Technical note · USDA PROBE
Why Ice-Bath Preparation Matters in USDA Probe Calibration
An ice-water bath is commonly used as a reference near 0.00°C when checking temperature probes for reefer cold treatment. The reliability of the check depends on the bath, the reference thermometer and the controller procedure—not just the probe.
What happens if the bath is actually +0.30°C?
0.00°C ≠ +0.30°C
Suppose a correctly functioning probe indicates +0.30°C in a bath that is actually +0.30°C. If the operator assumes the bath is 0.00°C, the correct indication can be mistaken for a +0.30°C sensor error. An adjustment based on that false assumption may introduce an offset. The exact result depends on the controller and its procedure.
Comparison of reference conditions
| Condition | Correct reference bath | Incorrectly assumed reference |
|---|---|---|
| Actual bath temperature | 0.00°C | +0.30°C |
| Correct indication from a functioning probe | 0.00°C | +0.30°C |
| Reference temperature assumed by operator | 0.00°C | 0.00°C |
| Apparent probe deviation | 0.00°C | +0.30°C |
Illustrative example only: actual controller response depends on the manufacturer’s calibration logic.
Preparing and checking an ice bath
- Use a clean, preferably insulated vessel and clean, finely crushed ice.
- Add clean, preferably pre-chilled water just to fill the gaps, keeping abundant visible ice throughout the check.
- Mix gently and allow the ice-water slurry to stabilize; maintain good temperature uniformity.
- Immerse the sensing portions sufficiently, without touching the walls, bottom or each other.
- Measure near the probes with a suitable calibrated reference thermometer, and record its reading and uncertainty.
- Follow the specific reefer controller manufacturer’s calibration/check procedure; verify the bath before rejecting or adjusting a probe.
A bath close to 0°C is not guaranteed merely by adding ice. Water quality, ice quantity, mixing, immersion and heat input matter. The +0.30°C example illustrates reference-temperature error; it does not prove that every failed probe is acceptable.