A proximity switch that triggered at different positions each cycle. The customer needed repeatability of ±0.1 mm. The switch was an inductive NPN, 8 mm range. It triggered at 7.8 mm one cycle, 8.2 mm the next. The customer thought the switch was defective. It wasn’t — the switch had 0.1 mm repeatability but the target was moving past it at high speed. The issue was sensor selection, not the switch. This is about proximity switch accuracy vs repeatability.
Repeatability vs accuracy
These are different specs:
- Repeatability: how consistently the switch triggers at the same position. A good inductive switch has 0.01 mm repeatability (for a shielded 8 mm sensor).
- Accuracy: how close the trigger point is to the rated distance. A switch rated for 8 mm might trigger at 7.5 mm. That’s 0.5 mm error — but repeatable.
- Hysteresis: the difference between trigger-on and trigger-off distance. Typically 10-20% of the rated distance. For an 8 mm switch, hysteresis is 0.8-1.6 mm.
For positioning, you care about repeatability (does it stop at the same spot every time?), not accuracy (what’s the absolute position). A switch with 0.01 mm repeatability is fine for a home position — it triggers at the same relative position every time.
Why the switch triggered at different positions
The target was moving past the sensor at 1 m/s. The sensor has a response time of 1 ms. At 1 m/s, the target moves 1 mm in 1 ms. The switch triggers at a slightly different point each cycle because of the response time jitter (about 0.5 ms). At 1 m/s, that’s 0.5 mm jitter. The 0.1 mm spec was for a stationary target.
What I changed
1. Slowed the approach speed. I added a slow-down zone before the home position. The axis runs at 0.05 m/s when approaching the sensor. At 0.05 m/s, 0.5 ms jitter = 0.025 mm. That’s within spec. The home position is repeatable to ±0.05 mm.
2. Used a higher-speed sensor. I swapped to a sensor with 0.1 ms response time (high-speed inductive). At 1 m/s, the jitter is 0.1 mm. Still not 0.01 mm. For absolute repeatability at high speed, I use a magnetic encoder or a linear scale — not a proximity switch.
3. Used a light barrier. For repeatable positioning at high speed, a through-beam photoelectric sensor has 0.01 mm repeatability regardless of speed (the beam is instant). I replaced the inductive switch with a through-beam sensor. The home position is now ±0.02 mm repeatable at any speed.
The target material and size
Inductive switch range depends on the target material. A standard switch rated for 8 mm on steel only gives 4 mm on stainless steel, 2 mm on aluminum. The target must be at least the sensor face size (8 mm × 8 mm). A smaller target reduces the range. I always check the target material and size against the sensor datasheet.
| Sensor type | Repeatability | Speed limitation | Use for |
|---|---|---|---|
| Inductive (standard) | 0.01 mm (static) | ~1 kHz response | Home position, low speed |
| Inductive (high-speed) | 0.01 mm (static) | ~5 kHz response | Medium speed positioning |
| Through-beam photoelectric | 0.001 mm | Instant | High-speed repeatable stop |
| Magnetic encoder | 0.001 mm (with interpolation) | Continuous | Precision positioning |
The sensor I choose: through-beam for high-speed repeatable stops, inductive for low-speed home positions. The jittery switch wasn’t defective — it was triggering at 1 m/s with 1 ms response time. Slow the approach or use a light barrier.
