Choose a physical sealed golden sample as your reference if you need defensible bulk approvals; choose photo-based approval only if you accept that lot two is a coin flip. This page gives wholesale buyers a repeatable observation log for bubble machines — setup, refill, leakage, and power — with checkpoints and supplier questions at each step, and no invented pass limits, AQL numbers, or defect rates.
The point is not to score the sample. The point is to record what you actually see, tie every line to a sample ID and a golden-sample reference, and turn every missing threshold into a written question on the PO. That log is what protects you when the bulk lot lands.
Key Takeaways
- Approve against a sealed physical golden sample on the QC shelf, not a photo. Compare bulk to that piece on colour, fit, flash and function.
- Setup, refill, leakage and power each get their own log section. If a value is missing, write the buyer's question, not a number. No supplied document sets a universal pass limit for bubble output, run time, or leak volume.
- Major and minor defect limits must be stated per lot or inspection plan; a factory's default percentages are not automatically your acceptance criteria. The AQL comes from the agreed inspection plan on the PO, not from a catalogue sentence.
- Keep paperwork separated by market, lot and file. A UKCA mark and a VOC file on an electric bubble machine are different papers covering different tests — paperwork that does not match the lot is not paperwork.
- Name the Incoterms version and the named place as actually booked. FOB without a named place turns into a freight argument that outlasts production.
Sample Check: What to Verify at Each Step
| Step 1 — Freeze sample identity | What to check: sample ID, SKU mapping, sealed golden-sample reference on the QC shelf, date, observer name. Red flag: approval based on a photo or a chat message with no physical reference piece. |
|---|---|
| Step 2 — Setup and assembly | What to check: assemble as received; record parts present/absent; note fit and flash against the golden sample. Red flag: seller says 'assembly is obvious' and ships no parts list. |
| Step 3 — Refill cycle | What to check: how the reservoir opens, fills and closes; whether the process matches the golden sample; fluid type stated on the pack. Red flag: reservoir cannot be opened and flushed without tools or force. |
| Step 4 — Leakage observation | What to check: where, when and under which condition leakage appears — upright, tilted, after refill, after handling. Red flag: any leak claimed as 'normal for this price'. |
| Step 5 — Power and switch behaviour | What to check: power source type, switch behaviour, whether operation matches the golden sample; plug and cord condition. Red flag: no stated power source or charging arrangement in writing. |
| Step 6 — Defect handling on a failed lot | What to check: isolate the lot, record defect class, release only after rework; defect limits per the agreed inspection plan. Red flag: a factory quoting its own default percentages as your acceptance criteria. |
| Step 7 — Paperwork match | What to check: which file covers which test, for which market, for which lot. Red flag: one certificate presented as covering a different market or a different test scope. |
| Step 8 — Shipping terms | What to check: Incoterms version and named place as actually booked (e.g. Shantou or Shenzhen). Red flag: 'FOB China' with no named port. |
| Step 9 — Tracking label and CPC | What to check: manufacturer, date and lot on product or pack; CPC matched to the SKU. Red flag: a tracking label with a barcode and no manufacturer, date or lot content. |
What does the observation log actually need to contain?
Every observation line must name the sample, the SKU and the golden-sample reference. A line that says 'leaks a bit' is unusable three months later; a line that says 'unit tilted 30 degrees ten seconds after refill, fluid at the seam under the reservoir cap, golden sample does not show this' is usable by anyone on the team.
Log the four categories separately — setup, refill, leakage, power — and keep an explicit 'not documented' column. That column is the most valuable part of the sheet, because it becomes your RFQ question list. A process-capability study quantifies variation against specified limits and is more informative than checking one sample against a nominal value, so treat single-sample readings as observations, never as capability data.
Worked example (illustrative, not a real shipment): one lot of about 600 units across two destinations — a US-bound carton and an EU-bound carton with different pack artwork — and two pack variants, a plain colour box and a window box. The observer logs setup and refill on three units, finds fluid residue under the cap on two of them, and records it as an observation with no pass limit. The step that stops the shipment is not the leak itself; it is that the EU carton's artwork file does not match the file mapped to that lot in the paperwork log. That mismatch holds the lot until the file, the market and the lot are reconciled.
How do you run the setup and refill checks without a benchmark?
Assemble the unit exactly as received, with no tools the buyer would not have. Record parts present and absent, then note fit and flash against the golden sample. Flash is the thin ridge of plastic where mould halves meet — the fastest visual tell that a tool has changed or that tooling is wearing, and the kind of drift a photo approval never catches.
Run one refill cycle and narrate it: how the reservoir is opened, how it is filled, how it closes, and whether fluid can be emptied and flushed. If the answer to 'can the reservoir be emptied and flushed' is no, that is a maintenance and mould-risk problem for the end user, not a cosmetic one.
Ask the supplier which physical golden sample is the reference for this SKU and whether it can be sealed and held on the QC shelf. Without a sealed reference you are comparing memories, not samples.
How should leakage and power observations be recorded?
For leakage, record the condition, not a verdict: upright, tilted, after refill, after handling, at rest. Note the location — seam, cap, base, nozzle — and whether the golden sample behaves the same way. Do not write 'passes leak test' unless a dated primary-source document supplies the test method and the limit; none was supplied for this article, so the honest entry is the observation plus the question.
For power, record the power source type, switch behaviour, cord and plug condition, and whether operation matches the golden sample. Note the charging arrangement for rechargeable units and whether the unit was supplied with the correct adapter for the destination market. Do not state run time, output or bubble count unless a dated document supplies the figure — those numbers are vendor claims until verified.
If the unit is intended for indoor venue use, moisture near sockets and cord routing are worth a line in the log. Certifications such as CE, UL or ETL are labels to record and verify; they are not guarantees of indoor safety or of leak-free operation.
What paperwork has to match the lot before you approve?
Map market, lot and file before you approve anything. A UKCA mark and a VOC file on an electric bubble machine are different papers covering different tests. Write down which file covers which test for which market and which lot, and treat any file that cannot be tied to the lot as missing rather than present.
For US children's product, the tracking label is not decoration: manufacturer, date and lot go on the product or the pack as required, and the CPC should be matched to the SKU, not to the factory door. For EU-bound goods, EN 71-2 covers flammability requirements for toys, and the digital product passport under the EU toys framework must state that compliance with the Regulation's requirements, including essential safety requirements, has been demonstrated. The digital product passport, GPSR and the toy safety directive are separate regimes — do not treat them as one deadline.
California Proposition 65 can require warnings for listed chemicals even when a toy already meets U.S. product safety documentation limits. It is a state right-to-know law, not a federal toy standard, so it belongs in your destination checklist, not in your compliance certificate folder.
On timing: where the official text does not confirm the date on which a requirement actually applies to this product category, record it as 'timeline not confirmed from official text' rather than writing a mandatory date. Entered into force, actually applies to the category, and uncertain are three different states.
What to ask suppliers
Which physical golden sample is the reference for this SKU, and can it be sealed and held on the QC shelf?
Which document covers which test, for which market and which lot — and can you send the file list rather than a single certificate?
What is the agreed inspection plan on the PO, and what defect limits does it specify for major and minor classes?
What Incoterms version and named place apply to this booking?
How is the reservoir emptied and flushed, and what fluid type is specified on the pack?
What power source, charging arrangement and destination-market plug are supplied?
What tracking-label content — manufacturer, date, lot — is applied, and how is the CPC matched to the SKU?
If tooling is involved: what is the tooling fee, who owns the tool, what is the storage period, and what happens if the tool sits idle?
FAQ
Do I need an AQL number to check bubble machine samples?
No — not at sample stage. Defect limits come from the agreed inspection plan on the PO, and a factory's default percentages are not automatically your acceptance criteria. Use the sample stage to record observations and to write down the questions that the PO will eventually answer.
How many units should I test from a sample order?
Enough to see setup, refill, leakage and power on more than one unit, and enough to include every pack variant you plan to buy. The exact count depends on your inspection plan and the sample quantity the supplier releases, so confirm it in writing rather than assuming a standard number.
Can I approve a bubble machine from photos and a video?
Only if you accept the risk on lot two. Photos do not resolve colour, fit, flash or function against the bulk lot; a sealed physical golden sample on the QC shelf does. If you must approve remotely, ask for the sealed sample to be couriered and keep it as the reference.
What certifications should a bubble machine carry for the US and EU?
It depends on the market and the SKU, not on a single document. For the EU, EN 71-2 covers flammability for toys and EN 71-3 covers heavy-metal migration; for the US, ASTM F963 and U.S. product safety documentation limits apply, with California Proposition 65 as a separate state right-to-know law. Ask for the file list mapped to market, lot and test — and confirm which requirements actually apply to your product category, because the applicable date is not always confirmed in the official text.
What are the biggest red flags when sourcing bubble machines wholesale?
A factory quoting its own default defect percentages as your acceptance criteria; one certificate offered as covering two markets or two tests; 'FOB China' with no named port; and a tracking label with a barcode but no manufacturer, date or lot content. Each of these is a paperwork failure that shows up as a shipment failure.
Should I ask for a process-capability study on the moulded parts?
Yes, if the parts are injection moulded and your volumes justify it. A process-capability study quantifies variation against specified limits and is more informative than checking one sample against a nominal value — it tells you what the tool does over a run, not what one unit looked like on a Tuesday.
Sources
Related Products & Categories
Ready to spec your next bubble machine order?
If you are building an observation log and want to compare it against real sample specs, start with our bubble toys range at /toys/bubble-toys/ and tell us which SKUs you want to log. Bring your golden-sample requirement, your destination market and your inspection plan; we will answer the paperwork questions in writing rather than in a photo.