Breadcrumb: Home > Solutions > Fluorescent Inspection
Updated:2026-11-02
Views: Three tasks come up almost daily in optical glass quality control: measuring how a lens blocks and transmits UV, checking surface cleanliness before coating, and looking for bubbles and stones inside a blank. They map to three bands - 365nm for fluorescence, 302nm for mid-wave blocking, 254nm for short-wave scattering. Three lamps used to be needed, taking up the bench and requiring re-aiming. The LELMS-38L puts three 8W tubes into a 330mm housing, with three switches, one per channel. The sections below cover the band principles, four typical uses on an optics bench, the conditions that make a reading valid, and the safety points that keep the inspection repeatable.
Three actions come up almost daily in optical glass quality control: measuring how a lens blocks and transmits UV, checking surface cleanliness before coating, and looking for bubbles and stones inside a blank. The three tasks map to three bands - 365nm for fluorescence, 302nm for mid-wave blocking, 254nm for short-wave scattering. Three lamps used to be needed, filling the bench and requiring re-aiming on every switch. Putting three 8W tubes side by side in one 330mm housing, with three switches at the tail, one per channel, gives the LELMS-38L handheld three-wavelength UV lamp. The sections below cover the band principles, four typical uses, the preconditions for a valid reading, and safety and maintenance.
UV absorption and scattering in glass are set jointly by the substrate, the dopants and the surface coating. The three bands interact with glass in different ways and reveal different information; mixing them up leads to wrong conclusions.
| Band | Center intensity at 10cm | Interaction with glass | What can be read | Common misuse |
|---|---|---|---|---|
| 365nm UV-A | 1236 μW/cm² | Low energy, penetrates fairly well, excites fluorescence in organic substances and some inorganic inclusions | Surface grease, polishing wax, mold release residue; cleanliness before coating; fluorescence differences of some substrates | Reading in a bright room, where weak fluorescence is swamped by ambient light |
| 302nm UV-B | 848 μW/cm² | Intermediate energy, approximating the mid-wave UV part of sunlight | Mid-wave blocking capability of lenses, protective eyewear and automotive glass; pre-positioning of UV adhesive | Using 365nm data in place of a mid-wave conclusion; the two do not convert |
| 254nm UV-C | 1170 μW/cm² | High photon energy, strongly absorbed and scattered by substrate and coating | Short-wave transmission of quartz glass and UV optics; scattering bright spots from bubbles, stones and microcracks | Looking into the spot for long periods, or ignoring ozone and corneal irritation |
One sentence to remember it by: 365nm shows "what is there", 302nm shows "does it hold back mid-wave", 254nm shows "does short wave get through, and is there anything inside". The three tubes switch independently precisely so that each conclusion maps to one band.
The housing is 330×100×80 mm, metal with a scratch-resistant coating; three rocker switches at the tail sit under the thumb so the band can be changed with one hand; the aperture is covered by a black UV filter of about 20×6 cm, which removes visible light so the background is darker when reading fluorescence; inside is an orange-peel aluminium reflector that returns light from the back of the tubes to the aperture, making the central spot more even. 2.5 m power cord, 220V supply (a 110V version is available), full-transparency UV protective glasses included, and a benchtop stand for long fixed use.
The length of the body sets its working boundary: a 330 mm tube suits bench work where parts are handled one at a time - piece-by-piece inspection, sampling, rework confirmation. For line inspection that has to cover a whole panel or batch at once, moving the body in sweeps is not efficient; a surface source or an irradiation unit with closed-loop control is the better fit.
This is a high-frequency item in incoming inspection of optical lenses, protective eyewear, automotive optical glass and UV filters. A handheld lamp can do a quick screen, but the order cannot be skipped:
One caution concerns the gap between a label and the band actually assessed. The common "UV400" marking means blocking UV below 400nm, while the assessment point often falls in the 380-400nm edge region; the 365nm channel of the LELMS-38L gives data at 365nm only and cannot be extrapolated to behavior near 400nm. For incoming sampling, batch comparison and post-rework recheck, an internal judgement is fine; an external test report still needs a spectrophotometer to provide the full spectral curve.
Three details that often go wrong: the spot is not aimed at the sensing area, so a low reading is misjudged as "good blocking"; the lamp is held unsteadily and a few centimeters of distance drift makes the reading jump; the reading is taken right after switch-on while the tubes are not yet stable. Fixing the body on a stand and warming up for a few minutes markedly improves repeatability.
In the coating process, residues left on the surface - polishing liquid, polishing wax, finger grease, traces of cleaning agent - cause pinholes, coating lift-off and local detachment. Many of these residues are the same color as glass under visible light and cannot be seen by eye, but they fluoresce under 365nm long-wave UV.
Operating points: work in a dark room or screen the workstation with a light-blocking cloth, switch on the 365nm channel alone, hold the body at about 45 degrees to the surface, and observe from a direction close to the normal - separating the angle of incidence from the viewing angle raises fluorescence contrast. Clean areas appear dark; contamination appears as bright white, bright blue or bright yellow-green spots and smears, graded by area and brightness to decide whether rework cleaning is needed.
Two preconditions for reading have to be stated: fluorescence is a screening method, not a composition analysis; some cleaning agents contain optical brighteners and leave false positives on a clean surface, while inorganic salts, water marks and some inorganic dust do not fluoresce and stay invisible at 365nm, so white-light grazing illumination or dark-field lighting is needed as a cross-check. For high-precision parts, locate by fluorescence and then confirm under a microscope.
Blanks of quartz glass and UV optical components are screened by scattering rather than fluorescence. 254nm photons carry high energy and scatter strongly at the refractive-index discontinuities of bubbles, stones and microcracks, showing up as bright spots or bright lines under oblique illumination in a dark room; 302nm mid-wave penetrates some substrates to a different depth and serves as a comparison view.
The procedure is to switch off 365nm, use 254nm only, hold the body close to the side or bevel of the workpiece, keep the eyes out of the direct beam and move slowly along the part. Scattering points that sit inside the workpiece rather than on the surface can be told apart from surface dust: surface dust wipes off, while an internal defect stays in the same place. This check covers transparent regions and near-surface zones; deep defects and defects under a coating still need an interferometer or microscopic imaging.
Lens assembly, prism bonding and fiber coupling often use UV-curing adhesive for temporary positioning: dispense, align, pre-cure for a few seconds to tens of seconds to lock the position, then continue to the curing step. 365nm is the positioning wavelength used in much of this work, with 302nm for adhesives that need mid-wave initiation or have a shorter absorption peak.
The role of a handheld lamp here is positioning, not final cure: a small spot at close range suits local spot curing; full-area curing needs a uniform surface source or a curing chamber, otherwise the degree of cure differs across one workpiece. Set the exposure time by calculating back from the energy value in the adhesive datasheet, not by habit.
The catalogue intensity is specified "at 10cm, dead center, new tubes"; leave those three conditions and the number in your hand changes. Understand this table before making quantitative records:
| Variable | Effect on the reading | Practical action |
|---|---|---|
| Exposure distance | The tube is a line source; at 20cm the value is already well below 10cm | Keep 10-15cm for reading detail; for large areas move back and extend the time |
| Off-center position | Lower at the two ends and outside the long edges of a 330mm body | Sweep along the long edge in steps of about half the lamp length, and touch up edges and corners separately |
| Tube aging | Output falls with cumulative burning hours, and the eye barely notices | Keep a burning-hour log, replace on schedule, re-measure with a radiometer after replacement |
| Filter contamination | Dust, fingerprints and reagent residue absorb UV | Wipe gently with anhydrous ethanol and lens tissue, not a dry coarse cloth |
| Supply and temperature | Output is low during cold start | Warm up a few minutes before reading or timing |
To convert to dose there is one formula: dose = intensity × time. Taking 254nm at 10cm center as an example, 1170 μW/cm² is 1.17 mW/cm²; for a target dose of 10 mJ/cm² the theoretical exposure is about 8.5 seconds. That holds for a new tube at dead center, so in practice substitute the radiometer reading at the working position, recalculate, and leave a margin.
254nm is the higher-risk channel in this lamp; it irritates the cornea and skin quickly, and a few seconds of looking at reflected light can cause discomfort. Put on the supplied protective glasses before switching on, and keep the beam facing away from people; where long exposure is needed, have personnel leave and post a notice at the door. The erythema effect of 302nm should not be ignored either - do not judge safety by "it doesn't feel hot". 254nm also reacts with oxygen in air to produce a small amount of ozone, so ventilate a small closed space after long exposure before people re-enter.
The filter is a glass part; after a drop or impact, check it against the light for cracks even if it looks intact - a crack lets visible light leak in, the fluorescence background brightens at once and all reading is distorted. Clean with anhydrous ethanol and lens tissue; abrasive cleaners are not to be used. Replace tubes on a schedule: common industry practice is about 1000 cumulative hours or 12 months of use, whichever comes sooner, with all three channels re-measured after replacement. The 110V version must be matched to the corresponding supply; the wrong voltage will damage the tube circuit directly.
Yes, the three rocker switches control one channel each, so single, double or all three can be on. For a judgement, though, switch on one channel at a time: with several bands superimposed, fluorescence, scattering and visible residue mix together and it becomes unclear which channel produced the effect. Switching all on suits work that needs broad-band illumination rather than precise reading.
Both are in the UV-B band, and the difference in nominal peak comes from the phosphor formulation of the tube. Real output is a band spectrum rather than a single line, so the nominal peak and intensity are what matter when selecting. If a customer's inspection document specifies 312nm in writing, run a comparison check before substituting, not a direct swap.
Not advisable. A handheld lamp gives a relative value at one point and one wavelength, affected by distance, angle and spot position, which suits internal judgements such as incoming sampling, batch comparison and rework recheck. An external report needs full-spectrum data and a traceable calibration chain, which a spectrophotometer provides. The two divide the work rather than replace each other.
Not necessarily. Whether a part passes depends on the assessed band and limit given in the product standard, not on whether the number looks high enough. Lenses of the UV400 type are assessed across all bands below 400nm, and 99% at 365nm alone does not show that 395nm also meets the requirement. Confirm the standard clause before measuring, then take values with an instrument at the corresponding wavelength.
Check in order: whether the baseline was re-measured, whether the distance and angle between sample and probe were the same, whether the spot was aimed at the sensing area, whether the tubes had warmed up, and whether there is a fingerprint on the filter. A change in any of these five is enough to shift the reading by more than ten percent.
The eye is not reliable for this. Keep a log of cumulative burning hours and periodically re-measure all three channels at a fixed position with a radiometer, comparing against the initial values; common industry criteria are output falling to about 70% of the initial value, or about 1000 cumulative hours or 12 months of use, whichever comes sooner.
A handheld lamp suits single parts and sampling. Where full-panel coverage and continuous exposure at a fixed station are needed, a tube-type surface source is an option, such as the LEAC-280L handheld dual-wavelength UV lamp (254+365nm, good filter contrast) or the XX-15 / XX-20 tube lamps in the same section; where quantitative exposure and temperature control are needed, an irradiation unit with closed-loop control is the fit. Three things decide the choice: how much area must be covered, how long the exposure runs each day, and whether the dose has to go into a record.
In UV inspection for the optics industry, the hard part is not switching the lamp on but fixing four things: which band, at what distance, for how long, and what reading. The value of the LELMS-38L is that it puts three bands into a housing one hand can hold, so the same sample at the same position can be viewed band by band; what decides whether a conclusion holds is still the habit of measuring a baseline, cross-checking with a radiometer and keeping a tube log. Do those three and a handheld lamp can still deliver a reproducible judgement.
To set a band combination and exposure condition for a specific process, send us the sample type, inspection standard and current workstation layout and we will check the selection together. Related reading: Using black lights in fluorescent penetrant testing, Workstation layout and compliance requirements for large-area UV inspection lamps, UV radiometers and optical power meters.
This article belongs to the fluorescence inspection section.
Other sections on the same product line include laboratory UV lamps (254 / 302 / 365nm single-tube and dual-wavelength benchtop lamps) and life science instruments (the entry point for fluorescence imaging, UV crosslinking, UV transmission and cell irradiation).
Models mentioned in this article: LELMS-38L, LUV-4B, ENF-280C. Ranges, accuracy and accessory lists for each model are listed on the product page.
Email: info@xinysun.com
Tel: 131-2093-6518
Address: 1525 Minqiang Road, Bldg 26, Songjiang, Shanghai
Copyright 2026 Xinysun Instrument (Shanghai) Co., Ltd. All rights reserved.
ICP:ICP-2026046447
