In This Guide
- 1. Separate visible level from quantity
- 2. Keep the run log and individual readings together
- 3. Compare heads before choosing a cause
- 4. Compare the trend with feed and refill events
- 5. Preserve the baseline and agree the next check
- 6. Download the log and send a focused support record
- Customer questions
- Product reference and related guides

Part of Filling Machines & Product Trials.
A changing bottle level is an observation, not yet a dosing diagnosis. First distinguish visible level from measured quantity, then retain head identity, sample order and the events occurring during the run.
1. Separate visible level from quantity
Record the visible level and foam/bubble observation without using them as substitutes for the agreed quantity measurement. Bottle internal geometry can affect the height presented by a given quantity. Identify bottle lot and drawing reference when investigating an appearance change.
For mass measurements, distinguish the filled gross mass, applicable empty-container tare and resulting net product mass. For example, a fictional gross mass of 124.5 g and tare of 24.5 g give a net mass of 100.0 g. That subtraction does not establish a volume of 100 mL: a volume inference also needs the applicable product density and conditions.
Retain instrument identity, unit, resolution and the measurement procedure. NIST’s weighing terminology explains the tare/net distinction; it does not provide acceptance limits for your filled product.
2. Keep the run log and individual readings together
| Record | Why it belongs in the log |
|---|---|
| Time, sample order and head identity | Separates a time trend from a difference between channels. |
| Gross mass, tare and net mass, or the agreed alternative quantity result | Keeps the measurement basis visible; do not mix units or silently substitute a mean tare. |
| Visible level and foam observation | Preserves appearance separately from measured quantity. |
| Product lot, temperature and measurement location | Identifies the product entering the machine; room temperature is a different observation. |
| Bottle lot, feed condition and refill/stop events | Supports comparisons before and after a recorded change. |
| Recipe revision and authorized adjustment | Distinguishes the original behavior from a changed setup. |
| Inspection status, suspected cause and confirmed finding | Keeps observation, hypothesis and disposition separate. |
Keep individual readings when the average looks stable
Retain readings before rounding or summarizing. Ask the measurement reviewer whether the method can resolve the change under investigation. A stable-looking combined average does not settle a concern about a particular head or pack.
Illustrative head-versus-time example
Fictional data for explanation only. Each point below is one invented net-mass reading. It is not an HM trial, a recommended sampling plan, a machine accuracy claim or an acceptance limit.
| Elapsed minutes | Head A: net g | Head B: net g | Combined mean: g |
|---|---|---|---|
| 0 | 100.0 | 100.0 | 100.0 |
| 15 | 100.2 | 99.8 | 100.0 |
| 30 | 100.4 | 99.6 | 100.0 |
| 45 | 100.6 | 99.4 | 100.0 |
The next useful question is why the two recorded channels differ, using the actual setup and sample identities. The illustration does not prove a faulty component or determine whether any pack passes. Compare real results with the agreed measurement method and acceptance criteria.
3. Compare heads before choosing a cause
Group real readings by head and by stage of the run. A change shared by every head suggests reviewing common events; a change confined to one channel makes that channel and its bottle presentation useful comparisons. Neither pattern proves a cause on its own.
Keep intermittent exceptions separate from a steady trend. Preserve sample order and do not delete an inconvenient reading without a documented reason. The NIST process-stability discussion explains why measurements are examined in time or sample order. This article’s eight invented readings are an illustration, not a stability study.
4. Compare the trend with feed and refill events
Record product-feed level, refill or supply-container exchanges, visible aeration, interruptions and product temperature at a stated location. Mark those events on the same timeline as the samples. If a difference follows a refill, retain that association as a question to investigate rather than naming the refill as the cause.
The significance of feed conditions depends on the actual dosing arrangement. Do not change heating, pressure, agitation or supply connections without the approved process and equipment instructions. For a restart-specific investigation, use the first-bottle checks after a long stop.
5. Preserve the baseline and agree the next check
Retain the approved recipe and the last acceptable run record. Make a controlled, authorized comparison with the same documented product and container combination, and record what changed. Simultaneously changing dose, nozzle timing and conveyor behavior prevents a clear attribution of the result.
Follow the site’s inspection and product-disposition procedure for sampled and affected packs. If valves, seals or connections require inspection, stop and isolate the equipment under its maintenance procedure. Do not loosen fittings, reach through guards or wipe a running nozzle as an investigation method.
6. Download the log and send a focused support record
Download the buyer review workbook (XLSX) and use the Fill drift log tab. Retain each sample's readings in the tare, gross-mass and temperature columns; the sheet calculates net mass. In the combined observation/reference column, clearly label visible level, foam, refill or stop events and the supporting record reference.
Send machine and dosing-method identity, recipe revision, affected heads, product/container lots, the original readings and the event timeline. Include what changed since the last acceptable run, measurement limitations and safe recordings from outside the guarded area. For weight-based checks, also use the tare, settling and net-weight trial guide.
Customer questions
Should I increase the dose when the visible level falls?
First compare the agreed measured quantity with the appearance, bottle geometry and foam observations. A lower visible surface alone does not establish that the dose setting should increase.
Can the batch average hide a filling-head problem?
Yes. Opposing changes between heads can offset in a combined average. Keep individual readings and head/time identity, then investigate using the actual configuration and acceptance method.
Does a change after a refill prove a supply problem?
No. It identifies a timing association to review. Preserve the refill event, product and temperature observations, sample identities and any other changes before deciding on a cause.
Product reference and related guides
The bottle liquid filling machine page describes the equipment family. Use the supplied machine configuration and approved operating requirements for a technical response. Browse the filling and product-trial hub for selection and run-review guides.
Discuss Your Machine Requirement
Share your product, container or closure samples, required output, process conditions and line layout with H M Pharma Machines for a technical review.
Request a Technical Review