Counterfeit and Remarked Components: A Buyer's Checklist
Most counterfeit parts are not clever fakes. They are ordinary parts with a new date code, a cleaned-up package, or paperwork that points at a factory the seller never bought from. All three are visible before you pay.
Quick Answer
- Paperwork first: certificate of conformance, lot traceability, and a factory path you can name.
- Date code: decode it, compare with the reel label, and check that one order is not three lots mixed.
- Physical: marking style, solder residue, lead condition, and whether the bag was opened.
- MSL: floor life starts when the bag is opened. A bag opened last month needs baking, not installing.
- Electrical: ID read, standby current, and a curve trace catch most remarks in minutes.
1. Why this lands on the buyer
Counterfeit risk is usually described as a supplier problem. In practice it appears at the buyer's desk as a quote: a part that has been on allocation for months, suddenly available, at a price that is either suspiciously low or suspiciously high. The decision to accept it is made by whoever signs the purchase order.
Industry bodies track the trend — ERAI maintains the largest counterfeit reporting database, and its alerts are the standard reference when a lot is questioned. Standards exist too: SAE AS6081 for distributor counterfeit avoidance, AS6171 for test methods, and IDEA-STD-1010 for inspection. None of them help if the buyer does not ask for the evidence.
The practical framing: you are not trying to prove a part is genuine. You are trying to establish that the paperwork, the physical part and the electrical behaviour agree with each other. When they disagree, that is the finding.
Four different problems, one word
"Counterfeit" is used for four different things, and they need different responses:
- Fake: a different die in a package marked as the real part. Fails ID read or curve trace quickly.
- Remarked: a genuine part of a lower grade or different speed, re-marked as the higher grade. Passes functional tests, fails at temperature or at speed.
- Refurbished: desoldered or reclaimed parts, cleaned and re-reeled. Genuine die, unknown thermal and moisture history.
- Overproduction: genuine parts made outside the authorized channel. Electrically correct, no factory warranty.
The first is easy to catch. The second and third are the ones that reach production, because they look right on a bench test and fail in the field or at temperature. That is why the inspection sequence matters more than the brand of the tester.
2. Check one: paperwork and chain of custody
Start with what the seller claims. A usable answer names the manufacturer, the factory or authorized channel, the lot, and the date the stock was acquired. "Original, in stock" is not an answer.
| Document | What it proves | What it does not |
|---|---|---|
| Certificate of conformance (CoC) | The seller asserts the parts meet the spec | That the seller bought from the factory |
| Factory invoice or packing list | A traceable purchase path | That the parts in the box match the invoice |
| Reel label with lot and date code | Batch identity | That the parts were not re-reeled |
| PCN / EOL notice | Die or process changes | Anything about this specific lot |
A seller who cannot name the path is not necessarily selling fakes. But the buyer's risk has changed from "is this part real" to "am I willing to bet a build on it", and that is a different decision with a different price.
Some parts attract remarks more than others. Commodity linear regulators, memory, and anything with a long allocation history — for example an AMS1117-3.3 or a W25Q32 — are worth a closer look because the volume and the price difference justify the effort. That is not a reason to distrust the channel; it is a reason to check the lot.
3. Check two: date code is not a serial number
Two things go wrong here. First, re-dating: a genuine older part is remarked with a recent code to look fresh. Second, mixing: a reel assembled from multiple lots, which is invisible until the codes are read across the strip.
What to do: read the code on several parts along the strip, not just the label. Compare with the label and the invoice. If the codes differ, ask why. Mixed lots are legal and common in small quantities, but they must be disclosed, and they change the warranty conversation.
Also check the marking method. Laser marking is standard on modern parts, but it should still match the manufacturer's style — font, depth, orientation, and the pin-1 orientation. A part with a perfect datasheet logo that is slightly the wrong shade of grey is a remark, not a variant.
Why re-dating exists: old stock sells at a discount, fresh stock sells at a premium. The gap is the incentive. The check is cheap: read three parts, compare with the label, and compare the label with the invoice date. If the numbers do not line up, ask before you accept.
4. Check three: the physical part
Ten minutes with a loupe and a phone camera catches a surprising amount:
- Leads: coplanarity, solder residue at the base, uneven plating, bent or re-straightened pins.
- Body: sanding marks, solvent haze, a package that looks newer than the date code implies.
- Marking: inconsistent laser depth across parts in the same lot; a logo that is the wrong weight.
- Packaging: re-reeled tape with a hand-written label, missing desiccant, a humidity card that is already pink.
- Mixed parts: two die shapes in one reel — visible under a magnifier at the edges.
Desoldered parts are the hardest to spot because they are genuine silicon. The tell is usually mechanical: solder wicking on the leads, a slightly dull finish, or a package that has been cleaned. It matters because a re-used part has unknown thermal and MSL history.
A cheap loupe and a phone camera are enough for the first pass. Photograph the label, the reel edge, and two or three parts at the ends of the strip. Those photos become the baseline if a dispute starts later, and they cost nothing to take at receiving.
Package-specific checks: for QFP and QFN, look at lead coplanarity and whether the edges show a uniform cut; for BGA, check ball diameter consistency and reflow marks; for SOIC, look for solder residue at the lead-to-body junction. None of these prove authenticity, but together they make a re-marked or re-used part expensive to disguise.
5. Check four: MSL and floor life
Moisture sensitivity is defined by J-STD-033. The floor life clock starts when the sealed bag is opened, and the bag is only valid if the humidity indicator card still reads dry.
For a buyer, the practical checks are: was the bag sealed, is the desiccant still active, and does the label carry an MSL and a date. If the answer is "we opened it to count", the parts need baking or a documented floor-life reset before reflow, otherwise the failure mode is popcorn cracking — which passes electrical test and fails in the field.
This is also why a "new" reel from a broker deserves the same question as a "new" reel from a distributor: the bag is the only evidence of how the parts were stored, and once it is open that evidence is gone.
6. Check five: electrical and physical tests
| Test | Catches | Misses |
|---|---|---|
| ID / JEDEC read | Wrong die, wrong density | Same die, different grade |
| Curve trace (pin-to-pin) | Open, short, wrong internal diode | Marginal speed or temperature grade |
| Standby current | Refurbished or damaged die | Functionally correct remarks |
| X-ray | Die size, bond wires, delamination | Correct die in the wrong package |
| XRF | Plating and RoHS differences | Anything electrical |
| Decap | Die markings, true manufacturer | Nothing — but the sample is destroyed |
For most buyers, the practical sequence is: ID read, curve trace, standby current. That combination catches the majority of remarks and wrong-die parts for a few minutes of technician time. X-ray and decap are for disputed lots, not routine receiving.
The economic logic is the same as any sampling plan: spend the cheap checks on every lot, and the expensive checks only where the cheap ones disagree. A curve trace on five parts from a reel takes minutes; a decap takes days and destroys the sample.
One caveat: electrical tests confirm that the part works now, at room temperature, at the test conditions. They do not confirm the temperature grade, the speed bin, or the long-term reliability. Those are paper claims, which is why the paperwork check is not optional.
Sample size matters too. Testing five parts from a mixed reel is close to pointless — the point of sampling is to catch lot-level differences, and a reel assembled from two lots can pass five random parts. If the packaging looks inconsistent, widen the sample or reject on packaging alone.
7. Red flags, in the order they appear
| Signal | What it often means | Ask for |
|---|---|---|
| Price 30 % below the market | Refurbished, mixed lot, or a trap | Factory invoice; photos of the reel |
| "In stock, same day" on an allocated part | Broker with no physical stock | Photo of the actual reel with a date |
| Label written by hand | Re-reeled stock | Original packaging photo |
| Part is "new" but the code is three years old | Re-dated, or genuinely old stock | Year of manufacture and storage history |
| Bag opened "to check" | MSL clock already running | Floor-life record or bake certificate |
8. What the failure costs
The arithmetic is worth doing before the negotiation, because it justifies the inspection and changes how much a few cents of traceability is worth.
Part: 1,000 pcs at $0.50 = $500 Inspection (ID + curve, 1 h) = $60 Line stop, 1 day, 20 people = $12,000 Field return + rework, 2 % = $8,000+ Relative risk: 12x to 40x the part value
A 5,000-piece build using a 1,000-piece reel at $0.50 is a $500 line item. Stopping the line for a day costs orders of magnitude more, and a recall costs another order again. The inspection budget should be set against the failure cost, not the part price.
This is also the argument for paying for traceability up front: the difference between an authorized channel and a broker is usually cents per part, and it is the cheapest insurance in the BOM.
9. Disposition: quarantine first
Two rules matter more than the flow itself. First, quarantine means a physically separate location with a label — not a different shelf. Mixed stock is how a single suspect reel becomes a suspect batch.
Second, document before you negotiate. Photos of the reel, label, date code and packaging; the invoice; the seller's written claim. If the lot turns out to be counterfeit, that record is what supports a claim, a report to ERAI, and a customs notification.
10. Put it in the purchase order
| Clause | Wording to use |
|---|---|
| Traceability | "Manufacturer and factory lot must be stated on the invoice; mixed lots require written approval." |
| Date code | "Date code within 24 months of shipment; older stock requires written disclosure." |
| Packaging | "Sealed factory packaging; MSL label and humidity indicator intact." |
| Inspection rights | "Buyer may inspect and test; nonconforming lots return at seller's cost." |
| Counterfeit | "Suspected counterfeit parts will be quarantined and reported; seller bears replacement and disposal cost." |
These clauses do not stop a bad lot. They change the negotiation after it is found, which is the part a buyer actually controls.
Two practical notes when writing them. First, "date code within 24 months" is easier to enforce than "fresh stock", because it is measurable at receiving. Second, require the seller to state the manufacturer explicitly — not "original" — because that single field is what makes a traceability claim checkable.
11. FAQ
Is buying from a broker always risky?
No. Vetted independent distributors with documented inspection are a normal part of the market, especially for obsolete stock. The risk is not the channel; it is the absence of evidence.
How old is too old for a component?
There is no universal limit. Solderability and MSL history matter more than the date itself, but many buyers set 24 months as a disclosure threshold and require baking beyond it.
Does a CoC prove authenticity?
No. It is the seller's assertion. It becomes useful when combined with a factory invoice and a lot that matches the physical parts.
Can I test my way out of the risk?
Partly. Electrical and X-ray tests catch most wrong-die and refurbished parts, but a correct die in a lower grade package can pass. That is why paperwork and physical checks come first.
What do I do with confirmed counterfeits?
Quarantine, photograph, notify the supplier in writing, and report to the relevant body (ERAI and, where applicable, customs). Do not resell them and do not return them before documenting.
How does this apply to small orders?
The same way, with cheaper checks. For ten pieces, the ID read and a photo of the reel cover most of the risk; for a production reel, add MSL and a sample curve trace.
What if the seller refuses inspection?
Treat it as a red flag and price the risk accordingly. A supplier who will not provide a factory invoice or allow a sample test is asking you to accept an unquantified risk on a production build.
12. Sources
- ERAI, Counterfeit electronics reporting and alerts — industry database used for suspect-lot checks.
- JEDEC, J-STD-033 — handling, packing and floor life of moisture-sensitive devices.
- SAE, AS6081 — counterfeit electronic parts avoidance for distributors.
- SAE, AS6171 — test methods for suspect counterfeit electronic parts.
- IDEA, IDEA-STD-1010 — inspection standard for electronic components.
Checking a suspect lot?
Send the manufacturer, date code and a photo of the reel. We will tell you what the paperwork should contain and which checks are worth running.