(Re)Design for availability

The global electronic industry (as well as multiple other industries) are all facing delivery problems due to "the Perfect Storm" referred to as Chipageddon or Chipocalypse...

Are your products suffering from the global semiconductor shortage too? Have you tried "everything"? Been up and down, and in and out of your supply chain? Vacuumed the brokers and the gray market (and other less predictable or secure channels)? Evaluated all the second sources (if any)? Turned every pocket inside out??

Fig. above: Supply chain disruption (Image: Pixabay)

Then now what to do?

Depending upon which news site you read, not to say believe, for sure the shortage will last for several months still, and the most pessimistic sources say 2023-2024. This is even if manufacturers are working around the clock to increase production capacity, and if all new component orders are accepted on a non-cancellable, non-returnable (NCNR) basis only (to prevent excessive stockpiling).

Now, what do to? Have you considered respin or redesign for availability? Poor availability of components is the most common reason for redesign also during more normal times. Respin may sound easy, but if you are a development engineer, you know it's not that straightforward (and it may impact firmware and software as well). There were good reasons the components in your current bill of material (BOM) were selected when you designed the product in the first place. Functionality was one important reason, assumably the most important. Perhaps finding it in your existing component library was another? Add to that environmental issues like extended temperature, EMI, shock and vibration, country of origin and Export restrictions, brand, and so on. And not to forget the cost of course!

Image: Both passive and active components are hard to get (Image: Pixabay)

More discrete logic or more SOMs/COMs?

So, if you are now forced to redesign for availability, you need to look for "something" that will fulfill the same spec. To succeed, you probably may have to accept compromise on your redesign, e.g. larger or more expensive components than what's ideal. You may also have to accept utilizing more PCB real estate to allow for the usage of multi-footprint compatibility components. Or perhaps you could replace an existing discrete logic with FPGA (programmable logic)? That said, the lead times of FPGA's are part of the current shortage situation. You may also have to split or subdivide your design into multiple cards (e.g. modularization) instead of a single module or PCBA only. This could cater for higher availability in terms of sourcing in the future, e.g. if you include an industry-standard "computer-on-module" (COM) in your design, you will accelerate the (re)design task, and you can focus on the application-specific parts of your product. And through COM's footprints and interface standardization, you can choose interchangeable COM's from multiple COTS manufacturers as well as adapt quickly to market trends, and upgrade to newer and more available technology simply by replacing your legacy COM.

Temporary solutions? What about regulatory requirements?

Another approach to handle the temporary component shortage is introducing a "Minimum Viable Product" (MVP): a similar product but with less - more basic - features (around the world, a lot of Original Equipment Manufacturer (OEM) companies now turns to Lower-Tech versions of their products to be able to supply). This could bridge you to when you can continue the supply of your existing, full-featured product. You will need your customers' approval, but this may work if they need your product.

But what could be more difficult, is if there are any regulatory requirements involved? If your existing product carries specific certifications like maritime approval, MIL-spec, RED, or similar, you may have to prove that the change will not affect the approval and certification, or you will have to check if you can waive that for the MVP.?

Design for availability

A respin or redesign may or may not solve your problem short term. But it can adapt your product into a more robust product to future shortages. How? You need a plan (Keep in mind the saying of Benjamin Franklin “If you fail to plan, you are planning to fail!). And you need "situational awareness" (typ. from design houses, EMS manufacturers, distributors, and component manufacturers).?

Be careful not to select a component by convenience, e.g. because it's already in your component library, but select the components carefully. Allowing for second-source or even multi-source components and alternative (ALT) parts in your design and BOM - the latter which can be considered as a shopping list for your product – should be pursued. But it’s not always possible, especially if you go with highly integrated components in your design (a more generic - discrete - design may safeguard component availability better). Still also here you may have to compromise on features to avoid single-source.

Talk with your purchase department, with your current EMS or ODM, and with your distributors to have their recommendation before design in. Check launch date and expected life cycle. Quite a large number of suppliers have their own, long availability road maps. Ensure the supplier your select has an active life cycle management policy, and that you will receive important life cycle info as well as PCN's. Here we could add; look to the automotive industry as components used here, typically offer long-term availability. However, as the automotive industry is part of the "Chipageddon" shortage, these days it may be easier to source "non-automotive components".

Multi-level and multi-source Bill of material

When talking about BOM, make sure to structure your bill of material in a good way. Include approved original manufacturers, and not (only) distributors on your approved vendor list (AVL). And unless you have very specific requirements, consider marking on the BOM “any/open” or tick the NO-box for Criticality. This indicates if the component must be exclusively sourced from one vendor or if your EMS or sourcing partner can suggest their preferred supplier. Thus, this is normally easier on passive components where alternatives are more interchangeable than active integrated components.??

If you have established a good and detailed BOM with prioritized multiple-source alternatives (where possible), you make the work much easier for the purchasing, sourcing, and planning departments to navigate through future shortage situations.

Fig. above: 3rd party electronic BOM analyze database (image: Z2Data)

Also, with a detailed BOM, you can run it through a third-party web-based BOM manager to get an instant lifecycle forecast, compliance status, and analyze all the parts in your bill of materials. Among the most popular online BOM managers, you will find IHS, Z2Data, and OpenBOM. Also, powerful PCB design software like i.e. Altium Designer offers so-called active BOM options fully integrated.

No quick fix and no guarantee

The above are all ideas on how to design (or redesign) for short-, mid-, and long-term availability - to give your products more component options to work with. However, it does not take into consideration raw material shortages (glass substrates, foil copper, wafer, plastic, epoxy resin, and more), hence nevertheless how good your engineering design flexibility and strategy is, unfortunately, there is no such thing as a guarantee for availability.?

BTW (did you see it coming?), did you know Data Respons Solutions can assist both with (re)design for availability and sourcing hard-to-get components, modules, and systems? Why not give us a call?

For more information, please also see our website:

https://datarespons.solutions/manoeuvring-the-perfect-storm/