A bayonet cap locks by pushing in and twisting. The two pins travel up a slot, over a shoulder, and drop into a detent. For that to work, the pins must be at the correct height above the reference surface, at the correct angular separation, and the correct diameter for the slot.
The standard fixes the pin height above the cap face, the pin diameter, and the angular position. The tolerances are tight because the detent is small and the spring that holds the lamp in place is strong.
| Feature | Nominal | Tolerance | What failure looks like |
| Pin height above face | 10.5 mm | ±0.2 mm | Lamp sits too high or does not reach the contact |
| Pin diameter | 4.7 mm | −0.1 mm | Loose in the slot, or will not enter |
| Pin angular position | Per drawing | ±2° | One pin enters, the other binds |
| Pin protrusion | Per drawing | ±0.15 mm | Will not pass the entry gauge |
| Cap shell diameter | 22 mm | Per drawing | Shell binds in the holder |
The go gauge checks the maximum dimensions: a cap that passes is not too large. The not-go gauge checks the minimum: a cap that fails it is not too small. Neither checks the angular position, which is why a separate entry gauge exists that requires both pins to pass through slots at the correct angle.
A cap that passes both go and not-go but fails entry has an angular problem or a pin height problem, not a size problem. Knowing which gauge failed tells you which process to look at.
Entry gauge failures cluster. If the failure rate rises from 0.5 % to 4 % in a week, the mould or the pin insertion station has moved. Tracking the failure by cavity, not just by day, usually finds it in an hour.
Standards referenced: IEC 60061-1 for B22 cap dimensions, IEC 60061-3 sheets 7006-11 and 7006-10 for the go and not-go gauges, 7006-4A for the entry gauge.
B22 pins are usually inserted after moulding, into holes in the shell, and then swaged or staked. The height depends on how far each pin is pressed; the angle depends on the hole positions in the mould.
Two pins pressed to different depths produce a cap that sits at an angle in the holder. The lamp works, but the contact pressure on one side is lower, and the cap runs warmer on that side.
A depth micrometer on each pin against the cap face gives the height directly. Measure both pins on ten caps from each cavity; a spread of more than 0.1 mm between the two pins of the same cap indicates a pressing problem.
The angle is harder to measure without a profile projector. A simple check is to insert the cap into a known-good holder and rock it: perceptible rock means the pins are not coplanar.
A gauge that has been dropped can have its slot edges burred. A burr of 0.05 mm will reject caps that are within tolerance, and the pattern of failures will look random.
Inspect the gauge with a magnifier monthly, and check the slot width with a pin gauge. Two minutes of inspection prevents a week of chasing a process problem that does not exist.