Choosing a Verification Method for a Used or Refurbished Aut

Choosing a Verification Method for a Used or Refurbished Automation Module From the Secondary Market

When a used or refurbished industrial automation module arrives from a non-franchised channel, the question is not whether to verify it but which verification to commission. Visual inspection, electrical bench test, X-ray, decapsulation and full teardown each answer a different question, and the wrong choice either wastes budget or — worse — leaves the very defect you were trying to find hidden. We are an independent industrial automation distributor and sourcing desk operating from China, and what follows is how we frame the choice with buyers and what we commission through third-party labs on their behalf. We hold no manufacturer authorization and no testing-laboratory accreditation of our own.

The honest starting point

The first thing to name is what the part is supposed to be. A module pulled from a working machine and sold as used has a different verification path than a part advertised as "new surplus" from a closed-channel warehouse, or a part described as "refurbished" with a bench-test record. The descriptor sets the bar, not the buyer's hope. A used unit does not owe a clean X-ray; a refurbished unit that has not been powered on has not earned the label. Condition, test status and warranty are stated per line before any commitment is taken — that is the rule the desk works to on every quote, and the rule the third-party lab work is layered onto when the buyer wants a deeper check.

The second thing to name is what would count as a failure. If the part will sit in a non-safety role inside a panel being rebuilt for a customer, a power-on bench test that exercises the I/O may be enough. If the part is destined for a machine whose documentation pins it to a specific module and whose warranty depends on traceability, the bar moves to lot-date inspection, X-ray and possibly decapsulation. Without naming the failure mode first, "more verification" becomes a budget line rather than a decision.

The verification ladder, method by method

Five methods come up in practice. Each one is a different lens on the part, and each has a cost and a turnaround that the desk quotes per line rather than publishing as a single fee.

Visual and marking inspection. Photograph every face of the unit, compare the label etch depth, font and spacing against a known-good reference, check the date code format and the lot trace, and look for re-marked or re-surfaced packages. This is the cheapest step and catches the largest single category of misrepresentation — wrong label on a similar-looking part, smoothed-over rework, missing or replaced QR codes. It does not catch a clean re-marking of an authentic-looking chip, and it does not say anything about whether the silicon inside is functional.

Electrical bench test. Power the module on a documented test rig, exercise the inputs and outputs against the manufacturer specification, log the result and return it to the buyer with the test record. For PLC I/O modules, signal-conditioning boards, power supplies and most drives this is the single most useful step. It catches dead channels, blown output stages, missing communications and a large share of refurbishment work that was done poorly. It does not catch latent defects that only show up under thermal stress, and it does not tell you whether the silicon on the board is the silicon the label claims.

X-ray (2D or CT). Image the internals non-destructively. Used to verify die presence, package authenticity and bond-wire layout, and to flag rework that left visible solder voids or extra bonds. Useful when the buyer's failure mode is a relabelled blank module with a clone chip and the visual inspection came back ambiguous. It does not prove the silicon works; an authentic-looking die from a counterfeit lot will still image like an authentic die.

Decapsulation. Chemically open the package and image or probe the die directly. The strongest authentication argument — if the markings and the die layout match the manufacturer's known-good reference, the silicon is the silicon. Costly, slow and destructive to the part, so it is commissioned only when the consequence of being wrong is larger than the cost of the test. Used most often for lot authentication on high-value drivers and CPUs.

Full teardown and component-level inspection. Open the housing, document the PCB revision, list every active and passive component, compare the bill of materials to the manufacturer's reference design, and look for substituted parts, missing passives or hand-rework. Most useful for power supplies and motor drives, where counterfeit practice has historically focused on swapping the main switching device and a handful of passives. It does not prove the part will run in your application, only that the parts on the board are what they claim to be.

The order above is not a hierarchy of merit; it is a menu of questions. Each method answers a different one. "Does the label look right?" — visual. "Does it power up and respond?" — bench. "Is the silicon inside the package what the datasheet describes?" — X-ray or decapsulation. "Are the parts on the board the parts the reference design calls for?" — teardown. Buyers should verify against the original manufacturer datasheet and run the result through their own qualification process before drawing any operational conclusion.

What each method does not prove

This is the section most verification reports leave out. Bench testing a refurbished I/O module tells you the I/O module worked on the test rig for the duration of the test. It does not tell you the I/O module will work inside your program, in your cabinet, under your ambient conditions, for the next ten thousand operating hours. X-ray tells you the die is present and the bond wires are intact. It does not tell you the part will pass a thermal cycle. Decapsulation tells you the silicon matches the datasheet photograph. It does not tell you the lot was not pulled from a production line running outside the manufacturer's quality system. Teardown tells you the BOM matches the reference. It does not tell you the reference design itself is appropriate for your load profile.

Buyers who treat a single verification step as the answer to all of those questions tend to be the ones whose "verified" parts later fail. The right mental model is that verification narrows the risk; it does not eliminate it.

A case where verification is the wrong tool entirely: when the buyer's own documentation pins the design to a specific part number and the change would alter the documented bill of materials, or when the buyer's quality system requires traceability that no third-party report can supply. In those situations the right answer is to keep sourcing the original — or to treat the change as a redesign rather than a replacement. No third-party lab report stands in for the buyer's own engineering and quality records, and the desk will decline to be the one to recommend the change.

Where the catalog anchors sit

For Phoenix Contact specifically — a brand whose parts pass through our catalogue as both current production and as surplus or refurbished stock — the same ladder applies with a few method-specific notes. Phoenix Contact's labelling uses laser-etched text on the housing rather than paper labels for most I/O modules, which makes a re-mark attempt easier to spot under raking light but easier to overlook in a flat photograph. The QUINT4-PS/3AC/24DC/20 series of power supplies, with diagnostic contact and signal output, exposes enough on a test rig to bench-validate a healthy unit, but a unit that has been powered up without its rated input filtering can pass the bench test and still fail under load. Inline terminal blocks such as the PR family and the UMK-2 FLK26/SO384 ribbon-cable adaptor are usually verified visually and by continuity rather than by powered test. Inline I/O such as the IB IL 24 SDI 8-PAC, being a bus node, is verified by bench with the upstream coupler present — a stand-alone power test tells you little. Those are the kinds of detail that decide which verification step is worth commissioning and which is theatre.

The condition on each of those catalog entries is quoted per line — new surplus, refurbished, or used, with the test record attached where one exists. Where a buyer wants the bench-test record, the unit ships with the rig printout; where the buyer wants a third-party X-ray or decapsulation, the original lab report ships with the unit. The desk does not summarise, paraphrase or "approve" the lab's findings; the original report is the deliverable.

Where this leads for a buyer

The decision flow the desk walks through with a buyer usually goes: name the failure mode you are guarding against, then choose the cheapest method that would catch it. If the answer is "I am worried the label is wrong," commission the visual record and stop there. If the answer is "I am worried the silicon inside is a clone," commission X-ray or decapsulation and accept that the part is destroyed in the process. If the answer is "I need to know whether it works in my machine," commission the bench test, then commission the integration test on your own rig, because no external verification can replicate that.

Two situations the desk will not take on: requests to certify that a substitute is compatible with a specific machine or program (that judgement belongs inside the buyer's engineering process) and requests to issue a verification result the desk did not actually commission and observe. What the desk can do is arrange the third-party lab work, hand the buyer the original report, and accompany it with the photos and bench record taken at receipt. The verification chain stops at "here is what we observed." Whether that is enough for the buyer's application is the buyer's call.

The next step is to send the part number — or, where the part number is not known, the marking photograph and the machine the part was pulled from — to the desk's inquiry inbox. The quote that comes back states condition, test record and warranty per line; the verification step is added on top of that quote once the buyer has named the failure mode they are guarding against.

Service limits, named up front

We are an independent industrial automation distributor and sourcing desk, not an authorized distributor for any manufacturer. We hold no manufacturer authorization, no manufacturer-issued warranty, no ISO 13485 or EAC or TR CU certification, and no testing-laboratory accreditation of our own. We commission third-party lab work and we pass the report to the buyer with whatever scope the lab and the buyer agreed in writing. We screen end users and end uses, classify before quoting, and decline transactions that cannot be screened. We do not certify compatibility between modules, we do not certify compliance with any functional-safety or regulatory regime, and we do not certify that a verification report from a third-party lab is sufficient for any particular standard or audit. The verification step answers a specific question; the buyer decides whether the answer is sufficient. For industrial automation modules — IP rating, ambient temperature, contamination level, vibration and EMC — those constraints are the buyer's engineering decision, not the desk's.

FAQ

What is the cheapest verification that catches most counterfeit modules?

Visual and marking inspection, paired with a date-code and lot-trace check against a known-good reference, catches the largest single class of misrepresented modules in the secondary market. It costs the least and turns around fastest, and it stops most attempts where a similar-looking part has been re-labelled. It does not catch every form of misrepresentation, so it is usually paired with at least one powered step.

Does a bench test prove a refurbished module will work in my machine?

No. A bench test on a documented test rig confirms that the module powered up, exercised its I/O against the manufacturer's specification, and was observed for the duration of the test. Whether it will work inside your program, your cabinet, your ambient conditions and your operating cycle is a separate question, and that question is answered by your own integration test, not by the supplier's bench record.

When is X-ray worth commissioning for an automation module?

X-ray is most useful when visual inspection has flagged an ambiguous label or when the buyer's failure mode is a relabelled module with a clone or remark-grade silicon inside. It images die presence, package authenticity and bond-wire layout non-destructively, and it is the step that catches what visual inspection cannot. It does not prove the silicon will work, only that the silicon inside the package matches the photograph the buyer's reference library expects to see.

Is decapsulation standard for industrial automation parts?

No. Decapsulation is destructive, slow and expensive, so it is reserved for high-value parts where the consequence of being wrong exceeds the cost of the test. It is most often commissioned for lot authentication on drivers, CPUs and other high-value silicon where a counterfeit's failure mode is too costly to leave unverified. For most I/O and signal-conditioning modules, bench testing and visual inspection cover what the buyer needs.

Can a supplier certify that a verification is sufficient for my regulatory regime?

No. Whether a verification report is sufficient for a buyer's regulatory regime is a determination that sits with the buyer's compliance and engineering functions, not with the supplier. The supplier's role stops at having commissioned the test, observed the result and passed the original report to the buyer. The buyer then evaluates whether that report covers what their own framework requires.

What is the difference between bench testing a used module and bench testing a refurbished one?

A used module is sold as pulled from service, and the bench test is descriptive — it records what the unit did on the rig on the day. A refurbished module is sold with a warranty, and the bench test is part of the evidence that the unit was returned to working order before it left the refurbishment line. The two records look similar in shape but carry different commitments, and the buyer's tolerance for what comes next is set by which one is in hand.

Can I send you a batch I bought on the Shenzhen market for inspection and lab work?

Yes. Where the buyer has sourced a batch themselves, the desk can act as their agent on the ground — receive the goods in Shenzhen, carry out visual and incoming inspection, and arrange third-party laboratory testing such as X-ray, decapsulation or electrical measurement. Scope, method and fee are agreed in writing before the work starts, and the original report is passed back to the buyer. The desk does not operate a certified laboratory of its own and does not issue findings it did not observe.

Data Notes

By the AO Control sourcing desk. First published September 23, 2026.
Independent industrial automation distributor and sourcing desk operating from China; not an authorized distributor for any manufacturer; not a certified testing laboratory.
Reference parts named in the body are real catalog entries: Phoenix Contact PR, QUINT4-PS/3AC/24DC/20, EMG 45-DIO 8E/LP, UMK-2 FLK26/SO384 and IB IL 24 SDI 8-PAC. Their current condition, price and lead time are quoted per line on request and are not stated in this article.
Verification method descriptions are drawn from general engineering practice for industrial automation electronics; the desk commissions third-party labs and passes their original reports back to the buyer without modification.
Industrial-environment constraints (IP rating, ambient temperature, contamination level, vibration, EMC) are the buyer's engineering decision, not the desk's; verification proves identity and function on the test rig, not fitness for the buyer's specific installation.
Service-limit language on screening, certification and laboratory ownership mirrors the site's published quality and terms-of-sale pages.
Send a part number or a marking photograph to the inquiry inbox to start the conversation; condition, test status and warranty are quoted per line.