Two reservoirs make a device more interesting to sell, but they also give a buyer two filling paths, two activation paths, and more ways for a sample to differ from the production unit. A promising first draw is not enough to approve a bulk order. The useful question is whether your team can repeat the same fill, switching, and storage results across identified units of the configuration you intend to buy.
Start with the specific model you plan to order. The empty dual-chamber AIO hardware collection includes models with different capacities and controls. Your checklist should follow the sample revision and its documented modes; do not assume that every device has a screen, a combined mode, or the same access to both chambers.
Lock the configuration before testing
For each sample, record the model and revision, supplier lot, chamber capacity, control method, firmware or interface version where relevant, closure type, and the two formulation codes. Keep the fill target and post-fill conditioning (rest time and conditions) with that record. Otherwise, a hardware or liquid change can quietly explain a later pass or fail, and the record will give no way to tell.
Ask the supplier for the current filling and closing instructions for that configuration. Check the physical sample against them before adding oil. If a specification, opening, or closure step is unclear, resolve it before setting pass criteria. Use the product-page description to frame your questions. How a particular production lot performs still has to be shown on the bench.
Treat the two fills as separate operations
Identify the left and right fill paths before dispensing. Record the mass added to each chamber, the visible headspace, the time between filling and closing, and any residue around the openings. Use the intended production formulation and process conditions; a convenient substitute oil can hide a fit or flow problem that matters later. After closure, photograph each unit from consistent angles and label the chamber assignments.
For storage observations, keep samples upright and in the other storage orientations relevant to your distribution plan. Record where liquid appears, when it appears, and whether it came from a fill opening, mouthpiece area, exterior seam, or a possible path between chambers. A visible mark is evidence to investigate; on its own, it does not establish cross-chamber migration. If the path cannot be identified, hold the unit for follow-up rather than recording a clean pass.
Check what the selector actually activates
Write down every mode the model under test is supposed to support. For a left/right device, run each side separately, noting the selected indication, activation response, and any vapor or taste from the chamber that should be idle. Test a combined mode only if the model offers one. Repeat switching rather than judging a single successful cycle, and preserve unit-level notes when the display, switch position, and observed response disagree.
Keep electrical behavior and liquid delivery separate in the log. A weak draw may come from a mode-selection issue, power delivery, airflow or intake behavior, or the filled formulation. Calling all four “switching failures” makes the supplier conversation less useful. There is no universal acceptable output number to borrow here. Agree on a model-specific test method and acceptance window with the supplier before production.
Decide pass, hold, or retest before seeing the results
Write the decision rules into the test sheet first. Define which observations stop sample approval, which can be retested after a documented process correction, and which require supplier engineering review. Include the number of units, lots, orientations, storage time, switching cycles, and who signs off. Each of these is a project decision; no universal industry threshold settles them for you.
Keep photographs, unit IDs, fill records, and the supplier’s response together. If a sample is modified during troubleshooting, mark it as a new configuration and retest the affected checks. A changed seal, firmware setting, or fill method should not inherit the earlier sample’s approval.
Once the configuration passes your agreed criteria, carry the same controls into the pilot-run and mass-production readiness checks. That handoff turns a good bench sample into evidence you can hold production to. Ask for the exact approved revision in the quotation and purchase order, along with the sample record that both sides signed off on.
A compact sample record to request
The record should let a different technician reconstruct the decision:
- Identity: unit and lot ID, hardware model and revision.
- Fill: left and right formulation codes, fill masses, and closure method.
- Function: each supported mode and observed response.
- Storage: orientation, time, leak or migration observations, and photos.
- Disposition: pass/hold/retest reason and the person who made the call.
If those fields are missing, the next batch will be difficult to compare with the sample you liked.
