An enclosure can be tested to IP66 and still leak in service through the one feature the test did not exercise: the cable entry. Glands are usually rated separately, fitted by someone following a torque they remember, and expected to deliver the same protection as the moulded body.
The thread is the part everyone specifies and the least important. The seal is the part that decides the rating, and it depends on the cable, not the enclosure.
What a gland specification actually contains
| Property | Typical value | Why it decides the result |
| Thread | M12 × 1.5, M16 × 1.5, M20 × 1.5 | Fits the enclosure, says nothing about sealing |
| Cable diameter range | e.g. 3 mm to 6.5 mm for an M12 | A cable at the bottom of the range may not be compressed |
| Sealing insert | One or two part, EPDM or TPE | Hardness sets the compression needed |
| Tightening torque | Often 2 N·m to 4 N·m for a plastic body | Too little leaks, too much splits the insert |
| Ingress rating | IP68 declared with a depth and time | The rating is only valid within the cable range |
| Strain relief | Required by many product standards | A pulled cable opens the seal |
The cable diameter range is the specification people skip. A gland that covers 3 mm to 6.5 mm has to compress its insert by a large fraction at 3 mm and almost not at all at 6.5 mm. At the top of the range the seal may be adequate; at the bottom it may be barely touching, and the IP rating still applies on paper.
Why the same gland passes and fails
Three variables decide it, and only one is on the datasheet:
- Cable jacket hardness and surface. A soft PVC jacket deforms into the seal; a hard PUR jacket does not, and a longitudinal rib on the jacket gives water a path.
- Hole preparation. A jacket cut with a dull knife has a burr that lifts the seal. A jacket stripped too far back leaves the shield exposed inside the gland.
- Torque application. Hand tight on a plastic body is usually under-torque; a wrench on the same body is over-torque and cracks the insert.
Testing a cable entry properly
Test with a real cable of the smallest diameter the gland claims, fitted as in service, with the gland torqued to the stated value. Then test with the largest diameter. Those are the two ends of the claimed range, and a gland that passes only one of them has a narrower range than its datasheet says.
Add a pull test and a bend test before the ingress test: a gland that holds water on a straight cable may leak after the cable has been pulled at 60 N and bent through 90 degrees. That sequence is what the installed product sees.
Recording it
Record the gland model, thread, cable diameter and jacket type, torque value, and whether strain relief was fitted. For an IP68 claim, record depth and duration as well, because IP68 without a condition is a marketing statement rather than a test result.
A cable gland test fixture that holds the enclosure, applies the pull and seals the entry during the immersion keeps the three variables separable, and the site index lists the gland sizes by thread and cable range.
Standards referenced: IEC 62444; IEC 60529.
