Most engineers assume that when a part is about to disappear, the manufacturer will tell them. That assumption no longer holds.
In 2025, more than 620,000 electronic parts were discontinued. The majority went end of life with no product change notification at all. No warning, no last chance to buy, no time to plan. One day the part is on your board, the next it cannot be bought.
The result lands on your desk as an unbudgeted redesign, a stalled production line, or a customer asking why their order is late.
Obsolescence used to be slow and fairly predictable. Today it is a constant background risk, driven by faster technology, tighter regulation and fragile supply chains. The good news is that you have more control than it feels like. The cheapest place to deal with obsolescence is not the warehouse or the purchasing desk. It is the schematic. This guide explains how good component obsolescence management starts at design.
What component obsolescence actually means
Component obsolescence is simple to define. A part is obsolete when it is no longer made or supported, so you cannot buy it for repairs or new builds.
It tends to arrive in one of two ways. The first is a managed end of life, where the manufacturer gives notice and a window for a final order. That is the version you can plan around. The second is instant obsolescence, where a part vanishes with no notice and no remaining stock. That is the version that hurts.
Two status labels are worth watching for every part you choose:
- EOL (End of Life): the part is being discontinued and will stop being produced.
- NRFND (Not Recommended for New Designs): the part is still made, but the manufacturer is steering you away from it, usually because it is heading towards EOL.
Obsolescence is not a fault or a failure. It is a normal stage in the life of every component. Treating it as an exception is what catches teams out.
A few terms worth knowing
- PCN (Product Change Notification): the formal alert a manufacturer issues when a part is changing or ending.
- LTB (Last Time Buy): your final chance to order a part before it stops.
- Form, fit and function: the test for a true replacement. A good alternative matches the original on physical size, footprint and electrical behaviour.
- AVL (Approved Vendor List): your documented list of trusted parts and their approved alternatives.
Why components get discontinued
It rarely has anything to do with the part working badly. The reasons are mostly economic and external.
The biggest single cause is low demand. More than three quarters of obsolescence events happen simply because a part stopped selling well, so the manufacturer stopped making it.
Lifecycles are also shrinking. Advanced chips now often last just two to five years, against the ten to thirty years many older parts enjoyed. If your bill of materials assumes a part will be around for a decade, the clock may be running far faster than you think.
Two newer pressures are worth noting. Tighter materials regulations, such as the rules on PFAS chemicals, are forcing manufacturers to requalify or quietly drop parts. At the same time, export controls and shifting global trade are changing which parts stay commercially viable. Both can remove a component with very little warning.
The takeaway is that electronic component obsolescence is mostly outside your direct control. What you can control is how exposed your design is to it.
What obsolescence really costs you
It helps to put numbers to the problem before looking at solutions.
A redesign forced by a single dead part is expensive. Estimates put the cost anywhere from around twenty thousand pounds to close to two million, depending on the product. Survey data suggests most manufacturers face rework costs of up to a quarter of a million.
Clinging on to a discontinued part is not cheap either. Once a chip is out of production, the remaining stock on the open market can sell for ten to fifteen times its original price.
There is a quieter cost too. Engineers already spend roughly a third of their time researching parts and hunting for replacements. Every obsolescence surprise adds to that pile and pulls skilled people away from actual design work.
Then there is the damage that does not show up on an invoice: a stopped line, a missed delivery, a service agreement breached, a customer who starts looking elsewhere.
Aerospace shows the stakes at the extreme end. Long programmes that must run for decades are regularly forced into last time buys, or into clever workarounds such as emulating an old chip with newer hardware, simply because an original part has gone.
How to design a PCB that resists obsolescence
This is where designing for longevity pays off. Strong component obsolescence management is built into the design, not bolted on later. Four habits make the biggest difference.
Design in a second source from the start
For every critical active part, qualify at least one alternative while you are still drawing the schematic. Choose alternatives that share the same footprint and the same key specifications, so a swap does not mean a new board.
Single sourcing feels simpler, but it is a hidden risk. If your one supplier runs short or raises the price, you have no leverage and no fallback. Designing second source components from day one removes that trap.
Build and keep an Approved Vendor List
Capture your validated parts and their approved alternatives in your design documentation. When a part goes EOL, a swap then becomes a quick decision rather than a fresh research project.
This also guards against rushed, “good enough” part choices. Nearly half of engineers admit to settling for a part rather than picking the best one, usually because checking alternatives by hand takes too long. A maintained list does that work in advance.
Check lifecycle status before you commit
Screen each part for EOL and NRFND status at the moment you select it, then keep an eye on PCNs through the product’s life. Catching a risky part at the schematic is cheap. Discovering it once the board is in manufacture, with lead times already stretched, is not.
Rationalise and standardise your bill of materials
Use fewer, more common parts where you can. Widely available standard components with several suppliers are far safer than niche parts from a single vendor. A leaner, more standard bill of materials is easier to source, cheaper to build and simpler to keep alive. This is exactly the kind of thinking that good design for manufacture brings to a project.
How DSL designs for the long term
This is where having design and manufacture under one roof changes things. Because we do both, in the UK, obsolescence is managed across the whole life of your product rather than lost in a handover between separate companies.
Our approach to electronic design is built around products that last decades, not months. We support a design for its entire life, actively managing and mitigating obsolescence as parts and pressures change around it.
We have done this in practice. On one remote monitoring unit, an ageing PCMCIA interface was becoming obsolete, pushing up cost and cutting availability. We redesigned it out, lowering the unit cost and moving the product clear of the obsolescence path while keeping the function the client’s customers expected.
The simplest place to start is our free Design Health Check. It is a no obligation, three step review, carried out under NDA, where our in house engineers examine your design, flag any obsolescent or not recommended parts, and suggest better options. If obsolescence is a worry on a current product, book your free review and let us take a look.
Build boards that outlive their parts
Obsolescence is inevitable. Unplanned obsolescence is a choice.
The pattern behind every point above is the same. The manufacturer may not warn you. Lifecycles are getting shorter. And the cheapest insurance you can buy is bought early, at the schematic, with second sources, a maintained vendor list and a simple lifecycle check before you commit.
Get those habits right and your boards keep running long after individual parts disappear. That protects your production, your margins and your reputation, which is the whole point of designing for longevity.
If you would like a second set of expert eyes on a design before obsolescence forces your hand, that is exactly what we are here for.
Frequently asked questions
What is the difference between EOL and NRFND?
EOL means a part is being discontinued and will stop being produced. NRFND means it is still made, but the manufacturer advises against using it in new designs, usually because it is heading towards end of life.
How will I know if a component is about to be discontinued?
Through Product Change Notifications and lifecycle status data. The catch is that many parts now go obsolete with no notification, so active monitoring and designed in alternatives matter more than ever.
How many alternate sources should I design in?
At least one validated alternative for every critical active part. It should match the original on form, fit and function so it can drop straight in without a board change.
Is it better to stockpile a part or redesign?
It depends on your volumes and how long the product must stay in production. Stockpiling buys time but ties up cash and can cost ten to fifteen times more once a part is discontinued. Redesigning costs more upfront but removes the risk. A design review helps you weigh it up.
When is the cheapest time to manage obsolescence?
At the design stage. Changing a part on the schematic costs very little. Discovering the problem in manufacture, with lead times already extended, costs the most.
