Though a medical device designer may be an expert in the specific application and sensing technology employed in a device, this expertise does not necessarily translate to building that technology on a large scale with stringent reliability constraints. Moreover, because the variables and potential combinations are difficult to define, so to are standard design guidelines.
Conventional flex circuits use rolls of Kapton, a Dupont™ polyimide film pre-clad with metal, and are bonded with adhesives between each layer. This process keeps layer thickness stuck at around 1-5 mils. Thin-film flex circuits rely on custom spun polyimide polymer films on glass, and this approach enables miniature flex circuits up to six layers, with layer thicknesses as thin as 10 microns or less. Trace lines and spaces for this technology can be as small as ten microns, and via hole diameters can be as low as 15 microns, or .00059 mil. The significant increase in circuit density and proximity of circuit traces and features, naturally leads to electrical and mechanical considerations more akin to integrated circuits (ICs) than prior flex circuit technologies.
This blog series aims to outline the top 7 considerations biosensor circuit designers should pay close attention to early in the stages of prototype development.
If certain special features used to enhance the competitiveness of your biosensor or medical device over other products is required it’s important these be on the table early with your manufacturing partner. This could be anything from bonding bumps, use of exotic metals, micro-circuit stiffeners, and extremely dense circuit patterns. All of which, need to be considered during the design stage to avoid unfortunate redesign and revalidation.
Exotic materials may incur substantially greater costs that will price them out of the market, and there may be a limit to how, or if, they can be processed. Some metals are unable to be processed in the presence of others, as certain metals and processing agents may interact chemically. For example, platinum can only be used on a gold circuit if it’s added to the metal stack during the optimal process step. This is a result of the platinum metal etchant also etching all other metal types. This is somewhat similar to a solder hierarchy used in PCB and PCBA manufacturing.
That if it Works in the Graduate Lab it Will Work at Volume. It’s common practice for many medical device companies to collaborate with advanced scientific and engineering Universities to develop next-generation technologies and devices. But, a common pitfall of this practice is that there’s a presumption that the prototype manufacturing tools and equipment used in a University lab—that may have led to working models and small volumes of a circuit—are the same as those required to produce a circuit at high volume. (This is not limited to University research. Expert medical designers at major brands often make similar errors in their own labs). Teaming up with a thin-film manufacturing partner early in the design phase can help to prevent development going down a road that later requires redesign when approaching manufacturing at volume.
It may be hard to imagine that a medical device may be specified to be “too good”. But as any engineering team knows, it’s not a designer’s nature to leave anything on the table in regards to performance. Overspecification occurs when a particular performance parameter is overvalued, or there isn’t a precise understanding of a nominal range for a parameter. This can lead to redesign, material swaps, and manufacturing trials that will increase costs and time-to-market. Ensuring that nominal parameters, and involving a manufacturing partner early in the design process, can help prevent over-specification and improve the competitive position of your biosensor or medical device in the market.
Though struggling through the minutia of correcting every design rules check (DRC) and layout versus schematic (LVS) error may be painful, it is necessary before handing off a design to a manufacturing partner. As the manufacturer may use different software tools to interpret the CAD files, any unaddressable errors could be compounded when being prepared by the manufacturer. Another reason, is that there could be nested errors that only reveal themselves after a particular DRC or LVS error has been corrected. Without correcting these errors, unintended circuit performance is possible, and tracking down and correcting these issues could delay time-to-market.
It is often more efficient to follow a systematic design process that minimizes errors, than having to backtrack each time an error is encountered, especially if correcting one error creates another.
During a CAD file review, your thin film manufacturing partner will analyze every individual line and feature to ensure that is properly drawn. Ideally, a customer supplied email of the DWG or DXF file with accompanying properly dimensioned PDF drawing of the final product can help to ensure the efficiency of this process.
The following are common errors that circuit designers often make:
• Overlapping and/or incomplete lines
• Inaccurate or missing labels of all layers
• Missing features
• CAD drawings that don’t exactly match the specification drawing
• Revision control related issues
Industry disrupting breakthroughs using more compact and flexible micro-circuits in biomedical sensors and other devices are enabling innovative medical sensing solutions with much greater performance, capabilities, and accuracy than ever before. However, these miniature, micron-scale electronics often require another level of circuit design expertise, especially as compared to traditional flex circuit or PCB design. This is especially the case when designing thin film circuits for manufacturability in medium to high volumes. In addition to all of the considerations mentioned above, the application engineering and DFM expertise of your thin film manufacturing partner can also be a make or break decision when pioneering new devices in this highly competitive market.