Product Design Tips for
Electronic Supply Chain Shortages
There is no need for me to tell you that the supply chain shortages are not just limited to the difficulty finding Bounty paper towels at the supermarket. It is also not just limited to microprocessors either. Those of us in the business of designing electronic devices over many years are used to sporadic supply shortages. We’ve seen them historically with either memory chips, radio modules, microprocessors, or even passive components. What we have never seen, however, are shortages in all these components at the same time. What’s an engineer to do?
A recent article in Reuters talks about how Apple is able to overcome supply chain shortages. Those of us in the electronics product business for many years know-how Apple does this. What Apple’s buying power enables them to do is “own” their suppliers. They can literally buy up manufacturing capacity enabling them to get allocations that leaves everyone else high and dry waiting for parts. In some ways, Apple’s purchasing power contributes to everyone else’s supply chain problem. Few companies in the world have that level of power. Unless you are selling 100K’s units per month (or more), you do not have that power.
At IPS, we are helping numerous clients trying to get products out the door in the face of severe, sporadic, and unpredictable part shortages. There are various techniques we, and our clients, are implementing to help deal with the current situation. Those of us savvy to the marketplace, know this problem is not going away soon (story for another day). While problems aligning supply and demand may improve within months or a few years, we have to get product out the door now!
Here are some of our suggestions as to how to work around the current supply chain shortage:
→ Design in an Auxiliary Micro Carrier Card
This will also require a close look at support circuitry to support more than one micro version or other parts in the product. Doing this provides flexibility and enables the ability to pivot as needed in response to flaky available supply sources. This may also have an impact on firmware which may need to be “smart” enough to adapt to alternate hardware. The carrier card may help get around problems where you can get the same part in multiple package variants. Using the existing PCB may offer an advantage in that it may have already gone through EMC testing or other regulatory certification processes.
→ Dual Footprint Design
In some cases, it may be possible to find components in multiple package configurations. PCB’s can be designed to accommodate multiple footprints. This particularly applies to alternate package variants of the same micro from a single vendor source. As with the carrier card concept, firmware intelligence or adaptability may be necessary. Again, designing for flexibility is key.
→ Design for Substitute Secondary Components
For secondary components, such as memory chips, it is important to look for similar parts with adaptability in the firmware to enable parts substitution based on available supply. In some cases, the engineer needs to pay attention to drivers that can detect subtle differences. As an example, one can design for the same function with slightly different pinout signals or pinout locations in the board pattern.
→ Accelerate to Next Generation
Not every product is working with the latest and greatest chips from their suppliers. If you are working with an older design, there could be components that are getting close to end-of-life. Of course, one can almost always do a last time buy but that is a stopgap. Supplies may be so short that there could be significant long delays in getting material for a last time buy. I use the word “could be” here because the notice given for last buys can be very short in duration. In fact, there are times we have seen when suppliers suddenly discontinue parts without the option for a last time buy.
If it rains lemons, make lemonade. This might be the time to proactively advance to the next generation chipset so that the product is adapting to parts the part manufacturers want to produce in volume. Oftentimes, end-of-life components are in low volumes to the manufacturer. This can result in escalating cost or reluctance on the part of the chip manufacturer to invest scarce resources producing a part they intend to discontinue.
→ Strip Parts from Eval Boards
Here is a desperation move we have used in extreme cases. Sometimes, we need parts to use in development or low volume early production and the parts are hard to get (or with long or unpredictable lead times) in the open market. In emergency cases, we have literally purchased evaluation boards that have the chipsets we want, harvested the parts off those boards and used them for our particular and most immediate needs. This is not something you really want to do but, as they say, necessity is the mother of invention.
→ Last Time & Risk Buys
Last, but not least, is the age-old practice of last time and risk buying. In “normal times”, when lead times were dragging out or end-of-life was announced on a part, companies would do a last time buy. Today, the situation is akin to hoarding consumer home products. Companies are now stockpiling parts to a degree heretofore never practiced. This is totally disrupting the normal market forces of supply and demand. In particular, companies that practiced just-in-time sourcing of parts are now stockpiling parts “just in case.” In such a market situation, companies need to take a careful look at supply and make unpleasant decisions about buying and inventorying materials well in advance of need. Eventually, the market will be less driven by fear as the supply side catches up with demand but, for now, it is a rational strategy. When developing a new design, the risk buy strategy may include acquiring parts that are a “best guess” at what will be needed later. The risk, of course, is that the wrong parts or quantities may have been pre-ordered. However, applying “engineering judgment” in the decision process can prove helpful in having the anticipated parts on hand when needed.
We live in interesting times. As someone in the business for more than 40 years, lead times of 50-100 weeks were never seen (yes… nearly 2-year lead times!). Now, we see this frequently. Worse, one cannot even trust the lead times being quoted. There is a new pattern of lead times routinely getting pushed 30-40 weeks beyond the quoted lead times at time of order. As engineers, our jobs just got harder. What used to be a minor checkbox of looking into parts availability and second sourcing is now a critical problem. It effects both new products and sustaining the life of existing products (and also the support of product already deployed in the field). Using some of the techniques previously described can hopefully get you through the crisis. If you need help, reach out to us here at IPS. We are helping several clients manage and engineer their way through the supply chain crunch.