Test 1 Can a transparent coating be detected at all?
A hydrocarbon-based acrylic coating applied to one third of the board, trained as two classes: coated and uncoated.
- Coated
- Uncoated
- Coated
- Uncoated
Material-level inspection for electronics manufacturing
Turn a visual judgement
into a documented measurement.
Identify coating presence and coating type in a single non-contact pass, including transparent coatings that look identical under normal light. No sample preparation, no destructive testing.
- Acrylic conformal coating
- Polyurethane-alkyd resin
- Clearcoat presence detected, two transparent coatings told apart
- HELIOS EC64 NIR line scan, 930–1700 nm, real-time classification
- Complements IPC-A-610 inspection, does not replace it
The challenge
The variation you cannot see
Conventional vision cannot see differences that are critical to quality. Variation in coatings, resins and solder alloys, or contamination, may be invisible, yet it can cause defects or reduce reliability.
Concealed coating defects
Thin or uneven coatings are invisible to the eye but change how the assembly survives moisture, dust and temperature.
Virtually identical materials
Different resins and polymers look the same under normal light, yet behave very differently over the product lifetime.
Hidden contamination
Flux residues, oils and particulates are hard to detect and can end up as resistance drift or field failures.
Two coatings. Same appearance.
Different protection.
One board, one measurement, two views. Drag to compare what an appearance-based check records with what the material signal shows.
- Left: the board as a conventional RGB camera records it.
- Right: the same board, same frame, with the hyperspectral layer on top.
- Recorded on a HELIOS EC64 in the Headwall proof of concept.
Both frames come from the same measurement, so every feature stays in register as you drag.
What hyperspectral imaging reveals
Coating type
Separates similar polymers, resins and conformal coatings by their spectral signature.
Material variation
Identifies subtle formulation or batch differences before they reach the field.
Process insight
Turns coating and material quality into a plain, recorded signal instead of an opinion.
Contamination detection
Flux residues, oils and particulates that stay invisible in a standard check.
Coating presence and coverage
Detects thin, uneven or missing coatings, including coatings with no UV tracer.
Proof point
Coating detection and coating differentiation
Two separate tests on PCB samples, each trained with two classes, carried out in a laboratory setup.
Test 2 Can two transparent coatings be told apart?
Acrylic conformal coating on the top half, polyurethane-alkyd resin on the bottom half, trained as one class each.
- Acrylic conformal
- Polyurethane-alkyd
- Acrylic conformal
- Polyurethane-alkyd
What was measured
Two tests on PCB samples. First, a hydrocarbon-based acrylic coating applied to one third of a board, trained as two classes: coated and uncoated. Second, an acrylic conformal coating on the top half and a polyurethane-alkyd resin coating on the bottom half, trained as one class each. All three coatings are transparent.
How
HELIOS EC64 real-time hyperspectral line-scan camera on an EVK LabStation: four halogen illumination units emitting across the NIR range of 930–1700 nm, a linear actuator moving the sample underneath the camera, and 256 wavelength pixels per spatial point. Training points were marked on the coated and uncoated areas, and the classification models were built in EVK SQALAR.
What it showed
Detecting a clearcoat is feasible, and two coatings can be told apart on a PCB. The classified images show slight misclassifications, mostly on areas that are lighter or darker, or on parts that return a strong spectral signal themselves.
Where this fits
Alongside the standards you already inspect against
Conformal coating is qualified under IPC-CC-830 and inspected against IPC-A-610 acceptance criteria, usually visually or under UV. Hyperspectral imaging adds what those checks cannot confirm: which material is actually there, including coatings that carry no UV tracer.
- IPC-CC-830
- IPC-A-610 rev. J
- J-STD-001
Applications
Across the electronics value chain
One application is validated in our lab. The others are plausible and start with a feasibility check.
Conformal coating inspection
Verify presence, coverage and coating type.
Validated in our labAdhesives and sealants
Detect voids, inclusions or the wrong material.
Feasibility on requestDielectric and surface analysis
Identify dielectric materials and surface compounds.
Feasibility on requestPolymers and material ID
Differentiate resins and polymers for consistency.
Feasibility on requestFlux and contamination
Find residues, oils, particulates and foreign material.
Feasibility on requestProcess monitoring
Track variation over time to improve line quality.
Feasibility on requestRelated resources
PCB Coating Feasibility Study
Setup, spectra and classified results for two transparent coating tests on a PCB, including where the method reaches its limits.
Get the feasibility studyWhen PCB Coating Inspection Misses What Matters
In electronics manufacturing, some of the most consequential coating problems are also the hardest to see.
Read blog postHELIOS EC64
The NIR line-scan camera used in the measurement: 930–1700 nm, 256 wavelength pixels per spatial point, real-time classification.
Download the data sheetFor your own assessment
Get the Feasibility Study
PCB coating inspection with hyperspectral imaging: setup, spectra, classified results, and the conditions under which detection reaches its limits.
- Two tests, measured and classified
- Setup, illumination and camera configuration
- Where the method reaches its limits
PCB Coating Feasibility Study
PDF · 9 pages · HELIOS EC64 laboratory setup, two tests with two classes each.