The Process of Developing Digital Display Devices for Machine Tools

• 1
  Key takeaways
• 2
  Define the axes and resolution before you buy anything
• 3
  Choose the scale technology that fits the machine
• 4
  Mount the scale and reader head so alignment holds
• 5
  Wire the display, set parameters, and verify counts
• 6
  Step by step: fitting a 2-axis DRO to a knee mill
• 7
  Which feedback type fits which machine

Start on paper. List every axis you intend to measure, the travel range of each, and the tightest tolerance the part actually carries. A two-axis knee mill usually needs X and Y only. Add Z when the quill depth controls a bore depth, and add a fourth axis when the rotary table is used for indexing rather than setup.

Resolution is a decision, not a default. A 0.005 mm scale on a machine with 0.05 mm of slide backlash gives the operator false confidence. Match the display resolution to roughly one tenth of the tolerance you can actually hold on that machine, and no finer.

Then measure the travel. Scales are sold by travel length, not by machine size, and a scale that is 50 mm short of the table travel will be hit by the way cover at the end of the stroke. Add the reader head width plus 30 mm of clearance at each end.

For a reference point, decide whether you want an absolute scale or an incremental one with reference marks. Absolute scales keep position after a power cut. Incremental scales cost less but need a datum routine at the start of each shift.

Write all of this down before contacting a supplier. A clear axis list with travels and resolution shortens the quotation cycle and prevents the most common retrofit mistake: ordering the wrong scale length.

• 1
  Axis listX, Y, Z, and any rotary axis, with the function each one serves.
• 2
  Tolerance per axisThe tightest feature the machine must produce on that axis.
• 3
  Travel plus clearanceTable travel + reader head width + 30 mm each end.
• 4
  Reference typeAbsolute for unattended work, incremental with marks for general use.

Linear scales come in two practical families for machine tools: glass grating scales and magnetic tape scales. Glass scales read finer and hold accuracy over long travels, but they dislike chips and coolant. Magnetic scales tolerate swarf and vibration, and they are the usual choice on lathes and on any machine cutting dry.

Encoders on the lead screw are a third option. They are cheap and easy to fit, but they measure screw rotation, not table position. Any backlash or screw wear shows up directly in the reading. Use them on light machines where the operator already works with backlash, and avoid them where the part tolerance is tight.

Resolution and accuracy grade are separate numbers. A 1 μm resolution scale with a ±10 μm accuracy grade still has ±10 μm of error over its length. Read the accuracy grade, not just the resolution, when the workpiece tolerance is below 0.05 mm.

For rotary axes, a ring encoder or a rotary table encoder gives angular resolution down to arc-seconds. If the machine only indexes to whole degrees, a coarser unit is fine and cheaper.

On our own 5-axis and mill-turn work, the position feedback choice is made at the same time as the fixture design, because both determine how the first article gets verified. Machine tools that carry a display retrofit follow the same logic.

The scale body mounts to a fixed surface, usually the column or the saddle, and the reader head mounts to the moving member. Keep the scale on the side that does not collect chips. On a mill, that is normally the column side facing away from the cut.

Alignment tolerances matter more than most installers expect. Hold parallelism between the scale and the axis of travel within 0.1 mm over the full length, and keep the gap between the reader head and the scale within the manufacturer's window, typically 0.1 to 0.5 mm. Twist is the killer: a head that is square at one end and canted at the other will drop counts in the middle of the stroke.

Use a dial indicator on the moving member to sweep the scale face along its length. Shim at the mounting points rather than bending the scale body. A bent scale reads well near the ends and badly in the centre.

Cable routing needs slack for the full travel plus a service loop. A cable that goes tight at the end of the stroke will pull the head out of alignment or break the conductor inside the jacket.

Cover the scale with the supplied shield or a way cover. Coolant mist and fine cast iron dust are the two most common causes of a scale that reads correctly on installation and drifts three weeks later.

• 1
  ParallelismWithin 0.1 mm over the full scale length.
• 2
  Reader head gapKeep inside the maker's window, often 0.1–0.5 mm.
• 3
  Cable slackService loop at both ends of travel.
• 4
  CoverShield or way cover fitted before the first cut.

Wire the display to the scales with shielded cable, and ground the shield at the display end only. Grounding both ends creates a loop that picks up drive noise, and the symptom is a count that jumps when the spindle starts. Keep the scale cable away from spindle and servo cables by at least 100 mm where the route allows.

Set the parameters before the first test: axis resolution, counting direction, and reference mark behaviour. Counting direction is wrong about half the time on a first install, and the fix is a parameter change, not a rewire.

Verify with a gauge. Move the axis a known distance with a gauge block or a granite square and dial indicator, and compare the display reading with the mechanical measurement. Do this at several positions along the travel, not just at the centre.

Check repeatability by returning to the same point ten times and recording the spread. A display that repeats within 0.005 mm is healthy for a manual machine. A spread of 0.02 mm points to a loose reader head or a scale that is not parallel.

Finally, check the reference mark routine. Power the display down, bring the axis back to the mark, and confirm the position returns to the same value. If it does not, the mark parameter or the mark position is wrong.

• 1
  Shield groundingDisplay end only, to avoid ground loops.
• 2
  Direction checkConfirm each axis counts the correct way.
• 3
  Position checkGauge block at three points along travel.
• 4
  Repeat checkTen returns to one point, record the spread.