Local Shop or Maker: The Three-Tier Boundary on Lens Faults

Service authorization is a scope decision, not a loyalty test. On a sealed industrial lens, the boundary between work a local shop may complete and work which must return to the maker sits at three things: the enclosure seal, the cemented groups, and the factory record keyed to the batch number. Consumer camera guidance puts the boundary at spare parts and firmware. A sealed endoscope objective contains neither. Lensmind Optics, registered in 2019, runs cementing, cleanroom assembly and final metrology in-house in Hangzhou, and every production batch ships with a test report stating MTF, distortion, back focal length and depth of field. FA family distortion is held below 0.05% of the field, which is a 0.05 mm budget across a 100 mm field. According to IEC 60529, the IP code defines how an enclosure resists ingress, and in practice that clause is what an opened barrel must prove again.

Which Lens Faults Can a Local Shop Touch and Which Must Go Back to the Maker?

A repairable fault is a fault whose fix does not break the sealed optical cavity and does not re-open a cemented joint. Under the cavity test, a local shop can own three groups of work: mount threads and adapters, external mechanics such as focus rings and set screws, and a protection window held in its own retainer. Everything else belongs upstream. Cavity faults are internal contamination, coating damage on a trapped surface, a shifted doublet, and any change which moves back focal length outside the drawing tolerance, which on an FA lens includes the 0.05% distortion budget. Those faults need a cementing bench and a centration station, not a clean mat and a steady hand. 100% of shipped batches left the factory with a test report. No local shop rebuilds that report from a bare bench. According to ISO 9001:2015, clause 8.7 requires a documented decision on every nonconforming output, and in practice the decision is written before the lens is packed.

Why Does the Consumer Camera Repair Rule Fail on a Sealed Industrial Lens?

The consumer repair rule is a spares-and-software test: a fault stays local when parts are obtainable and no factory calibration software is required. Three properties of a sealed industrial lens break the test. First, an endoscope objective and a fixed-focus FA lens carry no autofocus motor, no image stabilizer and no firmware, so the usual serviceability question has no target. Second, the cavity carries an ingress claim, and an IP67 or IP68 rating belongs to the sealed assembly rather than to individual parts. Third, the acceptance figure lives in the factory record: 0.05% distortion on an FA lens is a bench result at a stated image height, not a property a technician can eyeball. For example, the same 0.05% on a lens measured at one field position says nothing about the corner. According to ISO 14644-1:2015, air cleanliness is classified by airborne particle concentration, and a repair bench holds no class at all.

How Do the Three Service Tiers Split?

A service tier is a permission level tied to how much of the optical cavity a procedure opens. Tier 1 covers external work: mount threads, adapters, focus rings, set screws and a protection window seated in its own retainer. Tier 2 covers the front retaining ring and the outer face of the first element, with the cavity still closed. Tier 3 covers everything inside the cavity: spacers, cemented groups, the rear group, and any work which moves back focal length or eats into the 0.05% distortion budget. The tier decides the paperwork, not the fault name. A scratched protection window and a scratched first element look identical to a buyer and sit two tiers apart. According to ISO 9001:2015, clause 7.5 requires documented information to be controlled, so the authorization, the measurement and the sign-off belong in one file. In practice, 100 percent of tier assignments are written before a tool touches the barrel.

What Goes Into Each Service Tier?

A service tier table is a one-page map which names what each tier may open, what the work puts at risk, and who signs the result. The table is filled in before the barrel is touched, never after a finding appears.

Tier	What the shop may open	What is lost	Who signs off
1	Mount thread, focus ring, filter ring, set screws, window retainer	Nothing optical	Shop QA, thread gauge result photographed
2	Front retaining ring and the outer face of the first element	Factory seal claim until re-tested	Maker reviews bench data before release
3	Cavity, spacers, cemented groups, rear group	Seal, cement, traceability to the batch record	Maker only, with a fresh test report

For example, a scratch on a protection window sits in Tier 1 and costs nothing optical, while the same scratch on the first element sits in Tier 2. A 0.05% distortion budget cannot be reconstructed afterwards. According to ISO 9001:2015, clause 7.5.3 asks for the record to be controlled, and the table is the record.

What Breaks Inside the Barrel the Moment It Is Opened?

The optical cavity is the sealed volume which holds elements, spacers and cemented groups at their aligned positions. Opening the volume starts three clocks at once. First, particle load: a workshop holds no air cleanliness class, so dust settles on surfaces which were closed in a Class 10,000 cell zoned for medical builds. Second, moisture path: each gasket and O-ring has to return to the same groove, and a pinched seal turns an IP67 lens into an unrated one. Third, alignment: elements centred on a 1 μm-class active alignment station no longer sit where the drawing assumed, and a shift of a few micrometres shows up inside the 0.05% budget rather than outside it. Only the third loss appears on a 2026 acceptance test. In practice, the first two surface months later as haze, as a failed ingress check, or as a warranty argument nobody can settle. According to ISO 14644-1:2015, a cleanroom class is defined by particle concentration.

Which of the Three Losses Is Permanent?

A permanent loss is a loss which no later measurement can restore. Seal and alignment are recoverable when tooling and bench exist: a maker can re-seal a barrel, re-centre a group and prove both again. Traceability is not recoverable in the same way. Once a lens falls outside the factory record (the batch number, the test conditions and the measured MTF and distortion values on the original build), no local measurement re-creates the baseline. A shop can report what the lens does today. The shop cannot report what the lens did when the batch passed. 0.05% distortion at final test and 0.05% measured after a repair are different statements, and only the factory holds the first one. According to ISO 9001:2015, clause 8.5.2 requires identification and traceability where necessary, and a lens baseline is the case in point.

Can a Cemented Doublet Be Re-Glued Outside the Factory?

Cementing is the bonding of two or more elements into one group with an optical adhesive whose cure schedule and centration are controlled together. A local shop can physically re-glue a separated doublet. The shop cannot match the three variables which decide whether the group still works: adhesive refractive index and cure profile, centration of the group on the barrel axis, and a bench which repeats the original acceptance numbers. For example, a doublet which looks clean under a lamp can sit off axis and still pass a visual check. 0.05% distortion is a whole-lens figure measured at final test, and a re-cemented group moves the figure in ways a workshop cannot see. According to ISO 9001:2015, clause 8.5.1 requires controlled conditions for production, and cementing is a production process rather than a bench repair. Re-cementing outside the factory turns a repairable lens into a lens with no baseline.

What Should the Maker Put in Writing Before Tier 1 or Tier 2 Work Starts?

A service authorization letter is a short document which names the permitted procedures, the values to be restored, the measurement which proves restoration, and the person who signs the result. Four clauses carry the weight. 1. A closed list of permitted operations, written as procedures rather than intentions. 2. The values to be restored, quoted from the batch report: back focal length, the distortion budget at 0.05% of the field on an FA lens, and the MTF figure at a named image height. 3. The re-test method and its conditions, including aperture and field positions. 4. The void condition, which states which findings end local work and start a return. Lensmind Optics has held ISO 9001 certification since 2024, and the audit trail behind the certificate is what lets a maker sign such a letter. According to ISO 9001:2015, clause 8.4 requires control of externally provided processes.

Which Documents Travel With the Lens and Which Stay With the Maker?

The lens data packet is the set of records which lets a third party judge a lens without calling the factory. Four items belong in the packet: the batch test report with MTF, distortion, back focal length and depth of field plus test conditions and inspection coverage; the drawing with the tolerances that matter at the mount; the serial-to-batch link; and the coating run record for surfaces carrying a transmission claim. A local shop needs the first item and cannot use the fourth. For example, a shop can check a mount thread with a gauge and cannot check a 0.05% distortion budget without a target grid. Our batch data pairs each report with a batch number and the test conditions, which is what makes the comparison possible. According to ISO 10110-1:2019, drawings state optical tolerances through standardized indications, so the drawing travels with the report.

Who Re-Verifies the Optics After the Local Shop Finishes?

Post-repair verification is a measurement pass which repeats the factory acceptance values on the same parameters, under stated conditions, before the lens returns to the line. Three parties can own the pass, and the choice has to match the tier. For Tier 1 the shop owns a mechanical check: thread gauge, torque, and a photograph of the gauge result. For Tier 2 the maker owns the optical pass, because the shop holds no baseline for the sealed assembly. For Tier 3 the maker owns everything, including the re-seal and the ingress re-test. The signature matters less than the conditions recorded beside it. A distortion figure without an image height and an aperture is not a measurement, and 0.05% of a 100 mm field is 0.05 mm of allowed error. The as-installed lens record is where the signed result belongs, so the next drift has a comparison point. According to ISO 9001:2015, clause 8.6 controls release.

Is a Collimator and an Interferometer Enough to Qualify a Local Shop?

Equipment adequacy is a match between instrument and parameter, not a count of instruments on a bench. A collimator answers a different question from an interferometer. A collimator produces a wavefront from infinity and lets a shop measure back focal length and focus position. A laser Fizeau interferometer measures surface figure or transmitted wavefront of a flat or a sphere in a fixture. Neither instrument reports MTF at the working magnification unless the bench is built for the measurement, with a target at the specified object distance and a calibrated image analyzer. Centration needs a rotation stage and a software fit, which is what a Trioptics OptiCentric instrument provides. For example, the 0.05% distortion budget needs a target grid and a distortion calculation, not a look at a star target. A shop with three instruments and no grid cannot reproduce 0.05%. According to ISO 9001:2015, clause 7.1.5 requires measuring equipment to be fit for purpose.

What If the Fault Is Dust Inside the Front Group?

Internal contamination is particulate matter trapped between elements, and the correct tier depends on which surface holds the particle. A particle on the outer face of the first element is reachable after the front retaining ring comes off, which is Tier 2 work: the cavity stays closed and the seal stays untouched. A particle between the first and second elements is Tier 3 work, because the group has to come apart and the cavity opens. Position matters more than size. According to ISO 14644-1:2015, a cleanroom class describes airborne particle concentration, and a Class 10,000 cell is where the assembly was closed. Cleaning a trapped surface outside a controlled cell re-contaminates that surface within minutes of exposure. In practice, 100% of internal cleaning belongs in a controlled cell, followed by the same test sequence recorded in the batch report, and the 0.05% distortion budget is checked again at the end.

How Do You Write the Disposition So the Next Failure Is Traceable?

A disposition record is the entry which states what was found, what was done, who did the work, and which measurement released the lens. Six fields make the entry usable: the serial and batch number; the fault as observed before cleaning; the tier assigned and the authorization reference; the procedure performed, including any deviation; the measurement which released the lens, for example 0.05% distortion at a named image height with its aperture; and the signature with a date. The returned lens disposition record covers the inspection side in more detail. A record which lists the action but not the measurement cannot be compared with the batch report. A record which lists the measurement but not the conditions cannot be repeated. The bench evidence note shows the same field set on an impact case. According to ISO 9001:2015, clause 8.7 makes the disposition the controlled record.