Cable Assembly Samples in 7–10 Days: What to Do With the First Article

Cable assembly sample approval is the one gate in a China harness programme where a buyer holds real bargaining power and usually spends it badly. The box arrives, three leads sit in a poly bag with no paperwork, somebody plugs one into a test rig, it works, and an email goes out saying "samples approved, please proceed". That email has just approved an object. It has not approved a specification, a process, or anything a factory can be held to when the first thousand pieces land.

First article inspection on a harness is worth doing properly precisely because a cable assembly hides its own quality. The crimp that matters is inside a housing. The copper that matters is inside insulation. Nothing you can see from the outside distinguishes a lead that will survive two years of vibration from one that will not. What you actually check, and what you sign, are the two decisions that determine whether the production run resembles the sample.

Before the samples ship: the clock and the quantity

What "dispatched in 7–10 days" is a clock for

Stated sampling terms at this factory are 7–10 days to dispatch, with volume orders at 15–25 days to warehouse. Read the first figure carefully: it is a dispatch clock, not a delivery clock, and it starts when the specification is frozen, not when you first make contact.

Three things sit outside it and are worth confirming before you build a launch schedule on the number.

Connector procurement is the big one. If your bill of materials names housings, terminals or seals the factory does not hold, the sample cannot start until those arrive, and connector lead time is routinely longer than harness assembly time. Second, drawing clarification: every round of "which datum do you mean" adds days that nobody counts as sampling time. Third, your own approval loop — the sample sits in your goods-in for a week more often than the factory misses its dispatch window.

The practical move is to freeze the drawing revision and the connector BOM before you ask for the clock to start, which is the same discipline described in what goes in a wire harness RFQ. Then ask for a specific number of pieces rather than "a sample".

Ask for three pieces, not one

One sample is enough to look at and not enough to learn from. Three is the working minimum, and five is better if the piece price is low.

• One to destroy. Cross-sectioning a crimp and pull-testing a terminal are destructive. If you only have one sample, you will not do them, and those are the checks that predict field failure.
• One to retain, untouched. This becomes your reference — the golden sample you compare production against six months from now. It should be sealed, labelled with part number, drawing revision, date and a serial, and photographed from both faces and from the cavity side of each connector before it goes in the drawer.
• One to fit. Install it in the real assembly, in the real enclosure, with the real routing. Branch lengths that measure correctly on a bench can still fight the actual routing path, and this is the only check that catches it.

Ask, too, what the samples cost and whether that charge is credited against the first production order. Both are ordinary commercial terms and neither is published; get them in writing with the sample quotation rather than discovering them on the invoice.

The first hour: what to check before you approve anything

Work through this before anything gets plugged in. Every line takes under five minutes and each one has caught a real defect.

Two lines there deserve expanding. The terminal-seating tug is the highest-yield thirty seconds in the whole process: a terminal that is inserted but not fully latched will pass a continuity test on the bench, sit happily through your evaluation, and back out under vibration months later. And the marking check matters because printed circuit identification is applied at very different qualities — one that survives a thumb survives an installer.

The sample you have to destroy

The remaining checks require cutting something up, and they are the ones that separate a supplier who controls the crimp from one who merely produces crimps.

Cross-section

A crimped terminal sectioned across the wire barrel, polished and etched, shows what no external inspection can: how far the conductor has been compacted, whether every strand is inside the barrel, whether the barrel has folded correctly around the conductor, and whether the insulation barrel grips insulation rather than copper. It is the definitive evidence, and the full interpretation is covered in how harness crimps are actually judged.

What matters at sample stage is provenance. Ask for a cross-section image taken from your own sample lot, with the part number and date on it, rather than a reference image from a supplier's library. A factory with a terminal cross-section analyser on its bench can produce one; whether it does so as standard with samples is not published, so ask.

Pull-off force

Pull a crimped terminal off the wire on a tensile tester and record the force. The absolute number depends on conductor size and terminal design, so what you want is not a single figure but a stated acceptance value and evidence the sample met it. Ask which value the supplier applies for your conductor size and where it comes from — a terminal manufacturer's published data, an internal standard, or a customer drawing. Any of the three is a legitimate answer. "It passed" is not.

If you cannot do this yourself

Most buyers cannot section a terminal in-house, and that is fine. Two options work. Ask the supplier to perform the destructive checks on your sample lot and send images plus recorded values, or send one sample to a third-party laboratory in your own market. The first is faster and cheaper; the second is independent. On a first order with a new supplier, the combination of supplier evidence plus one independent check is a proportionate spend, and it is the same logic that makes pre-shipment inspection worth budgeting for later in the programme.

The question that actually predicts production quality

Here is the one to write into the email: was this sample built on the same equipment and by the same method as production will be?

Harness samples are frequently built by an engineer or a senior operator using benchtop crimpers, because setting up an automatic cut-strip-crimp line for three pieces makes no sense. That sample can be excellent. It also tells you very little about a thousand pieces coming off a servo crimping press with a different applicator and a different operator. This is the mechanism behind almost every "sample was perfect, production was not" story in cable assembly — not deception, just two different processes producing two different outcomes.

Ask three follow-ups, and put the answers in the approval record:

1. Which route built this sample — benchtop crimper or automatic line?
2. Which route will build the production order at my quantity?
3. If they differ, what evidence will you provide from the production route before shipment — a first-off cross-section, a crimp-height record, a pull-force record?

A supplier that answers those three cleanly has told you more about its process control than any document will. The same question governs how a programme moves from sample to mass production without a quality cliff in between, and it becomes more pointed once quantities are set — a family structured to share one setup, as discussed in planning a SKU mix around a 1,000-piece minimum, should also share one production route.

Common questions

What exactly should I sign?

A part number, a drawing revision, a date, and the serial of the sample you retained — not the word "samples". Write the approval as: "Part 12345, drawing rev C dated 2026-08-14, sample serial 002 retained by buyer, approved for production." That sentence is enforceable. "Samples look good, please proceed" is not, because there is no way afterwards to establish which geometry and which bill of materials were approved.

Should the factory keep a counter-sample?

Ideally both sides retain one, sealed and labelled to the same revision, so a dispute six months from now compares two controlled objects rather than two memories. Whether a supplier retains counter-samples, for how long, and how it re-issues them when a revision changes is worth asking about explicitly — it is not something most factories publish, and the answer tells you how mature the quality system is regardless of what paperwork accompanies it.

Does a passing continuity test on the sample mean much?

It means the sample is wired correctly, which is worth knowing and not the same as knowing production will be. The useful questions are about the production regime: is electrical test applied to every piece or to a sample per lot, what fixture is used, what the pass criteria are for withstand-voltage and insulation resistance, and whether records are retained per shipment. A bench that lists continuity, withstand-voltage and insulation testers can support all of that — what you need is the supplier's commitment on which of it runs on your order.

What to do before you sign off

1. Order three to five samples, and say what each is for so the supplier knows one will be destroyed.
2. Freeze the drawing revision and connector BOM before asking for the sampling clock to start.
3. Confirm sample cost and whether it is credited against the first production order.
4. Run the first-hour checklist above, including the terminal tug and the cavity-by-cavity comparison against the wire list.
5. Request a cross-section image and a pull-force value from your own sample lot, not a reference image.
6. Ask which route built the sample and which will build production, and what evidence comes from the production route.
7. Fit one sample into the real assembly before approving — bench measurements do not catch routing conflicts.
8. Retain, seal, serialise and photograph one untouched sample, and ask whether the factory retains a counter-sample.
9. Write the approval as part number, revision, date and sample serial.
10. On any certificate or test report presented alongside the samples, check the issuing body, the exact holder name and the validity date on the document itself.

Huisheng Electronics states samples dispatched in 7–10 days and volume at 15–25 days to warehouse against a 1,000-piece minimum per item, with a supplier-published equipment roster that lists automatic cut-strip-crimp machines, servo crimping presses and benchtop crimpers on the production side, and terminal cross-section analysis, tensile and insertion-force testing, withstand-voltage and insulation checks, salt-spray and cable continuity on the test side. That combination is what makes the questions above answerable: a shop that lists a cross-section analyser can be asked for a section from your lot, and a shop that lists both automatic and benchtop crimping can be asked which one builds your order. Ask to see the crimping and test benches live on a video call while your sample is on them — the general approach to running a sample order with a Chinese factory applies, with the harness-specific addition that you want to watch the crimp, not the finished lead.