Branch of Digital Display Device: How It Shapes Machine Tool Feedback

• 1
  What the branch of digital display device actually covers
• 2
  How a linear scale counts position
• 3
  Scale feedback compared with ballscrew feedback
• 4
  Why this matters on the parts you buy
• 5
  Where scale feedback stops helping
• 6
  Feedback choices by part and process

The branch of digital display device sits under the China Machine Tool & Tool Industry Association. Its members build two things that always travel together: the measuring head that reads position, and the display unit that shows it. Around 90% of the market in this category runs through that same network of sensors, so the branch writes standards for both halves rather than treating them as separate products.

A digital display device is a displacement measuring system. The sensor half counts how far the table, ram, or spindle has moved. The display half turns those counts into numbers an operator can read at the machine. Neither half is useful alone. A scale with no readout gives you nothing, and a readout with a drifting sensor gives you confident wrong numbers.

The branch works on three practical jobs. It drafts and revises technical standards for scales and readouts. It collects quality feedback from users and feeds that back to makers. It also gathers economic statistics on production and export so members can see where the market is going.

That mix matters to a machine shop more than it first appears. When a standard pins down resolution, repeatability, and signal output, a buyer can compare two scales on paper instead of on a test bench. When it does not, the numbers on a datasheet mean whatever the seller wants them to mean.

• 1
  SensorsLinear travel sensors and angular displacement sensors
• 2
  ReadoutsDigital display instruments with counting and compensation functions
• 3
  StandardsResolution, repeatability, signal format, environmental limits

Most machine tool scales are optical or magnetic. A glass or steel grating carries fine periodic marks, and a read head scans them as the axis moves. The head outputs two square waves in quadrature, offset by 90 electrical degrees. Direction comes from which wave leads, and distance comes from counting edges.

Resolution is set by grating pitch and by electronic interpolation. A 20 μm pitch with 20-fold interpolation gives 1 μm counts. Finer counts do not automatically mean better parts. What matters is repeatability, the spread of readings when you return to the same position, and accuracy, the difference between the reading and true position over the full travel.

Quadrature counting also tells the control which way the axis moved without a separate direction line. A common failure mode is a dirty or scratched grating. The head still sees alternating light and dark, so the count keeps running, but pulses drop out at one spot. The symptom is a repeating error at the same coordinate, not a random one.

Magnetic scales use a magnetized strip instead of glass. They tolerate coolant, chips, and vibration better, and they cost less per meter. Their resolution and long-term stability are lower than a good glass scale, so they suit general milling and turning rather than a jig grinder.

• 1
  Quadrature signalTwo channels 90° apart; leading channel sets direction
• 2
  Grating pitchCoarse pitch plus interpolation sets the count size
• 3
  Repeatability vs accuracyReturn spread is not the same as true-position error

A control can close the position loop in two places. Semi-closed feedback reads the motor or ballscrew, and the scale only appears on a separate digital readout for the operator. Closed-loop feedback reads a scale mounted on the slide itself, so the control sees the actual table position.

The difference shows up as heat. A 1,000 mm steel ballscrew grows about 12 μm per 1 °C rise. Run a machine hard for two hours and the screw can be 3–5 °C above ambient, which pushes the table 36–60 μm past where the control thinks it is. A slide-mounted scale does not see that error at all, because it measures the table, not the screw.

Backlash and screw wear follow the same logic. Semi-closed feedback cannot correct either, because the loop closes before the mechanical loss. Closed-loop feedback corrects both, but it costs more and it needs a scale that survives coolant and chips over years of service.

This is where the association branch earns its keep. It sets language that lets a shop compare a glass scale rated for a grinding machine against a magnetic strip rated for a router, without confusing the two spec sheets.

• 1
  Semi-closedEncoder on motor or screw; scale is display only
• 2
  Closed-loopScale on the slide; control corrects screw growth and backlash
• 3
  Cost tradeClosed loop adds hardware and sealing requirements

When we machine a part, the scale is what tells the operator the tool is where the program says it is. On a 5-axis job with a 4,000 mm envelope, a 40 μm positioning error is the difference between a bolt hole that clears and one that binds. We hold ±0.005 mm on features that need it, and that only holds if the feedback chain is honest.

You can often read the feedback type off a supplier's inspection report. A report that lists CMM results for critical dimensions, not just the machine's own display, has checked the part rather than the machine's opinion of the part. We run raw material checks, in-process monitoring, and a final inspection on every order, with reports on request.

There is a second-order effect on tool wear. If the scale drifts, the operator compensates by nudging the offset, and the next part runs with a wrong baseline. Errors like that repeat across a batch. They are hard to catch later because the parts are internally consistent and only fail at assembly.

For prototype work the risk is lower, because you inspect every piece. For a 10,000-part run it is the main risk. That is why we treat the feedback chain as a process variable, not as a machine specification.

• 1
  Check the reportLook for CMM numbers, not display printouts
• 2
  Batch riskOffset drift repeats across parts and hides until assembly
• 3
  Prototype vs runOne-off inspection masks drift that a long run exposes

Closed-loop feedback corrects the tool position. It does not correct the part. Thermal growth in the workpiece, fixture lift, and tool deflection all move the cutting edge after the control has done its job. A 20 mm end mill pushed at 0.1 mm per tooth can flex 30–50 μm in a deep pocket, and no scale on the slide will see it.

Vibration is another boundary. A scale reads average position over its sampling window. If the axis is chattering at 200 Hz, the reading is stable while the cut is not. Surface finish will show it: Ra 1.6–3.2 μm as machined is normal, but a jump toward Ra 0.8 μm or worse in one zone points at the setup, not the scale.

Environment sets a hard limit too. Glass scales dislike fine dust and direct coolant spray. Magnetic scales dislike strong external fields and nearby welding. Neither likes a slide that has lost its way lubrication, because stick-slip motion produces a position history that no feedback loop can undo.

Knowing these limits tells you when to spend on feedback and when to spend on the fixture instead. For thin-wall parts, the fixture usually wins.

• 1
  Tool deflectionCutting force bends the tool; the scale never sees it
• 2
  ChatterAveraged readings stay clean while the cut is not
• 3
  EnvironmentCoolant, dust, and stray fields cut scale life