Overrides are the operator’s direct control over a running program. They are essential for proving out a new job — and one of the most common reasons a machine runs slower than its programmed cycle time without anyone noticing.
The three overrides
Feed override
- Scales
- Programmed feed rate on cutting moves
- Typical range
- 0–120%, some controls up to 200%
Rapid override
- Scales
- Speed of non-cutting positioning moves
- Typical range
- Fixed steps, e.g. 25 / 50 / 100%
Spindle override
- Scales
- Programmed spindle speed
- Typical range
- About 50–120%
| Override | Scales | Typical range |
|---|---|---|
| Feed override | Programmed feed rate on cutting moves | 0–120%, some controls up to 200% |
| Rapid override | Speed of non-cutting positioning moves | Fixed steps, e.g. 25 / 50 / 100% |
| Spindle override | Programmed spindle speed | About 50–120% |
At 0% feed override the axes stop moving; ranges and steps differ between control makers.
What a reduced override costs
Illustrative example: a part with 60 seconds of cutting moves, run at 80% feed override, takes 75 seconds of cutting instead — 15 seconds longer per part. Over a shift of 300 parts that is 75 minutes, and nothing in the program shows it. It surfaces only as an actual cycle time that is longer than the quote.
Seeing overrides across the floor
ThingConnect reads the feed, rapid and spindle override values from the control on each poll, and the andon board shows feed and spindle override on each machine’s tile. A machine sitting at 70% for a whole shift stands out next to its neighbours at 100%.
A low override is not automatically wrong. Often it is the operator compensating for a real problem — a worn tool, a weak fixture, chatter on a thin wall. The useful question is why it was lowered: if the same job always runs at 80%, the program’s feeds are probably too aggressive for the setup and should be fixed at the source, so the quote and the floor agree on the cycle time.