When I considered the title for this column, I seriously considered calling it “From the frying pan into the fire.” I thought this because, as I’m sure you’ve noticed recently, the component shortage problem has only worsened; we’re now seeing other supply lines breaking down. For example, the other day, I made my usual lunchtime run to a fine food establishment—the local Sonic—hoping to indulge in my customary Chicago dog. Although I know it’s scary to consider Sonic as “fine dining,” imagine my horror when I heard that they were totally out of hot dogs—of all things! Now, I can handle component shortages and not bat an eye, but when something comes between my Chicago dog and me, that’s where I draw the line. Similarly, PCNAlert connected to IHS Markit confirms that EOL notices for components are increasing at an alarming rate. Furthermore, IHS says, “The estimated cost to manufacturers for missing one EOL notice is $20,000 to $50,000.” It seems everyone is feeling the pain.
Because of my position with Altium, I have the privilege to speak to customers every day. A common recurring theme is that companies are making a considerable paradigm shift from engineering new product lines to supporting and sustaining legacy product lines. Several well-known companies informed me just this week that they have ceased all recent engineering efforts or will delay the release of new products and concentrate solely on keeping the legacy products alive. That is a direct result related to the component shortages. Keeping those legacy products lines viable is key to whether a company will survive.
I am the type of person that when something like this happens, I want to know why. What is the root cause of the issue? First, the main reason for the changing company’s focus is the availability of components. A critical point to mention is that the parts are available, but the problem is with who has them. If you go to such sights as Octopart and look at who has the available stock, you’ll see that the smaller component broker houses hold much of the public component stock.
Furthermore, many of those companies are overseas. They have bought up and are hoarding (for lack of a better word) the components. By controlling inventory, and because of that pesky economics of supply and demand, you now see the result of the considerable increase in component cost. In a recent conversation with a company, they told me an FPGA that they used for some time at $75 per chip has now increased to over $1,000 per chip. To get the needed stock of 500 pieces to do a production build would cost over a half-million dollars. That is not sustainable for any company to do that. I might add that it’s also outrageous. That fact forces companies to move engineering efforts over to maintaining the legacy designs, as well as the status quo. We’ll never know the financial impact of such decisions.
Executing a plan for dealing with “Not Recommended for New design” (NRFND), or quickly obsolete components, means becoming very creative and avoiding the dreaded re-designing and new board spins.
Which I will say is not a solution at all, when you think about it. Instead, it turns into a classic case of the tail wagging the dog. You might solve the immediate emergency, but the situation is so volatile that it is only a temporary fix. Placing the component in a design is the easy part. The big question is whether a specific part will even be available.
Multiple-Part Choices vs. Alternate Components
The new focus is sustaining product lines and simply trying to find replacement components for those that are unavailable or have gone obsolete. I have heard the terms “multiple-part choices” and “alternate components” used interchangeably. But there is a difference between them.
As we all know (if you have been an avid reader of I-Connect007), a component has various parts of information and models. Some of that information is static, and others are dynamic; you guessed it, the part choices of manufacturer and suppliers are not just dynamic but very dynamic.
A multiple-part choice is another component that exactly matches the form-fit-function (FFF), also known as a drop-in replacement. This whole process usually occurs much more with the discrete parts. For non-discretes, you most likely only have a single manufacturer, especially with specialty integrated circuits (ICs), etc. But, because of the availability of MFG and vendors of discrete components, a common practice is to add as many multiple-part choices as possible. I always assure there are at least five to six different MFGs for each discrete. You’ll often find that you cycle through those part choices pretty quickly. And when a specific part choice begins to deprecate, they are taken off the preferred vendor list and replaced with another.
On the other hand is the alternate components. They are ones that first meet the FFF criteria of your circuit, which takes care of the first initial problem of getting the component on a specific footprint. But the alternate components may vary in some of the electrical parametric specifications.
Deciding on the alternate component is walking a fine line. We know how Indiana Jones felt when selecting which cup was the actual Holy Grail. We should consider the same warning to “choose wisely.” When you have a completed designed circuit, the hope is to have that device operate in a specific manner under particular conditions. Once either one of those is changed, you slowly drift away from the ideal operating situation. Making changes in the designs impact everything that comes after it—including any compliance testing or approvals already done.
When considering an alternate component, look at all aspects of the component, including electrical and mechanical specifications. However, unless you want to be going through this exercise again very shortly, the first thing needed is to look at the lifecycle information and predictions for the near future on availability. Unfortunately, that isn’t easy right now.
Keep in mind to remove as much of the risk as possible when looking at alternate components. That takes a broad understanding of the circuit. Finding an excellent replacement is simple enough by understanding the electrical parameters and the trade-offs on the specific differences and whether they are an improvement or a detriment to the circuit. Let’s say you are using a 1K resistor, 0402 size, at a 5% tolerance. Your alternate component should not be a “worse” component, such as selecting one with a 10% tolerance. You loosen the parameter, and the component quality worsens. A better choice would be to use a 1% tolerance. For each component in your library, know each specification/parameter and which “direction” changes can take.
The final point is to be proactive in working with both your multiple part choices and alternate components. The library is a living, breathing thing and is constantly changing; with that said, frequently check and know the status of your components by identifying earlier any problems or issues. That is something I would check daily for now. If you are reactive to component shortage and availability, it will substantially impact your product lines and probably take them down entirely. No products to your customers. Not good, not good at all.
John Watson, CID, is a customer success manager at Altium.