Home / All / Belt Selection / How much backlash does a timing belt drive have, and how do I reduce it?
How much backlash does a timing belt drive have, and how do I reduce it?
Direct answer: As a working figure, a correctly tensioned timing belt drive shows about 0.1 mm to 0.5 mm of lost motion at the driven pulley, measured at the pitch radius. It is not a catalogue number: pitch, span length, tension, load and pulley accuracy all move it. If the process needs tighter positioning, treat backlash as a design item and reduce it at the source rather than by over-tensioning.

Where the lost motion comes from
| Source | What it does | First thing to change |
|---|---|---|
| Designed tooth clearance | Every belt and pulley pair has clearance in the groove, so reversing the load crosses that gap | Accept it as designed, or move to a finer pitch with a matching profile |
| Cord elongation over the span | The free span stretches under load, so the driven pulley lags behind the driver | Shorten the centre distance, add support, or use a lower stretch cord |
| Tooth deflection | Rubber or PU teeth compress under torque, more so at high load and coarse pitch | Widen the belt or increase the number of teeth in mesh |
| Pulley runout and mounting | An eccentric or loose pulley adds play that looks exactly like belt backlash | Measure runout and keyway fit before blaming the belt |
| Tension | A slack belt adds play; an over-tight belt adds bearing load and noise without removing clearance | Set tension by deflection or sonic meter, then re-check after run-in |

Design choices that reduce backlash
- Use the finest pitch that still carries the load. A 2 mm or 3 mm pitch moves less per tooth than an 8 mm or 14 mm pitch.
- Maximise teeth in mesh. A larger driven pulley, or a layout that wraps more than 180 degrees, spreads torque over more teeth and cuts deflection.
- Keep the span short and stiff. A long unsupported span behaves like a spring; a shorter centre distance or an intermediate support helps.
- Choose the cord for stiffness. Steel cord stretches less than aramid under steady load, while aramid tolerates flex fatigue better. Match the cord to the duty cycle.
- Tension to the belt maker value, not above it. Extra tension cannot remove designed tooth clearance, and it costs bearing life.
- Close the loop where the process demands it. On positioning work an encoder, or a second belt opposed to the first, is often more effective than chasing the last hundredth of a millimetre with tension.

What to measure before changing anything
- Measure lost motion at the driven pulley with a dial indicator, loading the belt in both directions and reading the difference.
- Compare that reading with pulley runout. If runout is a large share of the total, the pulley is the problem.
- Inspect the tooth flanks. A worn groove increases play, and no new belt will remove it.
- Record the tension value and the method used, because a deflection gauge and a sonic meter can disagree by a meaningful margin.
Treat every figure on this page as a starting range. Backlash depends on the whole drive, and a number quoted without pitch, span and tension cannot be transferred to your machine.
Tell us the axis and the pitch
Email sales@goodlybelts.com with the belt marking, pitch, centre distance and the positioning tolerance you need, or use the enquiry form. We will suggest a pitch, width and cord combination that keeps lost motion inside your tolerance, and quote it.
Ask for a backlash checkAre you looking for a reliable manufacturer of conveyor belt products?
We can quickly provide customers with market analysis, technical support and customized services.Please send your message to us
- Name
- Phone
- *Title
- *Content


























