Can a correct gear backlash still coexist with poor tooth contact?

The short answer

Yes. Backlash measures a clearance relationship, while tooth contact describes where mating flanks carry load. A gear pair can have an acceptable local clearance but concentrate contact near one edge. Wind-yaw mesh acceptance should address both quantities using the selected manufacturer’s method.

Separate free motion from loaded geometry

Backlash concerns relative movement before opposite flanks engage under the defined measurement condition. Contact pattern concerns the distribution of engagement over the tooth surface. Shaft alignment, mounting-face geometry and tooth modifications can affect that pattern without being revealed by one backlash number. The method and applied load used to obtain a pattern also matter; an unloaded visual mark is not automatically a full-load distribution.

Check why local contact can change

A tilted pinion can contact near one face edge even if its centre-distance clearance seems reasonable. Deflection under torque can change that relationship again. Enlarging backlash does not inherently correct the alignment or support stiffness causing concentrated contact. Conversely, a broad-looking pattern does not establish adequate minimum clearance around an eccentric ring. The two checks answer different questions and should not be traded against each other.

Request two clearly defined acceptance results

Ask the supplier for the required backlash assessment and the separate contact-pattern criteria, including load, direction and ring position. Rothe Erde installation support includes both backlash setting and evaluation of gear contact, demonstrating their distinct roles. Its criteria do not establish MARSTON gear limits. A satisfactory clearance report should therefore not close an unresolved contact concern, and a pattern discrepancy should return to the gear designer rather than prompt undocumented alignment or tooth modification. A contact assessment should identify whether the pattern was obtained before or after the installed structure carried load. Deflection can change flank distribution, so an unloaded pattern has a deliberately limited evidence boundary.