What Is the Machine Home CNC?

• 1
  The home position, and the workpiece zero measured from it
• 2
  How homing works on the machine and what it can and cannot fix
• 3
  Where drift comes from: thermal growth, chips, and loose setups
• 4
  What changes when rotary axes join the home chain
• 5
  What machine home CNC can and cannot deliver
• 6
  Choosing a setup by part geometry and tolerance

Machine home is the reference position a CNC returns to before a job starts. It sits at the machine's own hard limits, and it exists for the machine, not for your part. Position is not geometry. The controller knows where the saddle is relative to its own frame, nothing more.

The workpiece zero is what you actually cut against. It is a stored offset measured from machine home to a corner or bore of the fixture. So the question of what is the machine home CNC really answers itself: home is the anchor, the offset is the part. Change the anchor and every dimension downstream moves with it.

On a 3-axis mill, home is usually the extreme positive corner of X, Y and Z travel. The operator jogs to that corner, the switches trip, the controller re-references, and the offset table is then filled by touching off the stock or the vise jaw. Nothing about the part is known to the machine until the operator measures it.

That matters when you read a drawing. A ±0.005 mm callout is not a statement about the machine's home accuracy. It is a statement about the whole chain: home repeatability, switch hysteresis, thermal growth, tool setting, and the rigidity of the setup. Home only sets the origin. The rest decides whether the origin holds.

• 1
  Home is a machine frame datumIt belongs to the machine builder, not to your drawing.
• 2
  Work offsets sit on topG54–G59 store measured distances from home to the part.
• 3
  Tool length is a third layerEach tool carries its own Z offset, measured or probed.
• 4
  Slug reads left to rightHome, then offset, then tool, then cut.

A typical homing cycle runs at reduced feed until a switch trips, backs off, then approaches again at a slower rate to find the switch's electrical edge. That second pass is what turns a crude proximity signal into a usable position. The repeatability of that two-pass approach is the number that matters, not the switch's nominal accuracy.

Mechanical limit switches with roller plungers are common and inexpensive. They re-trip to roughly ±0.01 mm in good conditions, which is enough for most 3-axis work. Optical or inductive proximity sensors remove the moving parts and hold tighter, often into the low single-digit microns. Absolute encoders skip the cycle entirely because the controller already knows the position after power-up.

None of these devices measure the part. They measure the machine's own frame. A homing switch that repeats perfectly still leaves you with a vise that moved 0.03 mm when the operator clamped it, or a casting that was 0.2 mm off nominal in the raw stock. Homing cannot correct either. It only guarantees that the machine starts from the same place each cycle.

This is the boundary of the concept. Machine home CNC gives you a repeatable starting point. It does not give you a reference to your part's features unless someone probes or indicates those features first. That extra step is where shops differ, and it is where the tolerance chain is either closed or left open.

• 1
  Two-pass approachFast trip, back off, slow re-trip. The slow pass sets repeatability.
• 2
  Switch type sets the floorMechanical around ±0.01 mm; inductive and optical tighter.
• 3
  Absolute encoders skip homingPosition survives power-down, so no cycle is needed.
• 4
  No switch measures the partPart reference comes from probing or indicating.

A machine that homes perfectly at 8 a.m. can be 0.02 mm out of true by 2 p.m. The ballscrew grows as it warms, the spindle cartridge grows, and the column leans a little as the bed temperature changes. This is why shops that hold tight tolerances run warm-up cycles before the first cut and re-reference between long jobs.

Chips and coolant are the second source. A chip packed under a vise jaw or on a locating face shifts the part by its own thickness. A vise that is wiped but not stoned will still hold a burr from the previous part. These are setup errors, not machine errors, and they are the most common cause of a dimension that was fine yesterday and is out today.

The third source is the offset stack itself. Every tool has a length offset, every fixture has a work offset, and every probe has its own calibration. Each layer adds a small uncertainty. A shop that measures tool length with a presetter and verifies the first part on the machine closes that stack quickly. A shop that trusts the last job's numbers does not.

You cannot inspect these away. They have to be controlled at the machine. That is why we run raw material checks before cutting, in-process monitoring during the run, and a full inspection before shipment, with reports available on request.

• 1
  Thermal growthScrews and spindles expand; warm up before tight work.
• 2
  Chip and burr seatingA chip under a jaw shifts the part by its thickness.
• 3
  Offset stack errorTool, work and probe offsets each add uncertainty.
• 4
  First-part checkVerifying the first part catches all three at once.

On a 4-axis mill, the rotary table adds an A or B axis to the same home logic. The table has its own home switch or encoder zero, and the rotation center has to be established relative to X, Y and Z. That center is often called the pivot point. If it is off, every feature cut at an angle is off by roughly the error times the distance from the pivot.

On a 5-axis machine, two rotary axes stack on top of each other, and the pivot point becomes a point in space rather than a line. The controller needs that point in machine coordinates, and it needs the tool length to go with it. Change a tool without updating the length and the pivot is wrong for that tool.

This is why 5-axis work is often probed rather than assumed. A spindle probe can touch a known artifact or a feature on the fixture, and the controller solves for the pivot. It takes a few minutes. It is also the difference between a bracket that fits and a bracket that needs a shim.

The trade-off is real. Five-axis simultaneous work buys you access to angles that a 3-axis machine cannot reach, and it costs you setup time and a longer offset chain. For a part with features on four faces, the rotary setup usually wins. For a flat plate with holes on one face, it does not.

• 1
  Rotary zero is its own datumThe table needs a reference, separate from the linear axes.
• 2
  Pivot point drives accuracyError scales with distance from the pivot.
• 3
  Tool length changes the pivotA new tool shifts the effective center.
• 4
  Probing closes the loopTouch a known feature and let the controller solve.

A tight machine home gets you a repeatable origin. It does not, by itself, get you a tight part. The parts we hold to ±0.005 mm do so because the home is stable, the offsets are measured rather than guessed, the tooling is rigid, and the finishing passes are light. Remove any one of those and the tolerance opens up.

Surface finish follows the same logic. A fine finish of Ra 0.2–0.8 μm needs a sharp tool, a stable setup, and a light finishing pass. A high finish of Ra 0.8–1.6 μm is the normal band for a well-run finishing pass. As-machined at Ra 1.6–3.2 μm is what you get when the priority is cycle time rather than finish.

Material changes the picture. Aluminum 6061 and 7075 cut freely and hold tight dimensions. Stainless 316 and 17-4PH work-harden and push back on the tool, so the setup has to be stiffer. Titanium TC4 and Inconel move the difficulty again: heat stays in the cut, tool life drops, and the finishing strategy has to be planned before the first pass.

The honest answer is that home is a precondition. It is necessary and it is not sufficient. When a shop tells you it holds ±0.005 mm, the question to ask is which machine, which setup, and how the first part was verified. The answer tells you more than the number.

• 1
  Home is necessary, not sufficientIt sets the origin; the setup decides the result.
• 2
  Finish is a setup decisionLight finishing passes and sharp tools set Ra.
• 3
  Material sets the difficultyAluminum, stainless and titanium need different rigidity.
• 4
  Ask how the first part was checkedThat answer predicts the run.