How Self-Learn Testers Differ from the Reliability of Cirris Testers
Expectations are high and delivery times are urgent. We operate in a high-volume, get-it-done-yesterday world. It’s how manufacturing and commerce is done these days. What happens often is that on a quick-turn project, due to time-sensitive elements, some of the final testing requirements get waived because you don’t have tools, resources—and time—to complete them. But the consequences of skipping this phase can be devastating if your cable and harness assemblies are faulty. A cable could ship 100% tested, but fail in the field.We all know how absolutely critical it is that you don’t have cables that arrive DOA. Proper testing avoids that, and anytime you are not using a validated testing method, there is risk. The Cirris Testing System guarantees robust, reliable testing. It’s why we use Cirris testing on all our cable assemblies and wire harness assemblies.
Let’s take a step back here and discuss what “self-learn” testing is, which is what most other contract manufacturers use in the cable and harness industry. With the self-learn testing process, step one is to build a first article. Once built, the tester looks for all the continuity points. One of the inherent risks with this is that if some of the wires are not connected properly, it is learned incorrectly.
Validation on a self-learn testing application is critical, and also can leave you vulnerable to failures. When validating a self-learn, the number of test points gets counted; they, of course, must equal the number of wires on the assembly. The danger is that you can miss something. When “learning,” you are acquiring a signature that is unique to that product. You have to match that same signature. If your testing results return anything different, it means it is not wired the same as the original. If you don’t produce the match and you cannot get that signature back, then you are left to wonder if that initial wiring was wrong. And you then need to explore whether it is the test fixturing or the cable itself—putting you back to square one. With self-learn testing, there must have been something wrong with the original (either testing or assembly), which can remain wrong, even if you set up another test.
Why We Test with the Cirris Easy-Wire System
We work against a schematic. Because it is clearly mapped out on a schematic, we know what the wiring is, and we program each connection accordingly. We don’t just work against a “known good,” but have verifiable back-up. The Easy-Wire system shows us if we are testing on a point that we are not supposed to be on—and it identifies where the missing point is.
We test to all points. The Easy-Wire tests to all points. It will identify, for example, 20 out of the 50 points, and it designates those in the program; it learns them as open. If there is a mis-wire, it will show it and provide us with an accurate readout. One of the many advantages of using Cirris testing is that we test the product and the test fixturing as well.
We use connector nomenclature. Sure, you could use an Excel spreadsheet and create a homemade matrix and spend time and resources cross-matching each time you test. We prefer the efficiency of loading the Easy-Wire database with the proper connector nomenclature, so it instantly matches every single wire, each time we test. The universal interface between the tester and the mating connector recognizes—and “translates”—each test point. It fully populates the connections in an easily readable form, so from the operator’s standpoint, it makes interpreting the results seamless.
The way it works: each tester is equal to an actual tester/connector on the assembly. This allows us, in connector nomenclature, to know if there is a missing wire (P17, for example), alerting us to the exact problem.
The benefits to this extend beyond the initial testing phase. The database information provided with Easy-Wire supports the testing scenario in other ways. If there is a missing wire, we know it is missing on the cable or the fixture doing the testing. This means we don’t have to count test points; we don’t have to come up with the signature—and, we don’t have to go looking for why.
Password protected. Importantly, the program schematics for these testers is done by qualified engineering or quality personnel (they retain the password), who understand the customer schematic and ensure that all test points are programmed and accounted for. Once the program is generated, the actual customer test programs are securely password protected and cannot be modified by the operators. Using Easy-Wire takes the guess work out of testing—operators don’t have to relearn the test every time they set up, as they would in self-testing.
Everything is documented. Documenting each connection with the original program means we will meet each test’s requirements when every wire matches up. That also means that if someone builds a new test fixture and it is not wired identically, then it will never pass the test. And that’s okay, because we want repeatability, reliability, and consistency in our testing. Everything is date coded and documented so we can replicate it every time. This is particularly helpful in situations when we only build a job once a year. When you set up the test 12 months later, it’s easy—everything is recorded and ready to go. If it’s a self-learn, you need to remember what was done 12 months ago … and you need to replicate that at this point.
Any quality process requires consistency, and because these tester programs are saved to a hard drive, they can be shared electronically and can be used simultaneously at multiple workstations. This capability also allows Fourstar to provide serialization and formal test reports with each shipment, as required by the customer.
Repeatable logic. We have a very specific protocol for how we set up our testing—every time. The system we developed is based on a repeatable logic. We are aware that common practice for some companies, especially when delivering to tight launch dates for time-to-market (TTM) projects, can be to rush through it just to get the test done. But, if there is a problem, they can’t go back. Having a set process allows consistency and care with every test.
Feedback loop. In the event an operator mis-wired and errors appear, the feedback on the tester tells us exactly what we need to fix. We get clear visuals—green and red dots—that inform us where and how to clear the fail. This feedback loop is very efficient and a valuable time saver.
Complete library synched with ERP. We keep a complete library of mating connectors across all products line, which is also synchronized into our ERP. This fully populated library guides us in determining what the unique run list is for each customer.
If someone enters a new connector part into our inventory, it sends an alert for a test fixture to be put into our system. It might have 30 different parts that plug into a certain fixture, but we will ensure we have a mate (or mates) for it. If we have a mate for it, we assign that mate to it. If the mating connector isn’t in the system, we assign it in the stock location so we have it in our library. This way, when/if we get the job, we have easy—and quick—access to it. Because that part number is in our library and inventory, the test fixture is available before the first article is built.
This helps us support the quick-turn contract manufacturing environment we operate in—especially with TTM and new product introduction (NPI) builds. We also maintain the test fixtures with the same quality standards we apply to our production lines. When we return test fixtures back into the library—each one gets inspected for wear and damage. If broken or worn, they get replaced right away, so they remain on hand if needed. Of course, everything related to this process is also documented.
Highly scalable. Self-learn testing is iterative, manual, and timely, requiring you to put the two fixtures on, plug in and test, and compile the learns. Cirris testing has the learning built in. It learns the connections, looks for opens and shorts, and knows there should be continuity between the two wires. Benchtop boxes typically have fixed test points—up to 250 test points. Cirris can be a stand-alone tester or is expandable to thousands of test points and multiple settings for test sensitivity, so is highly scalable.
Superior testing. We travel to customer sites and participate in trade shows and events regionally. From observations we make and feedback we receive from people in the field—both customers and third-parties with no skin in our game—we hear that our testing is superior to what electronic manufacturing services providers are doing regionally. At big shows, we repeatedly hear that we are the only cable shop they know of that is using the full capability of Cirris. We know that the big investment we’ve made—and continue to make—in time, money, and resources sets us apart from others in our industry. We are proud of this and happy to provide this reliable support to our customers through this important production phase.
Completing Your Product On Time!
Fourstar takes pride in our ability to be a reliable time-to-market and new product introduction partner in the electronic manufacturing services (EMS) industry. Providing Cirris testing is one vital part of what enables us to get your products to market on time, fully tested, and in top quality.
Fourstar is equipped to help you navigate through what can often be very challenging new product introductions and can quickly get you on a clear path to success. Fourstar can assist you with “finding a better way.” If you’re ready to begin the conversation now, contact us to explore how beneficial a supportive EMS partnership can be for your business.