How test is keeping up with technology changes.
Test is becoming increasingly complicated as new technologies such as flexible electronics begin playing mission-critical roles in applications where electronics have little or no history. Rigid Flex Printed Circuit Boards
Although flexible circuitry has been around for while, testing needs to catch up as these circuits are deployed across a variety of markets where conditions may be extreme. In many cases, sensors for monitoring what is going on inside a device, such as rate of flow in an industrial valve — or outside of that device, such as heat and noise in a mechanical robotic arm, or even in the sleeve of a jacket — have to be retrofitted to existing systems or designs. The easiest way to make that happen is through flexible electronics.
The flexible electronics field is growing. There are new varieties and hybrids that are stretchable, textile-based and printable. But because this technology has been more of a research project than a commercial venture, there are no standards for testing and little data to show how these materials age, drift or otherwise degrade over time, particularly in specific applications.
“We try to borrow as much as we can from the traditional rigid printed circuit board community,” said Will Stone, director of printed electronics integrations and operations at Brewer Science. “We’re looking at environmental stresses and other issues. The challenges are twofold. Number one, there are traditional components being mounted to a flexible circuit board. We need to make sure the joints, whether those are conductive epoxy or some other material, can withstand the motion, because they are not designed for that. Second, we have to look at the connecting traces themselves. Just by flexing you can get micro-cracking. We run those through the usual gamut of testing, such as environmental chambers, stressing the flexing thousands to millions of cycles, just to ensure the integrity of the device.”
On rigid chips, there are latent defects that may not show up for years. Conductive inks and solder add a whole new dimension to what needs to be tested.
“The inks are pretty stable once you get the process down,” Stone said. “It’s the epoxies and solders that we use. A lot of these weren’t designed to flex at all — especially the solders. We’re playing in a rigid world and trying to make a rigid world flexible. That’s the challenge. Any joint is where you will have the greatest area of concern.”
So far, there is too little history to show how well this kind of technology will fare over time under extreme conditions involving heat, cold and vibration. So while companies like Brewer are accelerating testing, both physically and through simulation, it still takes time to gather enough data about what can go wrong and how to fix it. But one thing is becoming clear in this segment of the market — if a defect is detected, it’s easier to replace a board than to fix it.
“It’s heading toward a disposable type of approach, because re-work is extremely difficult on these circuits,” he said. “We have had some success at doing rework, but it’s time-consuming and painstaking. It’s usually more economical to replace it with a new board, even in the R&D phases.”
Toward standards Some standards do exist. Flexible circuits are being used in low volumes as interconnects and circuit boards for electronics systems. The circuit wires are cut from a sheet of a copper suitable for the amount of bending the circuit will undergo in its lifetime. The circuit sits on a substrate of metal, glass or plastic, and is encased in plastic. It can be semi-rigid to flexible, and have one or more layers. There are even two-sided circuits.
The IPC standards (IPC 6013D) covers flexible circuit boards, splitting them into three classes, depending on the quality level needed. The standards generally describe the materials, manufacture, testing and quality, but flex circuit makers also follow other certifications, depending on the industry use case. And despite these efforts, gaps remain.
“While we can apply the majority of test methods and standards used for standard copper flex printed circuit boards for flexible hybrid PCBs, currently we see a gap in reliability standards for two very important areas for FHE — stretchable or skin/body-worn electronics,” said Wilfried Bair, vice president of engineering at NextFlex, a membership organization for the FHE industry that works with IPC on standards.
IPC (formerly the Institute for Printed Circuits, but now called the Association Connecting Electronics Industries even though it retains the IPC acronynm), is an industry standards body that grew out of the PCB industry. The group is working to produce standards for flexible hybrid electronics, stretchable wearables, and e-textile wearables.
“When you have an emerging technology it’s really hard to wrap standards around it because you’re dealing with IP companies that don’t want to share what they’re doing,” said Chris Jorgensen, IPC’s staff liaison for a half-dozen standards committees. “That’s understandable, but we see that in other areas, even for traditional electronics assembly. What our committee is doing is getting things in place so that when the flex hybrid group or several companies from within the supply chain say, ‘Hey we need to have a reliability specification for flex hybrids,’ or ‘We’re talking about something that is truly a hybrid product,’ we can say, ‘Here are two or three standards that were written for printed electronics. Take the content of these standards and revise it to where it meets your needs.’ So you’ve got a starting point.”
Because FHE is a hybrid of printed and flexible electronics, it uses some of the existing flexible circuit or printed circuit standards as a template, while also getting industry feedback on how the standard should be tweaked. The committees already have sketched out the standards and found holes that need to be filled.
How quickly those gaps are closed up depends on how quickly flexible electronics become commercialized and standardized.
“At some point, you’re going to go from onesie, twosie product development to scale-up,” said IPC’s Jorgensen. “When you go to scale-up, you’re really going to need to have standards that you can rely on because you’re going to have multiple customers that you’re going to deal with. And rather than dealing with each of their own standards, which is going to mean repeat testing, it’s better to have an industry standard that you can point them to.”
The rule with standards is that there must be multiple sites that can conduct the same tests before it can be called a standard. Since the first projects that industry uses to prove the technology can scale up are often one-offs or low numbers, there may be only one site that can do the tests. So the committee has to do a Gage R & R (Gage Repeatability & Reproducibility) evaluation of the test method before the test standard is published for industry. Industry cannot have a single source where it goes for test, and test equipment must be readily available.
e-textiles standards Creating standards for e-textiles, wearables is a work in progress now. IPC has begun the work but wearables and e-textile standards have some tricky issues. These new electronic media are on the standards frontier. “We don’t have an approved committee yet, or approved working group or an approved project, but there’s a proposal that’s been developed by the e-textiles industry and this includes manufacturers of wearables products to develop an IPC standard for e-textiles-wearables. This group right now is going through a laundry list of characteristics that would be important,” said Jorgensen.
Some of the difficulty is the analog to other flex hybrids breaks down with e-textiles, wearables, which means committees need to start from more fundamental definitions. Jorgensen at one time suggested the three class system used in flexible circuit standards for the new e-textile, wearable standards, but everyone realized that “it’s not a direct fit,” he said. “When you’re talking about worn on the human body, when is it not a class three? And think about the environment that it’s in. The human body is a really punishing environment.”
Washability, durability, sustainability, and safety are some of the concerns for industrial segments interested in wearables, e-textiles. How do you wash and dry an e-textile/wearable? With what chemicals if any and how often? “You don’t typically take a circuit board or a cell phone, and throw it into the washing machine. But you’re talking about having to take a product that somebody’s going to pay a considerable amount of money for and then how many wash cycles are you going to push it through? You’re looking at exterior environments. So, if it’s a jacket or if it’s a shirt, you’re going to be outside in the elements, so you could be looking at dust, humidity but rain, snow, heat, cold, so that’s why when we looked at it as a committee, we couldn’t really find direct lines from ‘class one’ [for flexible circuits] to what would be, you know, class one requirements for wearables, same thing all the way up to class three. And so that’s why they’re looking at general product areas.”
The product areas are still initial concepts, but they will feed into how and what tests must be performed. “The test methods that would need to be applied: how do we look at safety, how do we look at sustainability, and then breaking out the requirements based on product areas”, said Jorgensen. “Is it for fashion? Is it for personal protective equipment like damper for use for like fireman? Is it for the military? Is it for space applications? And so that way the characteristics are required testing will be more applicable to different markets, but everything would be in the same standard. Is it high durability, high risk? High durability, high risk are actually not health monitoring but [indicates] a medical device. Because if they if it fails, the risk is that the patient could then have adverse effects.”
To keep moving, IPC’s Jorgensen said, “We don’t really know exactly how [the product definitions are] going to play out right now. What the group is doing is they have a small ad hoc group, and they’re going to present their proposal to the committee chairs I asked them to do that by the end of this month. And then we would announce the formation of a national IPC working group to begin to develop the standard or standards — it could be multiple standards. I’ve asked for right now and the ad hoc group is agreed to this, keep it all in one standard. It makes it easier for industry.”
Machine learning’s role A key part of the manufacturing process in flexible circuits is subtractive. Building circuits is done by removing copper from what started as a full sheet. That means errors are visible and can be identified and characterized in the test and inspection processes.
Some companies already are collecting test and inspection data in this area, which can be used to identify latent and real issues. That data, in turn, can be used to train machine learning systems to identify patterns that can lead to reliability issues in the field.
“The greatest challenge of flexible circuits is actually inspection,” said Sam Jonaidi, vice president of Automotive Solutions at Optimal Plus. “Inspection to me is a form of a test. Instead of testing with electrical signals, we test with visual pictures — optical — because often enough, the defects that are in a printed circuit board or a flexible circuit board can be detected with AOI (automated optical inspection systems). You really need a very sophisticated image analytics capability.”
The AOI machines are the workhorse of flexible circuit testing, said Jonaidi, but image analytics is key to improving the process. By aggregating all the images that come off the AOIs, Optimal Plus produces intelligence about errors. “We do a lot of very fancy, very insightful things when we start looking at thousands of images, tens of thousands of images, millions of images. As we say, a picture’s worth 1,000 words. In this case, it’s actually worth millions of words. This is something new in the industry.”
When images are aggregated, “a whole bunch of other things emerge,” he said. “For example, we may actually see a particular kind of defect happening at a particular quarter of the panel more often than elsewhere. Or we actually find that certain defects are quite random. They’re happening all over the panel. By those kinds of observations, we actually can tell a lot more about the proceeding process.”
The data is useful in the upstream and downstream processes. “We can step in and say, ‘Your acid baths are not running correctly, or your photolithography is not going as planned. So you have issues upstream. We’re constantly trying to push the detection and the corrective action of upstream,” said Jonaidi.
The data also can be used for predicting reliability problems. “Having this information, we can also go forward — downstream. In other words, we can go to test. And once we all have the test information, we can start correlating the two. Let’s say you’re measuring a particular circuit for its resistive value, which often is what’s done in a flexible circuit. From reading the resistive value of these circuits, the narrower the circuit, the higher its resistance. The wider the circuit, the lower its resistance. So if I just take micro measurements of my circuitry, I now can connect it to my images. If I have all the images, I can predict what my test results should say and I can produce limits. For example, I can say, ‘Look, this particular circuitry that I just processed with this particular panel, by the time it gets to you Mr. Test, it better measure let’s say a 1.2120.’ That allows us to now be predictive. So we can say that for the next batch of flexible circuits you’re going to be producing, on average, it is drifting to the right, or on average is different to the left. Then we can prepare the downstream process.”
Inspection plays a key role with flexible electronics. OrboTech, now owned by KLA, and CyberOptics, both make automated optical inspection equipment. CyberOptics uses light reflections to find issues quickly. “We take the reflections of light — and light bounces around anytime it sees a reflective surface — and we create highly accurate measurements and inspection reports,” said Subodh Kulkarni, CyberOptics’ president and CEO. “The CyberOptics Multi-Reflection Suppression (MRS) technology is built into AOI systems and 3D metrology systems for semiconductor inspection. The MRS technology digitally tags the light and suppresses the multiple reflections of light.” (KLA uses CyberOptics’ MRS technology for semiconductor backend inspection and reports.)
With flexible circuit boards, most often a DC test is applied to circuit boards, unless the flexible circuit board is used in a high-speed application, such as an Apple iPhone. The AC test is used on the high-speed circuits. “If you’re coming from a semi world, flexible circuits is really not very sophisticated at all from every aspect,” said Optimal’s Jonaidi, “If you go to AC testing, which is high-speed testing, it becomes a lot more interesting. The tooling and measurements become more challenging. They have to be more high-precision, and usually you have to use full probe contact for every point. We’ve got to make sure that multiple probe heads come down, so the tooling becomes more expensive and more finicky. It has to be cleaned up more often, maintained and calibrated, because you’re running high-speed signals through these circuits. And, again, you’re trying to produce circuits that can operate and function properly.”
How the circuit is sandwiched together comes into play, too. “The entire process has a bearing on your end results,” he said. “So your scrap rate is higher, and to some extent you cannot avoid it. For example, if I’m trying to measure impedance of a particular circuit that runs on a couple of layers, if that layer has been sandwiched too hard, now the spacing is shorter and therefore it doesn’t meet my criteria. And I could have not prevented that very easily, because the circuits can only be tested once it’s been assembled in the multi-layer sandwich. And of course, scrap is higher and so are the costs. But again, these testers are not millions of dollars each. They are maybe $50,000 to $100,000 testers. And they’re usually slower because these tests take a little bit longer. And as I said, they’re finicky, so you have to make sure that you spend a good amount of time. But overall, this represents a fraction of the circuits being produced today in the world. Most other flexible circuits are just done in a passive manner.”
Conclusion Flexible electronics are just beginning to roll out across a variety of applications after years of research and development. It will take time to add the same kinds of standardized testing and inspection required for mass deployment of this technology. In effect, technology that is being used to improve reliability across a number of industrial and medical applications, needs some standardized ways of ensuring the technology itself is reliable.
“Standards for manufacturing, testing and reliability are increasingly important in flexible hybrid electronics, and while some initial work on testing as part of an early project call was completed at NextFlex, the community has looked to SDOs like IPC to apply the rigor around standards development that the industry requires,” said Scott Miller, director of Strategic Programs at NextFlex. “In the future, standards for design and test will absolutely be necessary to facilitate wide adoption of FHE technology.”
—Ed Sperling contributed to this report.
Coming next month: 5G and automotive test. Related Stories
Flexible Devices Drive New IoT Apps
Name* (Note: This name will be displayed publicly)
Multilayer Pcb Reverse Engineering Email* (This will not be displayed publicly)