Separation Records: What to Write Down After Each Run
Most separator problems are diagnosed by memory, and memory is unreliable over weeks. A farmer who says the cream has been thinner lately is usually right, but cannot say when it started, whether it moved gradually or in one step, or what else changed that week. Six fields written at the machine turn that conversation into a comparison.
This is not paperwork for its own sake. It is the cheapest instrument available: it costs about a minute per run and it is the only way to separate "the milk changed" from "the machine changed".
The six fields
Write these while the run is happening or immediately after. Anything reconstructed at the end of the week is an estimate, and estimates defeat the purpose.
1. Date and start time. A time as well as a date, because two runs on the same day at different temperatures are not comparable.
2. Milk volume in, measured. Not "about two cans". Use the same container each time so the measurement is consistent even if it is not exact.
3. Milk temperature at the start. The single most common cause of a run that behaves differently from the last one. A thermometer reading, not "cold".
4. Cream screw setting. Record the position, not "normal". If your machine has no scale on the adjuster, mark your own reference once and count turns from it.
5. Run duration and output. How long the run took end to end, and how much cream came out. Weighing the cream is better than judging the container.
6. One line of observation. Anything that differs from last time: sound, foam, overflow, cloudy skim, a slower start, a drip.
Six fields fit on one line. A notebook with a ruled margin and a pencil tied to the bench works as well as anything.
What the numbers let you answer later
With a few weeks of lines, four questions become answerable that are otherwise guesswork:
- Did the change happen in one step or gradually? A step points at a changed input or a changed setting. A gradual drift points at wear, or at the milk supply changing over the season.
- Did yield per litre move while the setting stayed the same? That is the input changing, not the machine. The yield arithmetic shows how much of a change a different fat content explains.
- Is cream consistency drifting between batches? Comparing setting against output over several runs is how you find out whether the drift is in your procedure. The consistency guide covers the variables worth controlling.
- Is the run taking longer for the same volume? Duration creeping up at a fixed volume and temperature is the earliest sign that something is restricting flow, well before it becomes a visible fault.
A layout that survives daily use
A wide notebook page, one run per row:
| Date | Start | Litres in | Temp °C | Screw | Minutes | Cream kg | Note |
|---|---|---|---|---|---|---|---|
| 12 Oct | 06:40 | 180 | 34 | 2.5 turns | 52 | 17.4 | normal |
| 14 Oct | 06:35 | 180 | 31 | 2.5 turns | 58 | 16.9 | slower start |
| 16 Oct | 06:45 | 175 | 37 | 2.5 turns | 50 | 17.8 | foam at outlet |
The point of the table is not neatness. It is that on 14 October the temperature and the duration both moved, and without both columns written down you would have blamed the machine.
What not to record
Do not record a number you did not measure. "Cream about 15 kg" repeated for a month is worse than an empty column, because it looks like data and will be compared against measured values later. An empty field is an honest signal that the run was not measured.
Do not record the setting as a word. "Normal" means different things in March and October.
When a record set is too short to conclude anything
Two runs do not establish a trend, and one unusual run is not a fault. The useful threshold is roughly five runs at a consistent procedure. Below that, treat any apparent change as noise and keep measuring rather than adjusting the machine.
If a change does persist past five runs and the inputs were stable, the throughput guide sets out what affects run time before you start dismantling anything.
Related maintenance resources
Use the ANDUUK replacement parts catalog to match seals, discs, outlets, motor components and cleaning tools to the correct machine assembly.
Browse replacement parts · Read the cleaning and reassembly checklist · Read separator FAQs