I like to have a little fun with this column while also trying to convey some lessons learned, but this month’s topic is more serious than most. If your television stops working, it’s no big deal, right? You probably have at least seven other ways to watch “The Bachelor” (Did I say, “The Bachelor?” I meant to say, “Nova…”). However, when you look at critical segments of the electronics industry—such as medical, aerospace, defense, automotive safety systems, and other on-demand hardware—the eye used for most electronics must be even more critical.
Overall cleanliness is essential for all classes of electronics, but none more so than Class III high-performance electronic products where continued or on-demand performance is vital. Cleanliness is a cumulative measurement of each material and process choice that contributes to the sum. Beyond the fluxing process, the raw components can also contribute to ionic contamination and need to be analyzed separately from the final assemblies. This is especially important in cases of products built with no-clean flux, as there is not a final wash process that can help overcome contaminated bare panels and components.
Generally, medical devices are classified as either IPC class II or class III based on what the expectation is for service and the effect if it fails. For example, a glucose meter is an important piece of medical equipment, but if that doesn’t work for some reason, it is relatively easy to obtain another one from a local drug store. In contrast, with devices like implantable cardioverter-defibrillators, failure could be a matter of life or death.
Active implantable medical devices would be in the Class III realm and have characteristics that are among the most difficult to maintain—extremely low levels of ionic cleanliness in conjunction with miniaturization. This means extensive testing of the assembly and cleaning processes are necessary to ensure that all flux and processing residues are removed from the assembly. Implantable devices demand the highest level of cleanliness on the surface of the product but also call for 100% protection from any outside influences as well. This is achieved using a hermetic seal that can be as rudimentary as a compressed gasket, but bioimplants are more likely to use a full penetration fusion weld.
Medical devices that aren’t placed in the human (or an animal’s) body could waiver between Class II or III based on the endues expectation. For instance, for monitoring equipment that is not hermetically sealed but has an uninterrupted on-demand operation, cleanliness is just as important. Consider one supplier that met all the cleanliness requirement as directed by the customer but saw a rash of field failures. They had tested new production using ion chromatography and global and localized extractions and had very low levels of ionics, but the failures showed high levels of chloride contamination. With the known baseline cleanliness levels, the focus became outside contributors.
The field return housings were visually inspected but nothing was found, so the next step was to look at the housing ionically. An ingress path of cleaning solution was found using localized extractions in several areas around the housing on the inner and outer surfaces. The hospital cleaning staff was using a bleach and water solution on the monitor housings, and because they were not a hermetically sealed product, the bleach mixture had flowed into the housing. Bleach is very high in chloride, so it only takes a small amount to create an electrical leakage path and corrosion. The fix was to include a small bead of industrial assembly adhesives (RTV) around the housings to help eliminate the ingress path, along with alerting the cleaning staff not to use excessive amounts of solution when cleaning that type of equipment. The point of this example is that it isn’t always the fault of the assembler when it came to the high-reliability product failure.
Figure 1: Contaminated ferrite filter.
The next example of a medical equipment failure concerns one of the previously mentioned devices—glucose testers. The problem was thousands of units were showing up at the customer location with dead batteries. The product was built with a no-clean flux, so there wasn’t an opportunity to remove excess manufacturing residues from contaminated components. This is another case of the contract manufacturer not being the root cause of the failure; instead, it was a component manufacturer. A capacitor on the battery line had enough plating residues to facilitate a leakage path that drained the battery in shipment. In this example, the component manufacturing process did a poor job at neutralizing the end termination plating chemistries. This issue was never realized until after the product was fully built because the supplier did not perform any cleanliness testing.
Another scenario I am familiar with concerned a contaminated ferrite capacitor but on a much more important piece of equipment—an external defibrillator. In the case of the glucose tester, the parts were scrapped and replaced because they were considered a low-dollar part that was less expensive to replace than repair.
The external defibrillator was the opposite, so the company decided to perform a secondary cleaning process to remove the contamination from the ferrite filter (Figure 1).
Table 1: Ion chromatography results of ferrite filter.
Another main difference between these two parts was the introduction of the contamination. With the glucose meter, it was an issue at the component manufacturer, but in the case of the external defibrillator, it was an issue with residual water-soluble flux remaining under the filter after an ineffective deionized water (DI) water-only wash process. The condition that made recovery of this even more difficult is there were already battery packs soldered onto the boards. This means the cleaning process had to be localized so moisture did not get on the rest of the board in areas that would be detrimental to the parts on the battery line.
You can see in Table 1 that the level of chloride and weak organic acid (WOA) was so high that cleaning was needed. The cleaning process required a specialized steam nozzle to be used at the filter location that was fortuitously located on the edge of the PCBA, so the likelihood of the steam condensate being deposited on moisture intolerant areas was very low. After the concentrated steam cleaning process, the PCBAs were placed in an elevated heat and humidity chamber for 500 hours to ensure that the removal of the residues was fully effective and the product was no longer at risk of field failure.
The point of this month’s column was that when you are manufacturing high-reliability assemblies related to medical industry, it is critical to take a very close look at the assembly process and all other processes that can influence the end-use reliability—even seemingly unrelated processes, such as post-installation cleaning. It really could be a matter of life or death.
Now, back to “The Bachelor…” I mean “Nova.”
Eric Camden is a lead investigator at Foresite Inc. To read past columns or contact Camden, click here.