Device Designarchiwum projektu „100 kampanii”

Lampa warsztatowa z lupą, regulowaną bielą i oświetleniem kontrastowym — wymagania v1

Archiwalna wersja 3 tej strony. Arkusz zmienił się po publikacji; najnowsza treść: /dd/produkty/lampa/wymagania/v1/.

Zaprojektuj lampę biurkową dla inżynierów do elektroniki i precyzyjnej pracy ręcznej, która łączy dużą szklaną lupę, regulowaną biel o wysokim CRI z natywnych białych diod (1800–6000 K) i przełączane oświetlenie kontrastowe z wielu kątów, dzięki któremu widać kontury, rysy i kształt lutów — wszystko w jednej, dobrze sterowanej głowicy na długim, stabilnym ramieniu, w cenie wyraźnie niższej niż dzisiejsze lampy z lupą klasy premium.

Status arkusza
wydany — zamrożony; poprawki trafiają do następnej wersji
Data
30 października 2026
Autor
W3 (agent-derived from market, photometry and standards research; v1 = v0 corrected after the fog-sheet review findings were applied to the lamp; Marcin's answers will become v2)
Liczba wymagań
59
Zdanie dla wariantu (po angielsku)
Design an engineering desk lamp for electronics and fine manual work that combines a large glass magnifier, tunable high-CRI white light from native white LEDs (1800-6000 K) and switchable multi-angle contrast lighting that makes contours, scratches and solder shapes visible, in one well-controlled head on a long, stable arm, at a price well below today's premium magnifier lamps.

Te wymagania widzi każdy wariant i według nich oceniamy projekty. Cel to poziom mocnego projektu, próg to minimum — poniżej progu wymaganie nie jest spełnione.

Wydajność (19)

LAM-PRF-01 waga 5/5

Tunable native-white range

White light is continuously tunable over this CCT range using only native (phosphor-converted) white LEDs, in steps of at most 100 K (target 50 K).

Cel
1800-6500 K, steps ≤ 50 K
Próg
1800-6000 K, steps ≤ 100 K
Miara
continuous CCT range of the white mode
Oceniane
od koncepcji
Jak sprawdzamy i źródło

obliczenie, sędziowie, plan testów — LED channel selection (datasheets) and mixing calculation in CIE 1960 uv; spectrometer sweep in the test plan.

Źródło: Marcin (30.10): native white LEDs 1800-6000 K; native 1800 K parts exist (Nichia NFSW757H-V1 R9050, Yuji BC 1800 K) — research.md §3

LAM-PRF-02 waga 4/5

Chromaticity on the blackbody curve

Over the whole CCT range and at 100 %, 10 % and 1 % output, the white point stays within this distance from the Planckian locus.

Cel
≤ 0,00 Duv
Próg
≤ 0,01 Duv
Miara
maximum |Duv| over the CCT range and dimming levels
Oceniane
od koncepcji
Jak sprawdzamy i źródło

obliczenie, plan testów — Mixing calculation from the LED datasheet chromaticities (worst case between channels, incl. temperature shift); spectrometer measurement at several CCTs and levels.

Źródło: two-channel 1800/6000 K mixing gives Duv ≈ -0.014 mid-range (CALCULATION, research.md §3); ANSI C78.377 tolerance ±0.006

LAM-PRF-03 waga 4/5

Colour rendering (CRI)

Colour rendering index Ra of the mixed white output at every CCT from 2700 K to 6000 K; at 1800-2700 K at least Ra 90.

Cel
≥ 95 Ra
Próg
≥ 90 Ra
Miara
minimum CRI Ra from 2700 to 6000 K
Oceniane
od koncepcji
Jak sprawdzamy i źródło

obliczenie, plan testów — LED datasheet values for each channel and the mix; spectrometer measurement.

Źródło: premium lamps Ra 80-90 (Luxo CRI 80, Waldmann TEVISIO Ra ≥ 90); Nichia Optisolis Ra ≥ 95 — research.md §2-3

LAM-PRF-04 waga 3/5

Saturated red rendering (R9)

Special colour rendering index R9 (saturated red, e.g. wire insulation and resistor bands) at every CCT from 2700 K to 6000 K.

Cel
≥ 80 R9
Próg
≥ 50 R9
Miara
minimum R9 from 2700 to 6000 K
Oceniane
od koncepcji
Jak sprawdzamy i źródło

obliczenie, plan testów — LED datasheet values; spectrometer measurement.

Źródło: only Waldmann publishes R9 (≥ 50); Optisolis R9 ≥ 80-85 — research.md §2-3

LAM-PRF-05 waga 2/5

Camera colour fidelity (TLCI)

TLCI-2012 of the white output from 2700 K to 6000 K, so photos and video of the work look right.

Cel
≥ 90 TLCI
Próg
≥ 85 TLCI
Miara
minimum TLCI from 2700 to 6000 K
Oceniane
w pełnym projekcie
Jak sprawdzamy i źródło

obliczenie, plan testów — Computed from the mixed spectrum (EBU Tech 3355 software) and confirmed by measurement.

Źródło: EBU Tech 3355 TLCI bands (85-100 top band) — research.md §3

LAM-PRF-06 waga 4/5

Illuminance at the work plane

Illuminance at the centre of the work plane at maximum white output at 4000 K; at 1800 K at least 40 % of this value.

Cel
≥ 4 000 lx
Próg
≥ 2 000 lx
Miara
illuminance at the work-plane centre, 4000 K, maximum output
Oceniane
od koncepcji
Jak sprawdzamy i źródło

obliczenie, plan testów — Photometric calculation from LED flux, optics and geometry (or ray tracing); lux-meter measurement.

Źródło: EN 12464-1 precision work 1000-2000 lx, raise one step for small low-contrast tasks; watchmaking 1500-3300 lx; market claims 3000-6000 lx at unstated distances — research.md §3

LAM-PRF-07 waga 3/5

Uniformity and no multiple shadows

Uniformity (Emin/Eavg) over a 150 mm diameter field at the work plane in the all-around white mode; tools and fingers cast one soft shadow, not several sharp or coloured ones.

Cel
≥ 0,70 ratio
Próg
≥ 0,60 ratio
Miara
Emin/Eavg over a Ø150 mm field
Oceniane
od koncepcji
Jak sprawdzamy i źródło

obliczenie, plan testów — Illuminance map (calculation or ray trace); measured grid and a shadow photograph in the test plan.

Źródło: EN 12464-1 Uo 0.70 for fine work; multiple shadows are a common complaint — research.md §3, §5

LAM-PRF-08 waga 2/5

Dimming range

Each light channel dims continuously from 100 % down to this level with a smooth (perceptually even) curve and no visible steps.

Cel
≤ 0,50 %
Próg
≤ 1 %
Miara
lowest stable output level
Oceniane
od koncepcji
Jak sprawdzamy i źródło

sędziowie, plan testów — Driver and dimming method (resolution, analog/PWM) reviewed; measurement at low levels.

Źródło: market 2-4 steps or 10-100 % (research.md §2)

LAM-PRF-09 waga 4/5

Flicker-free light

At every dimming level and CCT, light modulation is in the IEEE 1789 no-observable-effect region (target) and meets EU limits Pst LM ≤ 1.0 and SVM ≤ 0.4 (threshold); no banding on phone or camera video at 1/1000 s.

Cel
IEEE 1789 no-observable-effect at all levels; Pst LM ≤ 0.1 and SVM ≤ 0.1
Próg
IEEE 1789 low-risk at all levels; Pst LM ≤ 1.0 and SVM ≤ 0.4
Miara
flicker compliance at 100 / 50 / 10 / 1 % output
Oceniane
od koncepcji
Jak sprawdzamy i źródło

sędziowie, plan testów — Driver topology and PWM frequency reviewed; photodiode measurement of modulation at four levels.

Źródło: IEEE 1789-2015; Regulation (EU) 2019/2020 as amended by 2021/341 (SVM ≤ 0.4 from 1 Sept 2024); no magnifier lamp publishes flicker data — research.md §3

LAM-PRF-10 waga 4/5

Segmented light around the lens

The light around the lens is split into independently dimmable azimuth sectors, so light can come from chosen sides (shadow direction) and the sector blocked by the hand can be switched off.

Cel
≥ 8 sectors
Próg
≥ 4 sectors
Miara
number of independently dimmable sectors around the lens
Oceniane
od koncepcji
Jak sprawdzamy i źródło

sędziowie, plan testów — LED layout and driver channels in the schematic and CAD.

Źródło: Waldmann TEVISIO 4 segments (on/off scenes); microscope 4-zone rings — research.md §2

LAM-PRF-11 waga 4/5

Low-angle (raking) light

A low-angle light layer lights the work from a grazing elevation so that relief, scratches and edges stand out; it has its own sectors (at least 4) and works without moving the main head out of position.

Cel
≤ 20 deg
Próg
≤ 35 deg
Miara
elevation angle of the low-angle light at the centre of a Ø100 mm field
Oceniane
od koncepcji
Jak sprawdzamy i źródło

obliczenie, sędziowie — Geometry calculation (elevation = atan(height / radius) at field centre and edges) from CAD.

Źródło: dark-field / raking light is 10-30° elevation; a ring around the lens at 250 mm height gives ~60-70° only (CALCULATION, research.md §4)

LAM-PRF-12 waga 3/5

Multi-tier colour contrast

Like an AOI multi-layer ring, at least two (target three) elevation tiers can each be set to red, green, blue or white, so the colour seen on a surface tells its slope (e.g. solder fillets).

Cel
≥ 3 tiers
Próg
≥ 2 tiers
Miara
elevation tiers with selectable R/G/B/white
Oceniane
od koncepcji
Jak sprawdzamy i źródło

sędziowie — LED layout, tier elevations and channel map reviewed.

Źródło: Marcin's reference - Vision Datum KR multi-layer RGB ring; Omron colour highlight (R high, G mid, B low) — research.md §4

LAM-PRF-13 waga 4/5

Contour visibility

On three reference samples (scratched polished aluminium, an embossed/engraved part, a PCB with solder fillets), photographed through the lens at fixed exposure, the best contrast mode raises the Michelson contrast of the features by at least this factor over plain uniform white light.

Cel
≥ 3 x
Próg
≥ 2 x
Miara
contrast gain of the best contrast mode vs uniform white (Michelson contrast ratio), minimum over the samples
Oceniane
od koncepcji
Jak sprawdzamy i źródło

plan testów, sędziowie — Test plan contains the three-sample camera test with exposure fixed; concept judged on the lighting geometry.

Źródło: proposal from machine-vision practice (research.md §4); value is an ASSUMPTION to be calibrated on a prototype

LAM-PRF-14 waga 2/5

Glare-free mode for shiny parts

A cross-polarised mode (polarised illumination + analyser in the viewing path) suppresses specular glare from solder, metal and glossy surfaces.

Cel
≥ 1 000 :1
Próg
≥ 100 :1
Miara
polariser extinction ratio of the cross-polarised mode (500-700 nm)
Oceniane
w pełnym projekcie
Jak sprawdzamy i źródło

sędziowie, plan testów — Polariser films (datasheets), placement and light budget reviewed; glare test on a shiny sample.

Źródło: polariser films 100:1 to 10,000:1; cross-polarisation costs ~7-10x light — research.md §4

LAM-PRF-15 waga 1/5

UV inspection light

A 365 nm UV-A channel for conformal-coating and flux-residue inspection, eye-safe at the working position.

Cel
365 nm with visible-cut filter, exempt group (RG0) at 200 mm
Próg
not required; if present, 365 nm and at most RG1 with a warning label
Miara
UV channel at 365 nm and its IEC 62471 risk group at the working position
Oceniane
w pełnym projekcie
Jak sprawdzamy i źródło

sędziowie — UV LED and filter parts, irradiance estimate vs IEC 62471 limits.

Źródło: IPC-CC-830 fluorescence test at 365 nm; IEC 62471 UV-A limits — research.md §4

LAM-PRF-16 waga 4/5

Main lens size

Clear aperture of the main glass lens, large enough for comfortable two-eye viewing. For a rectangular lens the value is the diameter of the circle of equal area, and its shorter side is at least 110 mm.

Cel
≥ 160 mm
Próg
≥ 150 mm
Miara
clear aperture diameter (round lens) or equal-area diameter (rectangular lens)
Oceniane
od koncepcji
Jak sprawdzamy i źródło

automatycznie, sędziowie — Lens part (datasheet or drawing) and CAD aperture.

Źródło: TEVISIO 160 mm, Daylight 175-177 mm, Luxo 127 mm / 171 × 114 mm — research.md §2

LAM-PRF-17 waga 4/5

Magnification and working distance

Main lens of 3-3.5 D (M = 1 + D/4 ≈ 1.75-1.88x) with this free working clearance under the head, so that tools and hands fit between the head and the work.

Cel
≥ 240 mm
Próg
≥ 200 mm
Miara
free clearance from the lowest point of the head to the work plane with the main lens
Oceniane
od koncepcji
Jak sprawdzamy i źródło

automatycznie, obliczenie — Focal length from the lens power; clearance measured in 03-cad/assembly.step with the head at the focal position.

Źródło: 5 D lamps are too close for brush and iron (user reports); Luxo 3 D WD 292-318 mm — research.md §5-6

LAM-PRF-18 waga 3/5

Optical quality of the lens

Crown-type optical glass, anti-reflection coated on both sides, low distortion; the LED ring does not appear as a bright ghost reflection in the lens.

Cel
≤ 0,50 %
Próg
≤ 1,30 %
Miara
residual reflectance per lens surface (450-650 nm average)
Oceniane
w pełnym projekcie
Jak sprawdzamy i źródło

sędziowie, plan testów — Lens and coating specification; distortion ≤ 2 % (target) / ≤ 5 % at 80 % aperture checked in the test plan.

Źródło: uncoated glass 4 % per surface; MgF2 ≤ 1.3 %; complaints about fringes and reflections — research.md §5-6

LAM-PRF-19 waga 2/5

Close-work lens

A second lens for close work that is swapped in or flipped in without tools; its mass is within the arm's balance range (LAM-UX-01/02).

Cel
≥ 8 D
Próg
≥ 5 D
Miara
optical power of the close-work lens
Oceniane
od koncepcji
Jak sprawdzamy i źródło

automatycznie, sędziowie — Lens part (datasheet or drawing) and its mount in CAD.

Źródło: TEVISIO 8 D add-on, Luxo 4/6/10 D add-ons — research.md §2, §6

Wymiary (2)

LAM-SIZ-01 waga 3/5

Reach of the arm

Horizontal reach from the mounting axis to the lens centre with the arm extended; the balance holds over a vertical range of at least 600 mm.

Cel
≥ 1 000 mm
Próg
≥ 850 mm
Miara
maximum horizontal reach
Oceniane
od koncepcji
Jak sprawdzamy i źródło

automatycznie — Measured in 03-cad/assembly.step in the extended pose (deliver that pose or the joint angles that give it).

Źródło: TEVISIO 784 mm, Luxo 762-1143 mm, Daylight 1080-1180 mm — research.md §6

LAM-SIZ-02 waga 2/5

Head size

Longest outer dimension of the lamp head (lens, light rings and housing; arm excluded).

Cel
≤ 260 mm
Próg
≤ 320 mm
Miara
longest dimension of the head
Oceniane
od koncepcji
Jak sprawdzamy i źródło

automatycznie — Bounding box of 03-cad/head.step (file axes).

Źródło: assumption (W3) - a Ø160 lens plus a light ring and housing

Masa (2)

LAM-WGT-01 waga 3/5

Head mass

Mass of the head (lens, light rings, housing, controls) without the add-on lens or camera.

Cel
≤ 1 kg
Próg
≤ 1,40 kg
Miara
head mass
Oceniane
od koncepcji
Jak sprawdzamy i źródło

automatycznie, obliczenie — Mass of 03-cad/head.step with material densities plus purchased-part masses.

Źródło: 150 mm 3.5 D glass lens ≈ 0.5 kg (CALCULATION); research.md §6

LAM-WGT-02 waga 2/5

Total mass

Mass of the complete lamp as defined in the reference conditions (clamp and any low-angle light accessory included; power adapter and cables excluded).

Cel
≤ 3 kg
Próg
≤ 3,80 kg
Miara
total mass with clamp
Oceniane
od koncepcji
Jak sprawdzamy i źródło

automatycznie, obliczenie — Mass of 03-cad/assembly.step with material densities plus purchased-part masses.

Źródło: TEVISIO 3.4-3.6 kg, Daylight 2.9-3.3 kg — research.md §2

Zasilanie (3)

LAM-PWR-01 waga 3/5

Low-voltage supply

The lamp runs on a low-voltage (SELV) input - USB-C Power Delivery (target) or a bundled external 24 V supply - with no mains voltage inside the lamp; on a plain 5 V USB-C source it works at reduced brightness.

Cel
USB-C PD sink (9/15/20 V), full function from a 45 W adapter, reduced mode on 5 V / 3 A
Próg
external SELV adapter (≤ 60 V DC, Class 2), no mains inside the lamp
Miara
supply input
Oceniane
od koncepcji
Jak sprawdzamy i źródło

sędziowie — Power input circuit and power budget reviewed.

Źródło: USB PD SPR 15-60 W; USB-only lamps sit under GPSR rather than LVD — research.md §6

LAM-PWR-02 waga 2/5

Maximum input power

Input power with all white channels at maximum (other modes may not exceed it).

Cel
≤ 25 W
Próg
≤ 35 W
Miara
input power at maximum white output
Oceniane
od koncepcji
Jak sprawdzamy i źródło

obliczenie, plan testów — Power budget from LED currents, driver efficiency and electronics; measurement.

Źródło: market 5-16 W; Ecodesign Ponmax ≈ 12 W at 800 lm CRI 95 directional (CALCULATION, research.md §6)

LAM-PWR-03 waga 1/5

Standby power

Input power when the light is off but the lamp listens to its controls (and radio, if any).

Cel
≤ 0,30 W
Próg
≤ 0,50 W
Miara
standby power
Oceniane
w pełnym projekcie
Jak sprawdzamy i źródło

obliczenie, plan testów — Standby power budget; measurement.

Źródło: Regulation (EU) 2019/2020 standby and networked standby 0.5 W — research.md §6

Funkcje (6)

LAM-FUN-01 waga 3/5

Controls on the head

Brightness, CCT and lighting mode are adjusted at the head (or the arm end) with controls that need no force that moves the head (rotary with low torque, or touch); the current CCT and level are visible.

Cel
tak
Próg
tak
Miara
control concept with no head movement when operated
Oceniane
od koncepcji
Jak sprawdzamy i źródło

sędziowie — Control placement, type and operating force in CAD/BOM.

Źródło: pressing buttons on the head makes it droop (user reports) — research.md §5

LAM-FUN-02 waga 2/5

Presets

User presets store CCT, level and contrast mode (sectors, tiers, colours); the lamp powers up in the last or a chosen preset.

Cel
≥ 6 presets
Próg
≥ 4 presets
Miara
number of user presets
Oceniane
od koncepcji
Jak sprawdzamy i źródło

sędziowie — Firmware and UI description.

Źródło: no presets is a common complaint — research.md §5

LAM-FUN-03 waga 2/5

Hands-free control

A foot switch or remote input lets the user change presets or toggle a contrast mode without letting go of the tools.

Cel
foot-switch jack plus a wireless or USB control option
Próg
foot-switch jack
Miara
hands-free input
Oceniane
od koncepcji
Jak sprawdzamy i źródło

sędziowie — Connector and firmware function reviewed.

Źródło: hands-free control wanted (research.md §5)

LAM-FUN-04 waga 2/5

Computer control and camera provision

A documented USB control interface sets every light channel and scene (for documentation and photometric-stereo capture); the head has a mechanical, power and data provision for an optional documentation camera.

Cel
documented USB API with scene sync output; camera module (≥ 1080p30 UVC) designed
Próg
camera mount, power and data provision; channels controllable over USB
Miara
USB control interface and camera provision
Oceniane
w pełnym projekcie
Jak sprawdzamy i źródło

sędziowie — Interface documentation, connectors and CAD provision reviewed.

Źródło: users document their work; Aven Mighty Vue has a camera — research.md §2, §5

LAM-FUN-05 waga 2/5

No app needed

Every function works from the lamp's own controls; an app, if offered, is optional and needs no account or internet.

Cel
tak
Próg
tak
Miara
functions available offline without an app
Oceniane
od koncepcji
Jak sprawdzamy i źródło

sędziowie — UI and firmware description.

Źródło: assumption (W3); RED cybersecurity applies to internet-connected radios — research.md §6

LAM-FUN-06 waga 2/5

Mounting

Desk-edge clamp for tables of this thickness range, plus a 13 mm pin / grommet socket option; the clamp holds at least 1.5x (target 2x) the overturning moment at full reach without slipping or denting the desk.

Cel
8-80 mm and pin socket, 2x moment
Próg
10-60 mm, 1.5x moment
Miara
clamp table-thickness range
Oceniane
od koncepcji
Jak sprawdzamy i źródło

obliczenie, sędziowie — Clamp geometry in CAD; overturning-moment calculation and pad pressure.

Źródło: market clamps 0-75 mm; 12-24 N·m overturning moment at 1 m reach (CALCULATION, research.md §6)

Obsługa (5)

LAM-UX-01 waga 5/5

Arm holds position

The head stays where it is put - in any pose over the reach range, with or without the add-on lens - and starts moving at a light push of 3-10 N at the lens rim.

Cel
≤ 3 mm
Próg
≤ 10 mm
Miara
head drift in 1 hour at full reach, worst pose
Oceniane
od koncepcji
Jak sprawdzamy i źródło

obliczenie, plan testów — Spring/balance calculation over the poses (residual moment); drift test with the add-on lens on and off.

Źródło: drooping arms are the most repeated complaint (research.md §5)

LAM-UX-02 waga 3/5

Balance quality

Residual (unbalanced) moment of each arm joint across at least 5 poses per degree of freedom, for head masses from nominal to nominal + add-on lens, as a share of the gravity moment.

Cel
≤ 10 %
Próg
≤ 20 %
Miara
maximum residual moment / gravity moment
Oceniane
w pełnym projekcie
Jak sprawdzamy i źródło

obliczenie, automatycznie — Moment-balance calculation using CAD masses and spring parameters, re-run by the checker.

Źródło: zero-free-length spring theory k = m·g·r/(b·c) — research.md §6

LAM-UX-03 waga 2/5

Head articulation

The head tilts from horizontal to vertical and beyond, swivels, and the arm rotates on its mount, with stops so no cable twists more than 360°.

Cel
≥ 180 deg
Próg
≥ 135 deg
Miara
head tilt range
Oceniane
od koncepcji
Jak sprawdzamy i źródło

automatycznie, sędziowie — Joint ranges in CAD; swivel ≥ ±170° (target) / ±90° (threshold); base rotation 360°.

Źródło: TEVISIO head turns up to 90° in every direction — research.md §6

LAM-UX-04 waga 3/5

No direct glare

No bare LED emitter is visible from the eye zone (300-600 mm above the work, within ±30° of the lens axis); visible luminous surfaces have low luminance.

Cel
≤ 20 kcd/m²
Próg
≤ 500 kcd/m²
Miara
maximum luminance of luminous surfaces visible from the eye zone
Oceniane
od koncepcji
Jak sprawdzamy i źródło

obliczenie, sędziowie — Sight-line check in CAD from the eye zone; luminance estimate of the diffusers.

Źródło: EN 12464-1 shielding table (20 / 50 / 500 kcd/m²); bare 3030 LED ≈ 1,200 kcd/m² (CALCULATION, research.md §3)

LAM-UX-05 waga 1/5

Silent operation

No fan and no audible coil or driver noise at any dimming level.

Cel
≤ 20 dB(A)
Próg
≤ 25 dB(A)
Miara
sound pressure at 0.5 m from the head, any setting
Oceniane
w pełnym projekcie
Jak sprawdzamy i źródło

sędziowie, plan testów — Cooling concept and driver switching frequencies reviewed; measurement in a quiet room.

Źródło: assumption (W3)

Bezpieczeństwo (5)

LAM-SAF-01 waga 4/5

Photobiological safety

Every channel and every preset (white at all CCTs, colours, UV if present) is in the exempt or low-risk group at 200 mm per EN 62471 / IEC TR 62778, without a hazard-distance marking.

Cel
RG0 (exempt) for all channels
Próg
RG1 without a threshold-illuminance limitation
Miara
worst risk group at 200 mm over all channels and presets
Oceniane
od koncepcji
Jak sprawdzamy i źródło

obliczenie, plan testów — Radiance/irradiance estimate per channel vs EN 62471 limits; measurement per IEC TR 62778 in the test plan.

Źródło: EN 60598-1 4.24.2 and IEC TR 62778 (assessment at 200 mm) — research.md §3, §6

LAM-SAF-02 waga 3/5

Touch temperatures

At 25 °C ambient and maximum output, surfaces touched to adjust the lamp (knobs, joints, head handle, lens rim) stay at or below this temperature.

Cel
≤ 40 °C
Próg
≤ 60 °C
Miara
maximum temperature of surfaces touched in use
Oceniane
od koncepcji
Jak sprawdzamy i źródło

obliczenie, plan testów — Thermal estimate of the head; thermocouple test in the test plan.

Źródło: EN 60598-1 Table 12.1 (means of adjustment 60 °C metal / 75 °C non-metal) — research.md §6

LAM-SAF-03 waga 2/5

Burning-glass protection

The lens cannot focus sunlight onto the bench or the lamp when parked - an integrated cover or shutter, plus a warning.

Cel
integrated cover or shutter, closed in the parked position
Próg
supplied cover and a warning on the product and in the manual
Miara
lens protection
Oceniane
od koncepcji
Jak sprawdzamy i źródło

sędziowie — Cover design in CAD; warnings in the label and manual drafts.

Źródło: magnifier lamp manuals and fire reports (burning-glass effect) — research.md §6

LAM-SAF-04 waga 2/5

Mechanical safety

No pinch points in the arm joints and springs accessible to fingers; the head cannot fall if a spring breaks; cords pass the EN 60598-1 anchorage and flexing tests.

Cel
tak
Próg
tak
Miara
mechanical hazards addressed
Oceniane
w pełnym projekcie
Jak sprawdzamy i źródło

sędziowie, plan testów — Arm design (spring covers, captive springs, stops) reviewed; anchorage and flexing tests in the test plan.

Źródło: EN 60598-1 cord anchorage (25 pulls at 60 N) and 4.14.3 — research.md §6

LAM-SAF-05 waga 2/5

ESD-safe for electronics work

Surfaces near the work (head, lens rim, arm end) are static-dissipative and connected to a ground point, so the lamp can be used inside an ESD protected area.

Cel
1e5-1e9 Ω to a ground stud on all such surfaces, per IEC 61340-5-1 measurement methods
Próg
metal parts bonded to a ground stud (Rg < 1e9 Ω); no ESD claim without measurement
Miara
resistance to ground of surfaces within 30 cm of the work
Oceniane
w pełnym projekcie
Jak sprawdzamy i źródło

sędziowie, plan testów — Materials and ground path reviewed; resistance measurement in the test plan.

Źródło: IEC 61340-5-1:2024; Luxo/Waldmann sell ESD variants at +500-1,000 PLN — research.md §2, §6

Certyfikacja (3)

LAM-CRT-01 waga 5/5

EU conformity (CE)

A complete EU conformity plan for the product as designed (its power supply, any radio, any bundled adapter) - the legal route, every applicable EU directive and regulation and the harmonised standards with their current editions; for each, the design measure, the test and its cost; a draft EU Declaration of Conformity.

Cel
complete plan with costs and timeline, ready to send to a test lab
Próg
all applicable laws and main standards identified with design measures
Miara
completeness and correctness of the EU conformity plan
Oceniane
od koncepcji
Jak sprawdzamy i źródło

sędziowie, automatycznie — Compliance document reviewed; presence of the applicable law and standard ids checked automatically.

Źródło: research.md §6 (EN 60598-2-4 is the part 2 for portable luminaires)

LAM-CRT-02 waga 3/5

US conformity

A complete US plan - FCC authorisation for the electronics and any radio, a safety listing by a nationally recognised testing laboratory to the applicable standards (lamp, LED parts, power adapter) and the US efficiency rules for the adapter, with the tests, cost and timeline.

Cel
complete plan incl. test list, cost and timeline
Próg
FCC route and NRTL standard identified with design measures
Miara
completeness and correctness of the US plan
Oceniane
od koncepcji
Jak sprawdzamy i źródło

sędziowie, automatycznie — Compliance document reviewed; presence of the FCC rule parts and UL standards checked.

Źródło: UL 153 Ed. 13 (2025) covers USB-powered portable luminaires — research.md §6

LAM-CRT-03 waga 4/5

Truthful specification sheet

A one-page customer specification sheet in which every number is backed by a calculation or a planned test and is labelled; magnification given in dioptres with the convention named; illuminance with its distance; CRI, R9, Duv and flicker as specified here; no feature the design does not deliver.

Cel
every claim traced to a calculation or a planned test
Próg
every claim traced to a calculation or a planned test
Miara
claims on the spec sheet traceable to evidence
Oceniane
od koncepcji
Jak sprawdzamy i źródło

sędziowie, automatycznie — Each spec-sheet number cross-checked against calculations and the test plan.

Źródło: owner brief („zero ściemy”); inflated '10X' magnification and lux claims on the market — research.md §5

Koszt (5)

LAM-CST-01 waga 5/5

Unit cost at 1,000 units

Manufacturing cost per unit at a 1,000-unit batch, EXW - all parts incl. lens, LED boards, drivers, arm, clamp, power adapter, assembly, calibration, end-of-line test and packaging; one-off costs excluded (see LAM-CST-04).

Cel
≤ 95 EUR
Próg
≤ 140 EUR
Miara
unit cost at 1,000 units, excl. VAT
Oceniane
od koncepcji
Jak sprawdzamy i źródło

automatycznie, sędziowie — bom.csv quantities × unit prices at 1,000 must sum to the figure in cost.md; prices checked for realism.

Źródło: derived from the retail price target with ~40-50 % gross margin in a dealer channel (CALCULATION, research.md §7)

LAM-CST-02 waga 2/5

Unit cost at 100 units

Manufacturing cost per unit at a 100-unit batch (same scope; low-volume processes allowed).

Cel
≤ 160 EUR
Próg
≤ 230 EUR
Miara
unit cost at 100 units, excl. VAT
Oceniane
w pełnym projekcie
Jak sprawdzamy i źródło

automatycznie, sędziowie — bom.csv at 100 vs cost.md.

Źródło: assumption (W3)

LAM-CST-03 waga 2/5

Prototype cost

Cost of parts and outside machining for ONE working prototype (engineering labour excluded).

Cel
≤ 700 EUR
Próg
≤ 1 200 EUR
Miara
cost of one prototype, excl. VAT
Oceniane
w pełnym projekcie
Jak sprawdzamy i źródło

automatycznie, sędziowie — bom.csv at quantity 1 plus prototype processes in cost.md.

Źródło: assumption (W3)

LAM-CST-04 waga 3/5

One-off costs (tooling and certification)

Tooling (moulds, dies, fixtures), certification and photometric tests (EU and US) and other one-off costs before the first 1,000 units ship.

Cel
≤ 40 000 EUR
Próg
≤ 80 000 EUR
Miara
total one-off cost, excl. VAT
Oceniane
w pełnym projekcie
Jak sprawdzamy i źródło

sędziowie — Itemised one-off cost list in cost.md reviewed for completeness and realism.

Źródło: assumption (W3)

LAM-CST-05 waga 4/5

Retail price and unit economics

Recommended retail price in the EU including VAT, with a unit-economics table (price, VAT, channel or platform fees, shipping, unit cost, gross margin) for direct sales and for distributor sales.

Cel
≤ 349 EUR
Próg
≤ 499 EUR
Miara
recommended retail price incl. VAT
Oceniane
od koncepcji
Jak sprawdzamy i źródło

obliczenie, sędziowie — Unit-economics table re-computed; margin must be positive in both channels.

Źródło: Marcin: anything sensible costs 2-3 thousand PLN and has no options; PL ladder 100-350 / 1,100-2,150 / 2,650-4,700 PLN (research.md §2, §7)

Produkcja (3)

LAM-MFG-01 waga 4/5

Real, purchasable parts

Every BOM line has manufacturer, MPN, supplier, URL and unit prices at 1/100/1,000; purchased parts exist and are orderable (LEDs in the stated CCT, CRI and bin); custom parts (lens, housings, arm parts, springs) have a drawing, material, process and cost basis; critical parts (LEDs, lens, springs, LED drivers, USB-PD controller) have a second source or a stated single-source risk.

Cel
≥ 100 %
Próg
≥ 95 %
Miara
share of BOM lines with complete and valid sourcing data
Oceniane
w pełnym projekcie
Jak sprawdzamy i źródło

automatycznie, sędziowie — BOM checked line by line (fields present, part exists and is available); second sources reviewed.

Źródło: owner brief (BOM with real parts and prices)

LAM-MFG-02 waga 2/5

Assembly, calibration and test time

Labour time per unit at 1,000 units for final assembly, colour/flux calibration and end-of-line test.

Cel
≤ 25 min
Próg
≤ 40 min
Miara
labour minutes per unit
Oceniane
w pełnym projekcie
Jak sprawdzamy i źródło

sędziowie — Assembly and calibration sequence with time estimates per step.

Źródło: assumption (W3)

LAM-MFG-03 waga 3/5

Colour calibration and end-of-line test

Each unit is checked before packing - every channel and sector lights, CCT/Duv and flux at defined points (calibration data stored in the lamp), flicker, balance of the arm and the clamp - with pass limits.

Cel
per-unit spectral calibration stored in firmware plus full functional test
Próg
functional test plus CCT/Duv check at three points
Miara
end-of-line test and calibration defined with limits
Oceniane
w pełnym projekcie
Jak sprawdzamy i źródło

sędziowie — Production test plan reviewed.

Źródło: assumption (W3); LED bin spread drives colour mismatch (research.md §3)

Niezawodność (3)

LAM-REL-01 waga 2/5

Light output maintenance

Rated life to 90 % of initial flux (L90) of the LEDs at their operating temperature in the lamp, from LM-80 / TM-21 data.

Cel
≥ 25 000 h
Próg
≥ 15 000 h
Miara
L90 life at the in-lamp LED temperature
Oceniane
w pełnym projekcie
Jak sprawdzamy i źródło

obliczenie, sędziowie — LED case temperature estimate at maximum output and TM-21 projection from the LED maker's LM-80 data.

Źródło: LM-80/TM-21 example L90 ≈ 23,000 h at Ts 85 °C — research.md §3

LAM-REL-02 waga 3/5

Joint and cable endurance

Full-range movement cycles per joint without exceeding the drift requirement and without broken conductors.

Cel
≥ 10 000 cycles
Próg
≥ 1 500 cycles
Miara
endurance cycles per joint
Oceniane
w pełnym projekcie
Jak sprawdzamy i źródło

sędziowie, plan testów — Joint and cable design (bend radius, stops, materials) reviewed; endurance test in the test plan.

Źródło: EN 60598-1 flexing test 1,500 cycles for frequently adjusted luminaires — research.md §6

LAM-REL-03 waga 2/5

Colour stability

Shift of the white point after warm-up and over the life of the lamp at a fixed setting.

Cel
≤ 0,00 Δu'v'
Próg
≤ 0,00 Δu'v'
Miara
chromaticity shift from cold start to 30 min and over the rated life (Δu'v')
Oceniane
w pełnym projekcie
Jak sprawdzamy i źródło

obliczenie, plan testów — LED temperature drift data and compensation method reviewed; 30-minute warm-up measurement.

Źródło: assumption (W3); colour points of LEDs shift with junction temperature (research.md §3)

Serwis (3)

LAM-SRV-01 waga 3/5

Replaceable light sources and driver

LED boards and the driver can be replaced with common tools and connectors (no soldering) within this time; spare parts are listed.

Cel
≤ 5 min
Próg
≤ 15 min
Miara
time to replace an LED board or the driver
Oceniane
w pełnym projekcie
Jak sprawdzamy i źródło

sędziowie — CAD (access, connectors, fasteners) and service instructions reviewed.

Źródło: Regulation (EU) 2019/2020 Art. 4 (replaceable light sources and control gear) — research.md §6

LAM-SRV-02 waga 2/5

Lens swap and cleaning

The main lens and the add-on lens can be removed and refitted without tools for cleaning or exchange; the coating survives the stated cleaning method.

Cel
tak
Próg
tak
Miara
tool-free lens exchange
Oceniane
w pełnym projekcie
Jak sprawdzamy i źródło

sędziowie — Lens mount design in CAD and the cleaning instructions.

Źródło: TEVISIO bayonet lens change — research.md §2

LAM-SRV-03 waga 2/5

Firmware updates and adjustment

Firmware updates over USB; the arm balance can be re-adjusted by the user without tools after a lens or camera change.

Cel
tak
Próg
firmware update over USB; balance adjustable with a common tool
Miara
update path and user balance adjustment
Oceniane
w pełnym projekcie
Jak sprawdzamy i źródło

sędziowie — Firmware documentation and arm adjustment design reviewed.

Źródło: assumption (W3)

Założenia arkusza
Pytania do Pawła i Marcina

Lampa v1 — pytania do Marcina

Arkusz stoi na założeniach poniżej (v0: pierwsze przebiegi, v1: poprawki po recenzji, te same liczby). Twoje odpowiedzi trafią do wersji v2 (v0 i v1 zostają bez zmian).

1. Lupa

Założenie: szklana lupa ≥ 160 mm, 3–3,5 dioptrii (ok. 1,75–1,9×), ok. 25 cm wolnego miejsca pod głowicą na narzędzia, plus wymienna albo odchylana druga soczewka 8 dioptrii (3×) do drobnej roboty. Pasuje? Chcesz większe powiększenie, inny kształt (prostokątna) albo kamerę zamiast drugiej soczewki?

2. Oświetlenie kontrastowe — co najważniejsze?

Założenie: obowiązkowo (a) sektory wokół lupy (światło z wybranej strony, ≥ 8 sektorów) i (b) światło z niskiego kąta (≤ 20°, krawędzie i rysy). Jako cele dodatkowe: (c) warstwy RGB jak w pierścieniu Vision Datum (kolor = nachylenie lutu), (d) polaryzacja (bez odblasków z lutów), (e) UV 365 nm (lakier, topnik). Którą kolejność ważności przyjmujesz? Coś pominęliśmy?

3. Ramię i mocowanie

Założenie: zasięg ≥ 1 m, ramię trzyma pozycję bez opadania (≤ 3 mm na godzinę), zacisk do blatu 8–80 mm + gniazdo na trzpień 13 mm; bez ciężkiej podstawy. OK?

4. Cena i zasilanie

Założenie: ok. 349 € (ok. 1500 zł) brutto, koszt produkcji ≤ 95 € przy 1000 szt., zasilanie USB-C PD (zasilacz 45 W w zestawie), wersja bezpieczna dla ESD w standardzie. Zapłaciłbyś więcej za coś lepszego, czy cena ma być niższa?

5. Sterowanie i dokumentowanie

Założenie: pokrętło na głowicy (nie przesuwa lampy), ≥ 6 presetów, gniazdo przełącznika nożnego, sterowanie z komputera przez USB; aplikacja niepotrzebna; kamera jako opcjonalny moduł (1080p). Czego brakuje (np. pilot, aplikacja, wbudowana kamera z ekranem)?

Rozpoznanie konkurencji i źródła

Lampa v1 — research i konkurencja

Wersja v1 = v0 z poprawkami arkusza (lista w CHANGES.md); treść badań bez zmian.

Streszczenie (dla Marcina i Pawła)

Wszystkie liczby z etykietą pochodzenia; ceny widziane 30.10.2026. Szczegóły poniżej (po angielsku).


1. Scope and method

Six parallel web-research agents (2026-10-30, 03:22–03:32 Warsaw), one slice each: magnifier lamp market (L1), contrast/inspection lighting (L2), LEDs and photometry (L3), standards (L4), customer voice and alternatives (L5), mechanics and magnifier optics (L6). Every number in the raw data carries its URL; manufacturer claims are marked. Raw data (not in git): /srv/dd-data/w3/research/L*.json, checksums in /srv/dd-data/w3/research/SHA256SUMS. Known limits: Reddit and some shop pages were blocked (403/429) — customer complaints are a tally of distinct sources; the Vision Datum page returned 429 to one agent but was read by another (L2).

2. Market: magnifier and task lamps

product lens dioptre / M lens material illuminance CCT CRI dimming reach mounting power price (seen 2026-10) source
Waldmann TEVISIO (TVD 750/940/DM, EU) – 384 mm / 484 mm arm… lens 160 mm (Waldmann shop) / 153 mm (kaiserkra… 3.5 dpt (~1.88-2x); optional 8 dpt add-on (3.5+… plastic lens (cheapest variants), glass, or ant… 6000 lx max (distance not stated); 750 lm 4000 K fixed (EU /940); 5000 K in the US (TVD 7… Ra>=90, R9>=50 stepless dimming 100-10%; segment switching ('v… arm 384/484 mm, total reach 784 mm (kaiserkraft… table clamp, pin mount 13-14 W, 100-240 V; ESD version 16 W Waldmann DE shop, net: EUR 566 (plastic lens, 2D) / 749-781 (glass, 3… link
Waldmann MLD (MLD 750/850/D) lens 160 mm 3.5 dpt plastic lens lux not found; 800 lm 5000 K fixed Ra>80 dimmable 31 in double arm (US) table clamp 12 W, 100-240 V A1-ESD (DE): EUR 375.54 / 402.11 net; US: 481.50-587 USD link
Waldmann RING LED (RLLQ 63) lens about 2.8 in (~71 mm) 5.3 or 6 dpt (sources differ) material not found lux not found CCT not found CRI not found dimming not found 30 in reach clamp, universal base (+202 USD) or wall bracket 6 W US: 676.39-800 USD (Lighting Specialties 690.30 USD) link
Glamox Luxo (Vision-Luxo) WAVE LED / WAVE+ LED / WAVE LED E… lens rectangular 6.75 x 4.5 in (~171 x 114 mm) 3.5 dpt (1.88x) or 5 dpt (2.25x); clip-on secon… white crown optical glass 4600 lx at 11 in (279 mm) focal length 4000 K fixed CRI 80 2 x 6 W modules switched separately; step dimmi… internal-spring K-arm, 30 in or 45 in; Farnell … edge clamp or weighted base ESD versions available Farnell PL, net: 2,729.92 PLN (ESD 3.5 dpt), 2,878.04 PLN (ESD 5 dpt)… link
Glamox Luxo KFM LED / KFM LED-ESD lens 5 in / 127 mm 3 dpt (1.75x) or 5 dpt (2.25x); STAYS add-ons 4… crown glass lux conflicts by source: 5700 lx at 13 in (US r… 4000 K CRI 80 infinite dimming 1-100%; 9/4 h auto-off 30 in or 45 in arm clamp, base, floor caster ESD version not found in PLN; US resellers (prices not captured) link
Glamox Luxo LFM LED / LFM LED G2 lens 127 mm 3 or 5 dpt glass 3000 lx at 13 in (brochure, attributed to G2-li… 4000 K CRI 80 3-step dimmer 100/50/0% (LFM LED); G2: 9 W dimm… arm up to 1050 mm (RS PL), 30/45 in table clamp 7 W module / 9.5 W total, 24 V PSU, IP20 RS PL: 1,243.91 PLN net / 1,530.01 PLN gross (5 dpt LFG028261), 1,247… link
Glamox Luxo Circus LED (CIL026698 / CIL026699) lens 6.5 in / 165 mm round 3.5 dpt (1.88x) or 5 dpt (2.25x); STAYS add-ons white crown glass 3830 lx at 11 in focal length 3000 K or 4000 K (separate SKUs, not switchable) CRI 90 dimmable 10-100% 45 in steel parallel arm, 3 pivots edge clamp 10 W not found (gotopac etc. pages not captured) link
Daylight Company Omega 7 (E25130/D25130/U25130) lens 175 mm (7 in) 3 dpt (1.75x); optional 5 in 5 dpt and 12 dpt s… acrylic, anti-scratch coated (newer) 6300 lx at 15 cm; 814 lm 6000 K (older); newer version claimed by UK/US … CRI 95+ (one retailer, newer model) 4 steps 6300/5500/3500/1800 lx (older listings:… max reach 118 cm clamp 12-12.5 W Amazon.pl 1,343.30 PLN; maluje.pl 1,800 PLN gross; sklepplastyczny.pl… link
Daylight Company MAG Lamp XL (UN1300 / DN1300 / EN1300) and… XL: lens 177 mm (7 in) 3 dpt (1.75x) glass 5300 lx at 15 cm (3000 lx second step); 770 lm current official page: 3 presets 6000/4000/2700… CRI 95+ (one table says 80+) 2 brightness levels max reach 108 cm clamp 11.5 W XL: official US 239.99 USD; Bürklin about EUR 100.01; Heamar UK GBP 6… link
Daylight Company Omega 5 / Omega 5 ESD lens 127 mm 3 dpt and 5 dpt interchangeable glass 5500 lx at 15 cm; 548 lm 6000 K (6500 K elsewhere) CRI 95+ (ESD version: CRI 80+) 4-step touch dimmer reach not found clamp 11 W (ESD 8 W) not found link
Eschenbach varioLED flex (2781; XL 27815) and VarioLED+ flex: lens 76 mm 6 dpt (2.5x) material not stated about 6000 lx at 150 mm 9000 K (one dealer only) CRI not found ring/segment switchable (63 SMD LEDs) gooseneck 350 mm (XL 600 mm) base/clamp 5 W varioLED flex: geizhals.de from EUR 347.64 (2025); Hennes Optik EUR 4… link
Aven Tools Mighty Vue Inspector 26510-CAM (3 dpt), 26512-CA… lens 7.5 x 6.2 in rectangular (blog: 7.5 x 6.5 … 3 dpt (1.75x) or 5 dpt (2.25x) glass 1100 lm max; lux not found tunable 3500-6500 K from 30 white + 30 amber LE… CRI 95 (Zoro) brightness and CCT adjustable, IR remote 36 in reach table clamp 100-240 V list 577.17 USD (RS US, Newark, TestEquity); TEquipment 461.74 USD; M… link
Aven Tools ProVue SuperSlim LED 26501-LED (5 dpt; 8 dpt ver… lens 5 in (127 mm) 5 dpt (2.25x); 8 dpt (3x) version glass, retractable cover 7410 lx (Amazon, distance not stated); 560 lm 6500 K CRI not found dimming not found 36 in spring-balanced arm heavy-duty clamp or rolling stand 6 W, 20 V adapter aventools.com 160.73 USD (clamp); 8 dpt 206.58 USD; rolling stand 231… link
Brightech LightView PRO (desk clamp 1.75x/2.25x, floor vers… lens 4.5 in wide (desk); floor up to 6 x 4.5 in 3 dpt (1.75x) or 5 dpt (2.25x) glass lux not found; 650-900 lm 6000 K (6500 K on one listing) CRI not found some dimmable gooseneck/arm, height 4-33 in clamp (opens to 2.5 in) or floor stand 9 W not captured (Amazon.com) link
Lightcraft LC8064LED Classic; LC8093LED table; LC8072LED Pr… LC8064LED: 125 mm lens, 3 dpt (1.75x), 90 LEDs,… LC8064LED: CPC Farnell GBP 87; Newark 161.92 USD; EU EUR 116-146; She… link
Generic Chinese (Amazon/Allegro) fulingkeji B093SWNYTV; Xia… lens about 4.1-4.2 in (~105 mm) '5 dpt', marketed as 5X/10X glass (claimed) lux not found; 1500 lm claim (fulingkeji 10 W) 3 presets 3000/4000-4500/6000-6500 K CRI not found stepless or 10-step dimming, memory 13 x 13 in metal arm clamp + base USB or 10-12 W LEDarena.pl 104.99 PLN gross (12 W, 3000/4000/6500 K, 720 lm); Amazon… link
Generic PL electronics shops (Gotronik, Inelsc, Szafir, TME… Gotronik: NAR0460 90 mm 48 LED; LAM-042 127 mm … Gotronik 146.63 / 235 / 299 / 325 PLN; Inelsc 212-225 PLN; Szafir 350… link
Hakko (Singapore distributor) / O.C. White / View Solutions… Hakko: 6 in glass, 5 dpt, dimmer, 220-240 V, cl… O.C. White PLREC-45-B 419 USD; others not found link

Price ladder in Poland (SOURCE: L1, L5, L6; gross prices unless marked):

tier price what you get
generic 100-350 PLN 90-178 mm lens, mostly 5 D, 6500 K fixed or 3 presets, no CRI/lux data, weak arm
branded mid 850-2,150 PLN Luxo KFM/LFM/Wave (127 mm or 171 × 114 mm glass, 4000 K, CRI 80), Daylight Omega 7 (175 mm, 3 presets, CRI 95+ claim)
premium / ESD 2,650-4,700 PLN Luxo Wave ESD, Waldmann TEVISIO (160 mm AR glass, Ra ≥ 90, R9 ≥ 50, 4 switchable segments, 3.4-3.6 kg, 784 mm reach)
alternatives 1,550-30,000 PLN digital microscopes (Andonstar AD409 1,550-1,850 PLN), stereo microscopes, Vision Engineering Mantis (2,369 EUR+)

Observations (SOURCE unless marked):

3. LEDs and photometry

4. Contrast and inspection lighting

product key specs price (seen 2026-10) source
Vision Datum Multi-layer RGB Ring Illuminator for PCB AOI (VT-VT2-AOI… OD/ID/H: 73/8/28.5 mm; 100/7/48 mm; 123/33/45 mm; 198/42/68.8 mm; 250/84/81.2 mm / no magnifier / n/a / illuminance not found / RGB (250 mm model RGBW) / CRI n/a / each … from 609.00 USD (AOI73, without controller), shop.visiondatum.com, seen 2026-10… link
AmScope LED-144A (144-LED four-zone ring) / LED-72AM (72-LED four-zon… LED-144A: ID 64 mm, OD 100 mm, fits mounts up to 61 mm (sources disagree: some say 64 mm), 48 mm adapter included / 144 LEDs in 4 zones, each zone switched on/off separa… LED-144A 142.99 USD at amscope.com; 149.95 USD Volition; 173.99 USD Zoro; 263.9… link
Delta Optical LED-80T (DO-4720) 80 LEDs in 4 zones of 20 / fits objectives up to Ø60 mm / control panel built into the ring 479.00 PLN, deltaoptical.pl link
MAGUS LED 144-S4 144 LEDs, 4 sectors, adjustable brightness, aluminium body, ID 60-70 mm, remote control: a button selects the sector, a knob sets brightness; one or two sectors simulate… 699.95 PLN, pl.magusmicro.com link
Optek (Slezak-Bis) LED 144 zone ring YK-B144T / LED-60T / Multi illum… zoned 144-LED and 60-LED rings; MTL-3 is a 4-way multi illuminator; detailed specs not read YK-B144T 675.00 PLN gross; LED-60T 470.00 PLN gross; MTL-3 1,350.00 PLN gross (… link
Parco Scientific IFR-05 LED Quadrant Ring Light ID 64 mm, OD 106 mm / 64 white LEDs, 5,500 K / 0-100 % intensity / each quadrant on/off / WD 30-160 mm 97.00 USD, parcoscientific.com link
Vision Engineering S-414C LED Quadrant Ring-Light quadrant control; detailed specs not read 571.88 USD, diverseelectronics.com link
MBJ Imaging 4 Segment SRL-12 / RGBW SRL-12 / Angled Ringlight ARL-10 SRL-12: luminous field Di 87 / Da 121 mm, camera opening 80 mm, 4 individually controllable segments, made for 'shape from shading' elevation images, white/red/IR, M12 8… SRL-12 from 410 EUR; RGBW SRL-12 from 480 EUR; ARL-10 from 440 EUR (all net, B2… link
Smart Vision Lights RM140-4Z (KIT) / RM75-4Z mini low-angle ring ligh… dark field / radial / 4 zones, each independently controlled via the 4ZMD-250 driver, 10-100 % intensity, OverDrive strobe / WD 50-200 mm / IP40 not found (RFQ only) link
CCS LDR2-LA series low-angle ring lights (LDR2-48/74/132/170 RD2-LA, … LED mounting angle 30° (Edmund) / OD/ID/H: 48/22/18.5 mm, 2.1 W, 50 g; 74/48/19 mm, 4.6 W, 90 g; 132/96/22 mm, 13 W, 270 g; 170/134/22 mm, 18 W, 350 g / red 630 nm, also… not found (Edmund page did not load) link
CCS HPR2-DV segmented ring lights very diffuse output, 4 independently controllable quadrants, 7 sizes x 3 wavelengths not found link
Effilux Low-profile segmented square LED ring light / EFFI-SRing 4 separately controlled quadrants, AutoStrobe, mono or multi-colour, convertible to a dome base light not found; 4-channel switch 216.00 USD, flying-lead cable 110.00 USD… link
Omron VT-M12 MDMC (multi-direction multi-colour) AOI lighting 13 directions (coaxial + 3 dome elevation levels x 4 azimuth blocks) x RGB at 127 steps each = 39 channels; classic colour highlight: red from upper level, green middle,… not found link

5. Customer voice

SOURCE: L5 slice (EEVblog, elektroda.pl, hobby forums, reviews). Main themes:

  1. arms that droop, drift or wobble; clamps that slip; buttons on the head that push it down;
  2. wrong working distance — 5-8 D lenses put the focal point closer than a brush or soldering iron („don't go over 3 dioptres”; Polish users swapped 8 D lamps for 5 D);
  3. small lenses, distortion, red/blue fringes, reflections; inflated „10X” claims (8 D ≈ 3x);
  4. cold, harsh, point-like light; poor colour; glare on shiny solder (3,500-4,500 K preferred by some);
  5. flicker and banding on camera at low dimming;
  6. shadows from hands and tools; multiple shadows;
  7. heat in cheap heads (overheating driver transistors);
  8. controls in the base, no presets, no hands-free control;
  9. many owners end up using the lamp without the lens (OptiVisor 55-70 USD is the most recommended alternative); stereo microscopes „nothing beats” for SMD. Buyers: electronics engineers and repair techs, watchmakers, jewellers, miniature painters, model makers, QA. Price expectations: hobby 50-80 USD for a durable lamp, premium 150-300 USD; Polish market as in §2.

6. Optics, mechanics and standards

6.1 Optics (SOURCE: L1, L6 unless marked)

6.2 Mechanics (SOURCE: L6 unless marked)

6.3 Standards and regulations (SOURCE: L4)

7. Price, cost and unit economics

8. Gaps and uncertainties