Writer: admin Time:2026-11-16 09:30 Browse:℃
An ingress test looks like it is about the enclosure. Half the time it is about the water. Two samples from the same batch, tested in the same rig on the same day, can give opposite verdicts because one tank was at 18 °C and the other at 26 °C, or because the water had been reused since morning. Neither the drawing nor the test report mentions the water, which is exactly why it goes wrong quietly.
IEC 60529 addresses this with a single sentence that is easy to read past: the temperature of the water shall not differ from that of the specimen by more than 5 K. It is not a comfort rule. It is a rule against condensation.
Cool a warm enclosure in cold water and the air inside contracts. Air has to come from somewhere, and it comes in through the seal, carrying water vapour with it. Minutes later the enclosure warms back up, the vapour condenses on the inside of the lid, and the technician opens the box to find droplets sitting exactly where a leak would put them.
The physics is straightforward. A 20 K drop in a 1 litre air volume at 101 kPa shrinks that volume by roughly 7 %, which is 70 mL of air drawn in through whatever path exists. If that air is saturated at 25 °C, it carries about 23 g/m³ of water, so 70 mL brings in around 1.6 mg of water. That is small, but it lands as a film on a cold surface and reads as a failure.
| Scenario | Water vs sample | What you see afterwards | Verdict risk |
|---|---|---|---|
| Sample soaked at 25 °C, tank at 25 °C | 0 K | Nothing on the inside | Baseline, trustworthy |
| Sample at 25 °C, tank at 10 °C | 15 K colder | Fine droplets on the lid, no tracks | False fail, most common |
| Sample at 10 °C, tank at 25 °C | 15 K warmer | Air expands outward, seal pushed out | False pass, seal stressed the wrong way |
| Both left in the room for 24 h | Within 1 K | Nothing | Correct practice |
The practical fix costs one night. Leave the sample in the test room overnight so it reaches room temperature, and keep the tank in the same room. A 24 hour soak at 20 ± 5 °C removes the whole class of argument before the test starts.
If the pass or fail decision is made with an indicator lamp or a low-voltage signal, the water has to conduct. Tap water in most cities sits between 200 and 800 µS/cm, which is enough for a lamp circuit at 40 V through a series resistor. Deionised water can be below 5 µS/cm, and at that level a genuine bridge between two terminals may not light anything.
The reverse problem is dirt. Reused water accumulates metal particles, flux, and traces of oil from previous samples. Conductivity climbs, and a marginal leakage path that would not have shown in clean water now reads as a failure. Neither direction is recorded anywhere.
A useful water record has five fields: tank temperature, sample temperature before immersion, conductivity, hours since the last water change, and whether the sample was conditioned in the room overnight. Five lines turn "the results were inconsistent" into a finding.
That is the whole point of the 5 K rule. It is not there to make the setup harder. It is there so that the droplets you find afterwards belong to the product and not to the tank.
Contact: Eason Wang
Email: sales@china-gauges.com
Tel: +86-755-33168386
Address: 1F Junfeng Building, Gongle, Xixiang, Baoan District, Shenzhen, Guangdong, China