Infrared inspection benefits from image subtraction
Steve Scheiber, Contributing Technical Editor -- Test & Measurement World, 4/1/2007 2:00:00 AM
Most inspection techniques examine electronic devices and printed-circuit boards (PCBs) while they are in an unpowered state. Infrared inspection takes a different approach, requiring that circuits emit heat, which means the device or board must be operating at some level.
Infrared inspection can provide information about the board operation that an automated optical inspection (AOI) system can't see. Insufficient solder, for example, increases circuit resistance at the solder joint and therefore raises the temperature sufficiently to be detected by an infrared camera.
Also, a faulty circuit will exhibit a different temperature profile from a good one. Comparing the circuit's behavior with the corresponding behavior of a known-good equivalent (either a real one or a simulation) reveals whether the circuit should pass or fail.
|
|
|
The object on the top includes several materials and varying degrees of surface roughness. The infrared image of the object clearly shows differences in emissivity. Courtesy of Flir Systems. |
Like many theories, however, this one makes several assumptions. Most significantly, it assumes that the way in which a circuit emits infrared radiation is both known and predictable. The validity of that assumption depends on the materials and techniques used in the circuit's construction.
Of particular concern is the efficiency with which the circuit emits infrared radiation (emissivity) when compared with a theoretical “black body.” The left side of the figure shows an object made of several materials and with varying degrees of surface roughness. The right side of the figure shows a corresponding IR image made while holding that object at a constant temperature. The differences in emitted radiation result from variations in emissivity.
PCBs consist of a substrate and numerous individual components made of a variety of semiconductor materials exhibiting a range of emissivity values from reflective to flat black. How, then, can you measure temperature gradients accurately and reliably in this non-homogeneous environment? Ideally, you would need to know the actual emissivity at every point on the board.
Increase system accuracy
Chris Bainter, scientific segment engineer for Flir Systems, explained how manufacturers address this problem. “Some manufacturers literally paint the board black during board development to increase the accuracy of the results, but that approach would prove unacceptable during routine production,” said Bainter. “Other companies construct a 'black box,' putting the infrared camera and the board under inspection inside it. Aside from the need to compensate for the heat generated by the camera in that enclosed space, the approach is expensive and time-consuming.”
Another technique uses “image subtraction,” in which the infrared inspection system software creates a thermal baseline representing an ideal temperature profile, subtracting that ideal from the inspection results. Since the emissivity of a particular point on the board is unlikely to change much from one (allegedly identical) board to the next, the subtraction operation renders the differences largely irrelevant. Bainter noted, “Subtraction can occur directly in the camera as well as in turnkey or custom-developed software. In this situation, the software can adjust the measurement results to compensate for varying emissivities across the board.”
Like all test and inspection techniques, infrared inspection is not a panacea. Implementing it requires paying attention to both its principles and its limitations, as well as to the equipment available to make it work.
| For further information |
| For an introduction to infrared inspection, see “Characteristics and use of infrared detectors,” a 43-page pdf on the Hamamatsu Photonics Web site. sales.hamamatsu.com/assets/applications/SSD/Characteristics_and_use_of_infrared_detectors.pdf |
No related content found.
- 0 rated items found.
Datasheets.com Electronic Parts & Inventory Search
185 million searchable parts
- Part Number
- Description
- Inventory
- Products
- Manufacturers
View All Webcasts
