial number on a machined housing only works if the mark is repeatable, because the housing is already at final tolerance when it reaches the laser bench.
Starting parameters by material and goal
Values are starting points for a 5.5 W diode module. Adjust one variable at a time.
| Material | Goal | Power | Speed |
|---|---|---|---|
| 304 stainless | Colour mark | 60–70% | 600–800 mm/min |
| 304 stainless | Depth mark | 80–90% | 300–400 mm/min |
| Anodized aluminium | Light grey logo | 20–30% | 1,000–1,400 mm/min |
| Mild steel 1018 | Dark mark with spray | 50–60% | 700–900 mm/min |
| 3 mm birch ply | Clean dark line | 40% | 1,200 mm/min |
| PMMA cast | Frosted white | 35–45% | 900–1,100 mm/min |
| ABS | Shallow mark, fume control | 30–40% | 800–1,000 mm/min |
| Hardcoat anodize 25 μm | Not recommended | — | — |
What matters most
If you mark one material all day, spend your time on focus and optics; if you mark many materials, build the parameter library first and buy the air assist second.
Questions engineers ask next
Can a diode laser mark 304 stainless permanently?
Yes, but the mark is a surface oxide layer, not a deep cut. At 60–70% power and 600–800 mm/min on a 5.5 W module you get a dark grey to blue-grey mark that survives handling and light abrasion.
It will not survive aggressive bead blasting or heavy polishing. If the part needs a mark that outlasts a finishing operation, plan the laser step after finishing, or switch to a fiber laser.
Why is my engraving darker on one side of the plate?
Three causes are common: the bed is not level relative to the gantry, the module is overheating and losing output as the job runs, or the air assist is cooling one side more than the other.
Check bed flatness first with a dial indicator across the work area. Then run a 5-minute continuous job and touch the module heatsink. If it is too hot to hold, add a break or reduce power and increase passes.
Do I need a different lens for engraving and cutting?
Usually yes. A longer focal length gives a wider spot and better cutting depth on thick material. A shorter focal length concentrates energy for fine engraving detail.
Keep one lens per task and record the focus distance for each. Swapping lenses without re-running the ramp test is the single most common reason marks shift between jobs.
How do I stop burns on thin plywood?
Lower the power and add passes instead. Two passes at 35–40% produce a cleaner line than one pass at 70–80%, and the top veneer stays intact.
Raise air flow to the higher end of the range and move the workpiece up off the bed on standoffs so heat does not reflect back from the table.
What records should I keep for a repeatable process?
Material grade and supplier, thickness, focus distance, power, speed, line interval, number of passes, air flow and the date of the last lens change. One row per job setup.
Keep a physical coupon per material batch with the date written on it. When a mark looks wrong, the coupon tells you whether the material changed or the machine did.
Need engraving on a finished machined part?
We machine to ±0.005 mm, apply the finish, then laser mark the final surface so the mark is not disturbed by later operations. Upload your files for a quote and DFM feedback within 12 hours.
12-hour quote100% inspection±0.005 mm