Counterfeit Parts
How Counterfeits Enter the RF Supply Chain
Counterfeiting is overwhelmingly a problem of the secondary market. When a part goes obsolete or onto an extended lead time, a buyer under schedule pressure turns to independent distributors and brokers rather than the franchised channel. That is where fraudulent product enters: recovered die from electronic waste shipped overseas, sanded and re-marked to a higher grade, or rejected units diverted from a manufacturer's scrap stream. The 2012 U.S. Senate Armed Services Committee investigation traced roughly 1,800 counterfeit incidents covering more than one million suspect parts, the overwhelming majority of which originated from recycled e-waste. RF and microwave devices are disproportionately targeted because GaAs and GaN MMICs, low-noise amplifiers, mixers, and precision connectors carry high unit value and long lead times, making the economic incentive to counterfeit them strong.
The danger is that a counterfeit can pass a casual incoming check. A re-marked commercial-grade die in a military-marked package may power up and even meet room-temperature specs, then fail at the −55°C to +125°C extremes the genuine military part was qualified to, or degrade after a few hundred thermal cycles because the recycled solder joints and re-tinned leads were never designed for that stress. For an RF front end, a marginal counterfeit LNA can raise system noise figure by 1 to 3 dB or introduce intermodulation products that only appear at full drive, problems that surface in the field rather than on the bench.
Avoidance is therefore built on two pillars: controlling where parts come from, and verifying what actually arrived. The first pillar is procurement policy, sourcing from the original component manufacturer or its franchised distributors whenever possible and treating any open-market buy as high risk. The second pillar is physical and electrical inspection, escalating through a tiered test plan only as far as the assessed risk requires so that high-confidence lots are not over-tested while suspect lots are torn down to the die.
Tiered Detection Under AS6171
SAE AS6171 defines the test methods that the avoidance standards invoke, organized so inspection escalates from inexpensive and nondestructive to costly and destructive. The flow typically begins with external visual inspection under 10x to 40x magnification, looking for blacktop resurfacing, ghost marks from prior part numbers, inconsistent lead finish, and date or lot codes that do not match the claimed production window. Solvent scrub tests with acetone or Dynasolve strip painted-on remarks. X-ray imaging then compares internal wire-bond loops, die paddle size, and lead-frame geometry against a known-good golden sample without opening the package.
When visual and X-ray flag a unit, decapsulation with fuming nitric acid or a laser exposes the die so the mask revision, manufacturer logo, and die part number can be read directly. Electrical confirmation is the final arbiter for RF parts: a suspect LNA is swept for gain, noise figure, and S-parameters across its rated band, and any parametric outlier relative to the datasheet condemns the lot. XRF spectroscopy cross-checks the lead finish against the part's declared RoHS status, catching tin-lead parts masquerading as lead-free and vice versa.
Counterfeit Avoidance Equations
n ≈ ln(1 − C) / ln(1 − p)
For zero-acceptance (c=0): with confidence C = 0.95 and a target defect rate p = 0.05, n ≈ ln(0.05)/ln(0.95) ≈ 58 parts.
Noise-figure impact of a marginal LNA in a cascade:
Fsys = F1 + (F2 − 1) / G1 + (F3 − 1) / (G1 × G2)
Noise factors here are linear ratios, not dB. Because the first stage adds directly and its gain divides every later term, a counterfeit front-end LNA whose noise figure rises from 1.0 dB to 2.5 dB pushes a typical cascade from about 1.1 dB to 2.6 dB, nearly the full 1.5 dB stage change.
Authentication confidence after k independent tests:
Pmiss = ∏i=1k (1 − di)
where di is the detection probability of test i; stacking visual, X-ray, decap, and electrical drives the combined miss probability toward zero.
Counterfeit Class and Detection Reference
| Counterfeit Class | What Was Altered | Primary Detection Method | Tier / Cost | RF Risk |
|---|---|---|---|---|
| Recycled / harvested | Used die re-marked as new | Visual + decapsulation | Tier 1 to 4 | High (latent failure) |
| Remarked | Date/lot code or grade altered | Solvent scrub + XRF | Tier 1 to 2 | High (temp range) |
| Overproduced | Unauthorized extra build | Traceability audit | Documentation | Medium |
| Out-of-spec / reject | Failed units diverted | Electrical / S-parameter test | Tier 3 to 4 | High (parametric) |
| Cloned die | Unauthorized copy of IP | Decap + die imaging | Tier 4 | High (security) |
| Forged paperwork | Documents only, part genuine | GIDEP / certificate audit | Documentation | Low to Medium |
Frequently Asked Questions
What are the most common types of counterfeit electronic parts?
The AS6171 and GIDEP taxonomies recognize seven classes. Recycled parts harvested from scrapped boards and re-marked as new are the most common by volume; the 2012 Senate Armed Services Committee study tied the majority of its roughly 1,800 incidents to recycled e-waste. Other classes are remarked, overproduced, out-of-spec/defective, cloned die, tampered, and forged documentation. High-value RF parts such as GaAs MMICs, LNAs, and precision connectors are frequent targets.
How do you detect a counterfeit GaAs MMIC or RF connector?
Detection escalates through the AS6171 tiers: external visual at 10x to 40x for blacktop and ghost marks, solvent scrub for painted remarks, X-ray against a golden sample, then decapsulation to read the die and mask revision. For RF parts the deciding step is electrical screening, sweeping gain, noise figure, and S-parameters across band so parametric outliers condemn the lot. XRF spectroscopy verifies the lead finish against the declared RoHS status.
What is the difference between SAE AS5553 and AS6081?
AS5553 is the OEM/integrator control plan covering franchised sourcing, risk assessment, incoming inspection, and GIDEP reporting. AS6081 governs independent and broker distributors who trade outside the franchised channel and mandates a far more aggressive test regime. AS6171 supplies the actual test methods and sampling plans that both invoke, and DFARS 252.246-7007 makes flow-down of these controls contractual for U.S. defense work.
Can a counterfeit part pass an initial power-on test?
Yes, and that is precisely the hazard. A re-marked commercial die in a military-marked package often meets room-temperature specs, then fails at the qualified temperature extremes or after a few hundred thermal cycles. A marginal counterfeit LNA may add 1 to 3 dB of noise figure or generate intermodulation products only at full drive, so the defect surfaces in the field rather than on the bench, which is why temperature and life screening per MIL-STD-883 matter.