Application Note | PCB Inspection
Automated Large Area Profilometry of Printed Circuit Boards
Fast, Macro-Driven 3D Inspection of Six PCBs: Step Heights, Dimensions, Intensity and Warpage
Research & Experimental Testing
Frank Liu
Visual Design & Editorial
Andrew Shore
Introduction
Scaling up of manufacturing processes is necessary for industries to grow and keep up with constantly increasing demands. As manufacturing process scales up, the tools used in quality control also need to be scaled up. These tools must be fast to keep up with the production rate, while still maintaining high accuracy to meet product tolerance limits. Here, the NANOVEA HS2000 Profilometer, with Line Sensor, showcases its value as a quality control instrument with its fast, automated, and high-resolution large area profilometry capabilities.
Importance of 3D Non-Contact Profilometry for Quality Control
The technology behind the topography is non-contact axial chromatism. This technology is a non-destructive technique that is able to measure the surface of all types of samples. Axial chromatism associates wavelength with a specific height, making reflectivity, transparency, and curvature have no influence on accuracy.
With the NANOVEA HS2000 Line Sensor, multiple samples can be repeatedly scanned by configuring a simple macro. The use of a macro allows users to easily use the instrument, minimizing constant instrument handling and human error. With scanning speeds up to 1m/s and 192 points per pass, the NANOVEA HS2000 is able to quickly obtain surface data on a large number of samples.
ℹ️ Learn more about surface profiling lab services.
Measurement Objective
In this study, the NANOVEA HS2000 Line Sensor showcases its fast, automated abilities by scanning six different printed circuit boards (PCB). Each PCB had differing heights and dimensions, but the full surface of each board was captured by setting an 18-scan macro.
NANOVEA ST500 Large Area
Optical 3D Profilometer
Discussion
In the results below, the false color view and 3D view for each PCB are presented. Additional analysis examples were also included for some results to showcase the capabilities of our analysis software.
In the scan conducted, the maximum height range of our pen was limited to 4mm. Large height features had to be captured by scanning the same PCB multiple times at varying heights and patching the data together. The total time for the macro to complete was approximately 40 minutes, averaging slightly over 2 minutes per scan. If the samples were flat, it would only take 12 minutes to scan the same area since patching would not be needed.
Results
PCB 1
| Plane parameter (mm) | Plane 1 | Plane 2 | Plane 3 | Plane 4 | Plane 5 |
|---|---|---|---|---|---|
| Zmin | 0.2989 | 5.534 | 6.610 | 11.53 | 13.99 |
| ZMean | 2.126 | 6.326 | 6.976 | 12.50 | 15.16 |
| Zmax | 2.921 | 6.764 | 7.225 | 13.07 | 15.37 |
Table 1: Step height analysis of PCB 1: component planes.
PCB 2
PCB 3
| Parameter | Value | Tolerance | Result |
|---|---|---|---|
| Distance, dimension 1 | 14.30 mm | 13.00 mm ± 1.00 mm | Out of tolerance |
| Distance, dimension 2 | 13.00 mm | 13.00 mm ± 1.00 mm | Within |
| Distance, dimension 3 | 101.00 mm | 100.00 mm ± 2.00 mm | Within |
Table 2: Dimensional tolerance check on PCB 3.
PCB 4
PCB 5
PCB 6
| ISO 12781 flatness parameter | Value |
|---|---|
| FLTt, peak-to-valley flatness deviation | 601.5 µm |
| FLTp, peak-to-reference flatness deviation | 336.6 µm |
| FLTv, reference-to-valley flatness deviation | 264.9 µm |
| FLTq, root-mean-square flatness deviation | 149.1 µm |
Table 3: Flatness parameters of PCB 6 (warpage study).
Conclusion
The NANOVEA HS2000 Line Sensor conducted profilometry scans on six different PCBs. Each PCB had unique features that were all captured with a single macro. This macro covered a large area with a high resolution of 25um x 25um step size in 40 minutes, obtaining the full surface of each PCB board. The false color view and 3D view were shown for each PCB board, with the addition of automatic step height analysis, contour analysis with dimensional tolerance limits, intensity views, and warpage study. These are just a few examples of the capabilities of our analysis software.
From this study, the automation of high resolution, large area profilometry scans has shown to be beneficial for quality control applications. It meets the requirement of speed and accuracy that quality control tools need to fulfill. Its automation abilities also minimize user handling, making the instrument easier to use and decreasing time required for setup. With its ability to scan all types of samples, the NANOVEA HS2000 Line Sensor can be used in all industrial fields as well.
Frequently Asked Questions About Automated PCB Profilometry
Can a profilometer inspect several PCBs automatically?
Yes. A simple macro scans multiple samples repeatedly, minimizing instrument handling and human error. Here one 18-scan macro captured the full surface of six different boards.
How long did the scan take?
About 40 minutes for all six boards, slightly over 2 minutes per scan, at a 25 µm x 25 µm step. Because tall components exceeded the 4 mm height range of the pen, boards were scanned at several heights and the data patched; flat samples would take 12 minutes.
Why non-contact axial chromatism for PCBs?
It is non-destructive and measures all types of surfaces: each wavelength is associated with a specific height, so reflectivity, transparency and curvature have no influence on accuracy, which suits the mix of metal, solder, plastic and laminate on a board.
What analyses can be run on the scans?
Besides false color and 3D views: automatic step height analysis of component planes, contour analysis with dimensional tolerance limits, intensity views of the surface, and warpage studies with ISO 12781 flatness parameters.
How flat was the board in the warpage study?
PCB 6 gave a total flatness deviation FLTt of 601.5 µm (peak 336.6 µm, valley 264.9 µm, RMS 149.1 µm) under ISO 12781.
Can it flag out-of-tolerance dimensions?
Yes. In the contour analysis of PCB 3, one dimension measured 14.30 mm against a 13.00 ± 1.00 mm tolerance and was flagged, while the other two were within their limits.
