-
- News
- Books
Featured Books
- smt007 Magazine
Latest Issues
Current IssueBox Build
One trend is to add box build and final assembly to your product offering. In this issue, we explore the opportunities and risks of adding system assembly to your service portfolio.
IPC APEX EXPO 2024 Pre-show
This month’s issue devotes its pages to a comprehensive preview of the IPC APEX EXPO 2024 event. Whether your role is technical or business, if you're new-to-the-industry or seasoned veteran, you'll find value throughout this program.
Boost Your Sales
Every part of your business can be evaluated as a process, including your sales funnel. Optimizing your selling process requires a coordinated effort between marketing and sales. In this issue, industry experts in marketing and sales offer their best advice on how to boost your sales efforts.
- Articles
- Columns
Search Console
- Links
- Events
||| MENU - smt007 Magazine
Choosing the Most Appropriate AOI System:
Clarifying Common Misconceptions
December 31, 1969 |
Estimated reading time: 7 minutes
By James Gibson, Landrex Technologies Inc. and Pamela Lipson, Ph.D., Imagen Incorporated
All AOI systems have features and characteristics in common; it's what they don't have in common that can make the difference between choosing the correct one for current and anticipated applications, and making an expensive error. It's also not just about the vendor; there are many excellent systems available on the market from a number of established suppliers. However, many misconceptions about AOI systems can result in choosing the wrong one for a given application, and that can be costly in many ways. This article looks over the range of AOI systems, and then examines some common myths about them, with a view toward debunking those myths.
Features in CommonMost automated AOI system solutions have the following features in common from the very basic to more sophisticated. This list could be longer, but the fundamentals include:
- • An image capture system, comprising illumination and cameras;
- • a mechanical stage that moves either the board or the camera head;
- • A set of algorithms to inspect the board;
- • a programming paradigm;
- • a defect review and analysis software package;
- • and all the above packaged as an in-line or off-line system.
A closer look at the individual categories will reveal significant divergence. To make it even harder for an equipment purchaser, even if the features sound good, it is difficult to predict how well these individual assets work until they are put to a long-term test, e.g., operating in-line and running millions of boards. The most common question the supplier is asked is how the purchasing engineer will be able to tell if a given machine will fit the company's inspection needs. Here, common misconceptions about evaluating AOI systems come into play, misconceptions that can result in a poor purchasing choice.
Misconception: "I can assess an AOI machine by looking at each of the subsystems independently."Not exactly. Most engineers want to dissect an optical inspection machine into its subparts and then make comparisons between machines based on these subsystems. To be more specific, most people want to focus on the tangibles that they can measure, such as hardware specifications. Unfortunately, breaking an AOI system down into subparts and then making comparisons is not equivalent to assessing the system as a whole. The system camera is an example of why it is not possible to use specs alone to benchmark an AOI system.
The camera has many measurable attributes. People like to ask how many pixels the camera has. Other questions include "What is the size of a pixel?" or "Is the camera color?" Actually, the camera cannot be assessed without understanding the image processing technology that is being used to analyze the images it captures. On one hand, more pixels could mean higher resolution and therefore better accuracy, which is beneficial. Or, on the other hand, it could mean more noise for the system and a slower processing time, which is detrimental. Without comparing the hardware and the software together, it is hard to know the effects on an AOI system.
One can have a similar discussion about whether the camera has color or not. Color looks more pleasing to the eye and allows the system to make color-based measurements, which seems good. On the flip side, most solder joint systems inspect joints that intrinsically do not have any color, possibly mitigating the benefit of adding color. In addition, color triples the amount of information that is transmitted to the system and can slow processing down.
The main question to ask about the camera and, more generally, the image capture and lighting system, is whether or not this configuration makes defects more evident to the AOI system. Additionally, the question should be amended to include whether or not the image analysis algorithms can identify the particular signature of defects coming from the camera and lighting system. Basing purchasing decisions on the answer to any single question is an incomplete characterization of the system. In the context of evaluating an AOI system, we need to simultaneously review the camera specifications, lighting configuration, and image analysis algorithms together, because they are so tightly coupled.
Misconception: "The price of an AOI machine is the price of the base system plus normal support."Many people purchase AOI machines based on the base price plus some support costs. The rationale is that most AOI systems likely will deliver similar performance plus or minus some delta. Thus, given that one can't really tell how well these machines will work in one's manufacturing environment without trying them for an extended period of time, the assumption is that choosing the lower-priced system will reduce the risk of the purchase.
From an overall, detached perspective, this appears both logical and sensible. The hypothesis that most AOI machines deliver roughly the same performance eliminates steps like carefully scrutinizing them, making it easy to make a decision based on one attribute, i.e. price.
The challenge to this line of thought is that the base price of the system rarely reflects its true cost. There are two hidden costs that make some systems more expensive than others. The first is the amount of false failures and false accepts that the system generates. False failures are good items that are flagged defective by the system. They require human inspection cycles to review and clear them. Large numbers of false failures also have a negative psychological effect in that they cause the human inspectors to lose faith in the system. If the human inspectors expect to see only false failures, they might miss a true defect, or worse, stop using the system.
False accepts are even more costly. False accepts are true defects that are flagged as good. Undetected, these defects remain in the product, which results in lower end-of-line yields or failures in the field.
The second hidden cost is the programming time required to tune the system to reduce the false accepts and false rejects after initial board setup and programming. This hidden cost is rarely mentioned because it is difficult to quantify. However, in the worst case, we have seen where each AOI system at a facility requires a dedicated programmer to manage the false accepts and false rejects on a board-by-board basis. One should ask why the system, if programmed on a set of sample boards, would exhibit varying performance on the next set of boards. The reason is that there is a tremendous amount of variability in how the boards, parts, silkscreen, orientation marks, solder, etc. look from one instance to another. In the PCB component industry, there is tight control of how these materials perform, but little control of how these materials look.
AOI systems that can automatically tolerate normal process variation but still identify true defects will be much less costly than those that require constant tweaking. The base price of an AOI system rarely reflects the true cost of ownership. False accepts, false rejects, and programming time can easily inflate the true cost and render the base price of the system irrelevant.
Misconception: "Post-reflow is the best place in the line for an AOI inspection system."Most EMS providers planning to purchase an AOI system look for a post-reflow system, primarily because they want to catch all of their defects in one place with one machine. Thus, on the cost side, post-reflow provides the most advantageous inspection opportunities for the price of one machine. From the quality perspective, most people are concerned about the integrity of their solder joints and feel that post-reflow is the best place to assess these defects.
Considerable research has shown that if solder paste is applied properly and parts are positioned correctly, in concert with a well-designed board, there generally will be fewer defects at post-reflow. From a cost point of view, it is better to catch errors just as they are created, or in-process, rather than after reflow, when many boards will have been built with the same defect. Post-reflow inspection often obscures what is happening earlier in the process. The oven can change the defect profile. Grossly mis-positioned parts may become tombstoned. Parts that are placed incorrectly may be pulled toward the center with leaded solder, and to a lesser degree with lead-free solder. By measuring the parts after placement but before processing in the reflow oven, one can assess the true performance of the placement system and make adjustments in real time. This enables you to not only catch the defects as they occur, but to prevent future gross defects. In-process detection and measurement may be just as effective in catching defects and more cost-effective by finding defects early and preventing future defects.
Conclusion Misconceptions about AOI systems and how to evaluate them in advance of a purchase can result in costly mistakes. These can include a higher- or lower-than-needed equipment investment, higher cost to maintain the systems, and cost impact based on where in the line the systems are deployed. The key to making the right choice is to avoid false logic in making comparative assessments of only a limited scope of the machine's attributes. A better working knowledge of how AOI systems gather and process information is essential. To make the right choice, one has to ask the right questions, and then get the right answers.
James Gibson, president, Landrex Technologies Inc., may be reached at www.landrex-us.com. Pamela Lipson, Ph.D., CEO, Imagen Incorporated, may be reached at www.imagen-inc.com.