How to Source an EOL or Obsolete Electronic Component

Obsolete sourcing is lifecycle work, not a purchase order

When a discontinued part shows up on a shortage report, the request that reaches sourcing is usually "find us 4,000 of these." That framing hides the real question, which is how many you will ever need again and what supply strategy carries the product to its own end of life. Answer the second question and the first one usually answers itself.

Treating it as lifecycle work also changes who is in the room. A buyer alone cannot size a lifetime demand, approve an alternate, or accept the risk profile of an open-market purchase. A component engineer alone cannot commit the cash or judge whether a supplier's paperwork is credible. The steps below assume engineering, sourcing, quality, and whoever owns the product's P&L are all reachable within the week.

Step 1 — Quantify remaining demand

Every later step is scored against this number, so get it before you get quotes. State a horizon explicitly: "as long as we sell it" is not a quantity. Pick a date the business will defend, then add what you owe after that date.

• +Production forecast to the stated horizon, by product, including any variant or regional build that uses the same part.
• +Service, repair, and warranty demand, which usually runs years past the last production build and is the component most often forgotten.
• +Contractual spares commitments — some customers bought a support term, and that term is a quantity.
• +Scrap, yield, test attrition, and rework allowance, expressed as a percentage you can defend rather than a round number.
• +Internal consumption: engineering builds, field trials, qualification lots, and the boards that live on someone's bench.

Step 2 — Check the inventory you already control

The cheapest obsolete part is one you already own. Before anyone opens a distributor search, reconcile every pool of inventory your organization or its partners hold. This step routinely closes the gap outright, and it takes hours rather than weeks.

• +Your own stockrooms and MRP on-hand, checked physically rather than trusted from the system — the two diverge on slow-moving parts.
• +Inventory consigned at your contract manufacturer, and any buffer stock the CM bought against your own forecast.
• +Kitted material left over from a cancelled or reduced build, which often sits outside normal on-hand reporting.
• +Field-service and repair-depot stock, including units earmarked for a program that ended.
• +Excess held by a different site, business unit, or product line — larger organizations regularly buy on the open market a part another division is writing off.
• +Material on order that has not yet arrived, with its schedule confirmed rather than assumed.

Step 3 — Work the authorized channel before anything else

Authorized (franchised) distributors hold inventory well past a manufacturer's last-ship date, and that inventory carries the strongest traceability available to you: an unbroken chain back to the original manufacturer, the manufacturer's warranty terms where they still apply, and returns handled through a normal commercial relationship. If the quantity you need exists there, the sourcing decision is usually finished.

Ask your distributor about more than the number on the website. Franchised distributors can often see stock at other regional warehouses, know whether a factory buffer or a die bank exists, and will tell you when a line has been transferred rather than truly ended. Some discontinued devices are picked up by an authorized aftermarket manufacturer — Rochester Electronics is the best-known example — which continues to supply the original die or a licensed rebuild through the authorized channel. That is not the same thing as the open market, and it is worth checking before you assume the part is unobtainable.

One caution shapes everything downstream: a genuinely discontinued device does not have a meaningful factory lead time. Whatever schedule you are quoted reflects handling and shipping against inventory that already exists, or the build schedule of an aftermarket arrangement if one exists. Nobody is going to start the fab again because your order is large.

Step 4 — Understand what a last-time buy actually commits you to

If the last-time-buy window is still open, you have an option that closes on a date. The last-time-buy date is the deadline to place the order; the last-ship date, usually months later, only constrains when the material arrives. Plan backwards from the earlier one, and leave room for the funding conversation, because a lifetime buy is a cash decision before it is an engineering one.

Sizing is the whole exercise. Take the remaining-demand number from step 1, add the scrap allowance, subtract confirmed inventory, and then sanity-check the result against three constraints that quietly invalidate large buys: shelf life and moisture-sensitivity handling for the packages involved, the physical storage and stock-rotation capacity to hold parts for years, and the write-off exposure if the product's forecast is wrong by the amount forecasts are usually wrong.

A bridge buy is the underused middle option. Instead of buying the full remaining life, buy enough to cover production until an alternate can be qualified, and start the qualification the same week. It converts an all-or-nothing forecast bet into a smaller, time-boxed one, and it keeps the schedule honest because the bridge quantity makes the qualification deadline explicit.

Step 5 — Evaluate independent-market inventory

Once authorized stock is exhausted, the remaining material sits with independent distributors and brokers: excess from OEMs and contract manufacturers, distributor de-stocking, cancelled programs, and inventory bought specifically to be resold later. This is a legitimate and often necessary channel, and for a truly obsolete part it may be the only one that has the quantity you need.

It is also a different channel, with different economics and different provenance. Price tracks scarcity rather than a published break schedule, availability is a specific lot rather than a replenished line, and the seller's knowledge of where the parts have been varies enormously. The questions worth asking are practical ones: who held this material and for how long, what documentation comes with it, what testing has already been performed and by whom, what the return terms are if it fails your inspection, and whether the quantity offered is in hand or being sourced from someone else on your behalf.

Step 6 — Assess counterfeit and traceability risk deliberately

Counterfeit risk is why the open market is treated differently, and it is managed rather than eliminated. The industry has written the practice down. SAE AS5553 covers avoidance, detection, mitigation, and disposition of counterfeit electronic parts for the organizations that buy them; AS6081 addresses the same problem specifically for independent distributors; AS6496 covers the authorized channel; and IDEA-STD-1010 is a widely referenced visual acceptance criteria document for parts bought on the open market. Defense, aerospace, medical, and some industrial customers require conformance to one or more of these by contract, so check your own flow-downs before you assume the choice is yours.

The practices teams reference span a wide range of cost and depth: documented traceability back to the original component manufacturer, incoming visual inspection against a published acceptance standard, X-ray of the package and lead frame, solvent or XRF checks for remarking, decapsulation and die inspection, electrical test to the datasheet over temperature, and solderability testing. Payment terms that release funds only after inspection passes are common. Industry reporting programs such as ERAI and GIDEP exist so that a suspect part or supplier is not each organization's private discovery.

How much of this you apply is a policy decision, not a universal rule. Depth should track application criticality, order value, and the supplier's history with you — a consumer accessory and a life-supporting instrument do not warrant the same regime, and pretending otherwise either wastes money or accepts risk without saying so. Decide the regime before you are under schedule pressure, because that is when it gets skipped.

Step 7 — Evaluate alternates in parallel, not afterwards

Qualification is almost always the long pole, so alternate evaluation should start at the same time as sourcing, not after sourcing fails. Running them in parallel is also what makes a bridge buy sizeable: you cannot scope a bridge without an estimate of how long the alternate takes to approve.

Two kinds of candidate exist and they should not be blurred together. The first is the replacement the manufacturer names in the discontinuance notice — a claim from the party that knows the original best, and usually the right first look, though the manufacturer's definition of "compatible" is not automatically yours. The second is a functional alternate found by comparing specifications across the market, which may fit better commercially but carries no vendor statement behind it.

Specification similarity, however it is computed, is a prioritization signal. It tells you which candidates deserve engineering time first. It is not approval, not a drop-in guarantee, and not a substitute for review — package and footprint, electrical margins across temperature, timing, errata, firmware and register-map differences, qualification status, and long-term availability of the candidate itself all remain engineering's call.

Step 8 — Compare lifetime buy, qualification, and redesign on the same terms

These three options are usually argued in different currencies — the buy in dollars, the qualification in engineer-weeks, the redesign in schedule — which is why the argument goes in circles. Force them onto one page with the same three columns: cash out and when, calendar time to a shippable product, and residual risk if the assumption behind the option turns out to be wrong.

An illustrative case makes the shape clear. Suppose a motor-control board has four years of remaining production at roughly 6,000 units a year, plus a ten-year service obligation on units already shipped. A lifetime buy has to cover about 24,000 production pieces plus a service estimate plus scrap, paid for this quarter, stored for a decade, at whatever the open market is charging for a part that no longer has a list price. Qualifying a functional alternate might cost a couple of engineer-months plus one customer requalification cycle. Redesigning the power stage might cost a quarter of engineering time and a fresh EMC campaign. Written down side by side with real numbers, that comparison usually resolves in an afternoon; argued from memory in a meeting, it does not resolve at all.

Combinations are normal and often correct. Bridge buy plus qualification is the most common. Qualify an alternate for new production while servicing the installed base from a smaller final buy is another. The point of the comparison is not to pick one column but to know which risk you are choosing to carry.

Step 9 — Record the decision, including the options you rejected

The sourcing decision will be questioned later — by a customer audit, by a quality investigation into a field failure, or by the engineer who inherits the product and wants to know why it uses a part with no authorized supply. A decision that is not written down gets re-litigated from scratch, usually under worse conditions than the first time.

• +The part, and the evidence that it is obsolete: the discontinuance notice, the distributor status change, or the stated source for the status.
• +The demand figures and the horizon they assume, so a future reader can see what would have to change for the decision to be wrong.
• +The options considered and the ones rejected, with the reason — this is the part everyone skips and the part auditors ask about.
• +The channel chosen, the supplier, and the traceability documentation received, including any inspection or test performed on receipt.
• +Who approved it, on what date, and whether customer, quality, or regulatory sign-off was required.
• +A review date, because a five-year inventory position is a decision that should be revisited before year five.

Where PCNshark fits

PCNshark covers the evidence and the record around this path rather than the buying itself. It works from the notices your team already receives — uploaded or emailed in, not discovered from distributor feeds — extracts the affected part numbers and dates, and matches them against your BOMs so exposure is a list rather than a hunch. On a component record it holds a manual lifecycle status with the evidence behind it, distributor sourcing context across roughly 15 distributors (stock, MOQ, price breaks, estimated lead time, whether the source is authorized or independent, and a market median at 1k), and alternate candidates on Team plans and above: the manufacturer's recommended replacement extracted from the notice, and functional candidates ranked by specification similarity. Distributor figures are refreshed on demand and are estimates, not quotes. The demand math, the supplier qualification, and the buy-versus-qualify-versus-redesign decision stay with your team; what PCNshark keeps is the part-level decision history, so the reasoning outlives the people who were in the room.