Одобряйте предпроизводственный образец, только если он изготовлен в тех же условиях, что и ваш опт: та же восковая ёлка или полость стальной формы, тот же сплав и партия литья, та же ванна покрытия, тот же комплект муассанита, та же последовательность пайки, обточки и полировки, те же замки и цепочки. Изделие из образцовой комнаты, собранное техником, покрытое в верстачной ванне и с камнями, подобранными вручную, доказывает лишь то, что дизайн можно сделать, — а не то, что линия его повторит. Правильный шаг — запросить образец производственной линии: первое изделие с реальной оптовой партии, измеренное, сфотографированное, опечатанное как эталон хранения и согласованное письменно. Если фабрика не может дать такой образец до финализации оснастки — не одобряйте ничего.
Чем изделие из образцовой комнаты отличается от производственного?
A sample room is built for speed and control. One technician files the piece, sets the stones one at a time, solders at a bench torch and walks it to a small plating bath. Production is built for repetition: dozens of pieces on a wax tree, a shared flask, a setters’ line, a rack in a larger bath, and a barrel polisher that touches every surface at once.
In silver jewelry, that difference is not cosmetic. Casting shrinkage changes with tree position. Plating thickness changes with rack position and current density. Solder joints that look tight on a bench piece can open when a batch is re-flowed, and tumble polishing removes metal and softens a stamp that was cut cleanly by hand.
So treat the two samples as different documents. A sample-room piece answers “can this design be made?” A production-run sample answers “will this line make it the same way ten thousand times?” Only the second question belongs in a purchase order.
Какие шесть жёстких показателей обязан доказать предпроизводственный образец?
Generic checklists stop at dimensions, color and stone count. For S925 moissanite jewelry, six process indicators decide whether the approved sample and the bulk order are the same product.
1. Same Mold, Same Cavity, Same Casting Lot
Ask for the mold number and the cavity number, not just “the same mold.” In a multi-cavity steel mold, cavity-to-cavity weight variation is the defect no photograph will show you. For a custom piece, the wax print file and wax batch must match the approved wax, and the sample must be cast with the same sprue position and tree density planned for bulk. The casting lot matters as much as the mold: alloy grain lot, flask temperature and burnout cycle all move porosity. Request the work order and alloy lot, weigh the piece against the CAD, and inspect the shank and stone seats at 10x for pits.
2. Same Plating Bath and a Measured Coating Thickness
A bench bath and a production bath are different even with the same chemistry sheet. Temperature, drag-out, filtration, current density and rack position all shift the deposit. Ask which bath ID plated the sample, on what date, and where the piece sat on the rack, then ask for thickness in microns with the method named.
Handheld XRF is a screen, not a verdict: its beam reaches only a few microns and can mix coating and substrate signals. A mounted cross-section read by XRF line scan or SEM-EDS is the defensible number. If the finish is vermeil, the floor is regulatory, not promotional: 16 CFR 23.4 defines vermeil as a base of sterling silver coated or plated on all significant surfaces with gold, or gold alloy of at least 10 karat, at a minimum thickness throughout equivalent to 2.5 microns of fine gold. That 2.5 microns is measured as fine gold, not as the alloy, so a 10K coating must be about 6 microns thick to reach the floor. Amara Jeweiry publishes ≥2.5 µm for its 18K vermeil and rhodium finishes, nickel-free. Check the clasp tongue, jump rings, inner band and prong edges, where coverage thins first.
3. Same Moissanite Parcel and Setting Parameters
“D/VVS1” is a grade, not a parcel. Two parcels can both be sold as D/VVS1 and still differ in tint, table size and how a pavé row sparkles. Ask for the parcel reference that filled the sample, the GRA certificate numbers that came with it, and written confirmation that bulk will be filled from the same parcel or one matched to it before production starts.
Setting is a process, not a look. Ask for seat depth, prong contact rules and stone position tolerance in millimeters. Amara Jeweiry publishes a ±0.05 mm stone position tolerance and a 40x micro-pavé inspection. On the sample, count stones against the CAD, rock every stone with a needle, and read the girdle inscription at 10x against the certificate number.
4. Same Soldering, Filing and Polishing Sequence
In a sample room, one pair of hands solders or laser-welds, files and polishes. In production, those are three stations, often three people, and the piece may be tumbled with hundreds of others. The sequence matters: solder before or after setting, mark before or after plating, polish before or after assembly.
Ask for the sample’s routing sheet and compare it with the bulk routing. Look at the clasp seam, bail and every link joint under 10x. Confirm that file marks were removed rather than plated over, and that the tumble did not round the prong tips or loosen a stone.
5. S925 Stamp Position, Depth and Legibility
The stamp is the one feature a customer can use to challenge you later. 16 CFR 23.10 states that a quality mark should be legible to a person of normal vision, placed where a purchaser is likely to see it, and should disclose the part of the piece it applies to when it does not cover the whole product.
Check three things. Position: beside the brand mark, not across the band, with space reserved for UK or EU assay marks if those markets are in scope. Depth: enough to survive polishing and wear, never cut over a solder joint or stone seat. Legibility: judge it at 10x before plating and again after plating, because plating can fill the characters, while post-plating laser marking can open the gold layer inside the letterforms on vermeil. Ask whether the 925 is cast in the mold or laser-etched, then lock the method and character height into the approval. Amara Jeweiry’s guide to laser etching a 925 hallmark covers the placement trade-offs in more detail.
6. Clasp and Chain Tensile Strength
The clasp, jump ring and end caps fail first. “Pull tested” with no number is not a specification. Agree the force in newtons, the pull direction, the sample size and the pass criterion, then ask for the record on the pre-production sample and on the first bulk lot. Test a sacrificial piece to failure and record where it opened: spring, jump ring solder, end cap crimp or chain link. Amara Jeweiry states that pull testing covers 100% of pieces before plating and again before packing, but the force still has to be written into your purchase order.
Как испытать образец и его не уничтожить?
Split the sample into a decision piece and a record piece. Keep one untouched, photograph it beside the CAD with a scale, seal it and co-sign it as the retention standard. Use the second for tests that consume material: a cross-section for plating, a pull to failure for the clasp, and a stone removal if a setting is disputed. One sample cannot be both the standard and the test piece.
- Measure, do not admire: grams, stone count, character height, microns, newtons.
- Match magnification to the feature: 10x for the stamp and girdle inscription, 40x for pavé seats.
- Name the method: XRF screen or cross-section, salt spray or thioacetamide, bench pull or rig pull.
- Log the lot: mold and cavity, alloy lot, casting lot, bath ID, stone parcel, work order.
Что должен содержать документ согласования образца?
One page, signed by both sides, with a photograph of the approved piece and every number the bulk order will be judged against: SKU and revision, mold and cavity, alloy and casting lot, plating bath ID with thickness and method, stone parcel with certificate numbers and setting tolerance, stamp position and character height, clasp test force and result, polishing route, AQL level, and the sealed retention sample ID. Reject “same as sample” as a specification; it is the sentence that ends every dispute with no evidence. A measured sheet plus a sealed reference is what survives a claim three months later, which is the same logic Amara Jeweiry applies in its guide to plating defect claims.
Когда изделие из образцовой комнаты ещё приемлемо?
Three cases. Before tooling exists, a sample-room piece is a design proof, and it should be labeled that way: design sample, not production approval. With stock molds, the piece may legitimately come from the actual mold and the actual plating line, so verify the mold number and bath ID before treating it as a pre-production sample. In a small pilot, pull one to three first articles from the first bulk lot and approve those, keeping the sample-room piece as reference only.
What is not acceptable is approving a hand-built, bench-plated piece as the standard for a production run and then discovering the difference in the first shipment.
Часто задаваемые вопросы
Is a “golden sample” the same as a production-run sample?
Not necessarily. Golden sample is a generic term for the reference piece a line is measured against, and it can be either a sample-room build or a first article. What gives it authority is the paperwork: mold, lot, bath and parcel references. A golden sample without lot references is a nice piece with no authority. Ask for the production-run version and seal it.
How many pre-production samples should I approve?
Two per SKU and finish: one sealed as the retention standard, one for destructive testing. For multi-plating or multi-color orders, two per colorway. Approving a single piece and testing it leaves nothing against which to compare the bulk.
What if the mold is the same but the casting lot differs?
Treat it as a different sample. A new alloy lot, flask load or burnout cycle changes porosity, shrinkage and final weight. Ask for a piece from the bulk casting lot before you approve, or approve a pilot only and re-approve the first article taken from the production tree.
Can I check plating thickness without cutting the piece?
You can screen it with handheld XRF, but the beam reaches only a few microns and can mix coating and substrate signals. A mounted cross-section read by XRF line scan or SEM-EDS is the number you can hold a supplier to. Keep the screen for receiving checks and the cross-section for claims.
Should the S925 stamp go on before or after plating?
Either can work, but the choice must be written down. Marking before plating can leave characters partly filled; laser marking after plating can open the gold layer inside the letterforms on vermeil. Judge legibility at 10x in the finished state, and reserve space for assay marks if you sell in the UK or EU.
Каков итог?
Insist on a sample pulled from the actual production run, and treat the sample-room piece as a design proof. If a supplier cannot provide a first article from the bulk lot, ask for a pilot of one to three pieces from that lot and approve those instead. Amara Jeweiry publishes a 7–10 day sample lead time after CAD sign-off and states that samples and bulk run on the same line with the same checks; its custom moissanite jewelry process и FAQ set out the wax approval, inspection and warranty terms described above. Hold every supplier to the same six indicators, and put the numbers in writing.
