Achieving Machined Surface Finish on a CNC Router: Ra, Stepover, and Toolpath Strategy to Skip Hand-Sanding

Feeds-and-speeds content answers how fast you can cut without breaking a bit or stalling the spindle. Surface finish is a different question entirely: how smooth is the surface left behind once the cut is done, and how much of that smoothness comes from the toolpath itself versus how much you're going to have to sand out afterward. On a desktop router like the Wolfpawn 4040 Pro, a few toolpath decisions make the difference between a part that comes off the machine ready to finish and one that needs twenty minutes of hand-sanding to erase visible scallops.

What Ra Actually Means

Ra (roughness average) is the arithmetic mean of the peaks and valleys on a surface, measured in micrometers or microinches. It's an industrial spec, and desktop CNC routers running wood, plastic, and soft aluminum don't need to hit the same numbers a production aerospace shop does — but the concept still matters for understanding what's actually happening at the tool-material interface. On a router, the dominant source of surface roughness isn't tool wear or machine rigidity (though both contribute) — it's cusp height, the scalloped ridges left between adjacent toolpasses, and that's almost entirely under your control through stepover selection.

Stepover and Cusp Height

Stepover is the distance the tool moves sideways between passes, usually expressed as a percentage of tool diameter. Cusp height — the ridge left behind between passes — is a direct, calculable function of stepover and tool geometry:

Tool TypeCusp Height Formula (approx.)Notes Flat end mill (finishing a flat face)Cusp height ≈ stepover² / (8 × tool radius)Doubling stepover roughly quadruples cusp height — this relationship is not linear Ball nose (3D relief/contour finishing)Cusp height ≈ stepover² / (8 × ball radius)Same relationship; ball nose is chosen specifically because it holds this formula cleanly on curved surfaces

The practical takeaway: a small reduction in stepover buys a disproportionately large improvement in surface finish, which is why finishing passes commonly use 5–10% stepover even though roughing passes might run at 40–50%. The cost is toolpath time — cutting a finishing pass at 10% stepover instead of 40% roughly quadruples the number of passes and the machining time for that operation, so this is always a deliberate trade of machine time for less sanding time.

Roughing vs. Finishing: Use Two Different Toolpaths

The single biggest surface finish mistake on hobby CNC is running one toolpath, at one stepover, for the entire job. A proper two-stage strategy:

1. Roughing pass: Larger stepover (30–50%), leaves a stock allowance of 0.010–0.030" on all finished surfaces, prioritizes removing material fast over surface quality. This is where you push feed rate and chip load hard.
2. Finishing pass: Small stepover (5–15%), removes only the stock allowance left by roughing, runs at a lighter chip load and often a higher spindle RPM specifically to improve finish rather than material removal rate. On flat surfaces, a finishing pass can also run as a single-pass "surfacing" or "facing" operation across the whole area rather than following the part contour.

Beyond Stepover: Other Levers That Affect Finish

• Climb vs. conventional milling. Climb milling (cutter rotation matches feed direction) generally leaves a cleaner finish on the "climb" side of a cut, with less fuzzing on wood and less work-hardening on soft metals — but it demands a rigid machine and tight backlash, since it can pull the workpiece into the cutter if there's slop in the gantry.
• Spindle RPM relative to feed rate. Too slow an RPM for the feed rate leaves visible tool marks (each flute engagement is too far apart); too fast for a soft or gummy material can cause melting or burnishing instead of clean cutting, especially on plastics like acrylic and HDPE. Surface finish issues are frequently an RPM problem disguised as a feed problem.
• Tool sharpness and flute count. A dull bit tears fibers instead of shearing them, especially visible on wood grain reversal areas; more flutes generally means a finer finish at the same feed rate (more cuts per revolution), at the cost of chip clearance — this is why finishing bits and roughing bits are often genuinely different tools, not just the same tool run differently.
• Deflection. A long, thin end mill deflects under cutting load, and that deflection shows up as a wavy, inconsistent surface rather than uniform scalloping. If a finishing pass looks worse than the math predicts, check tool stickout length before blaming the toolpath.
• Down-cut vs. up-cut vs. compression bits change where the visible finish quality lands on a cut — an up-cut bit pulls chips up and leaves a cleaner bottom face but can leave a fuzzy top edge on plywood, a down-cut does the opposite, and a compression bit (up-cut lower, down-cut upper) is specifically designed to leave clean edges on both faces at the cost of being a more specialized, harder-to-source tool.

Setting This Up in Your CAM Software

Most CAM packages used with a Wolfpawn-class router — Vectric Aspire, Fusion 360, Carbide Create — separate roughing and finishing as distinct toolpath operations by default, but it's easy to accept the default stepover and move on. Deliberately set roughing stepover in the 30–50% range and drop finishing stepover to single digits on any surface where visible scalloping matters — a sign face, an edge-lit panel, anything meant to be seen rather than just structural. For 3D relief carving specifically, this is also where ball-nose finishing toolpaths (as opposed to flat-end roughing) become worth the extra machining time, since the ball geometry is what makes the cusp-height formula behave predictably on curved surfaces in the first place.

When to Stop Chasing Toolpath and Just Sand

Not every project justifies a 10% stepover finishing pass that triples your machine time. For parts that get painted, primed, or otherwise finished by hand anyway (most V-carved signs, most structural parts, anything getting a coat of polyurethane), a moderate finishing stepover (15–25%) followed by a quick hand-sand is often the faster overall workflow than chasing a toolpath-perfect surface. Reserve the tight stepovers for surfaces that are genuinely difficult to sand by hand — tight interior curves, deep 3D relief detail, or edge-lit acrylic where visible tool marks scatter light unevenly.