Home   |   Technical Articles

Technical Articles

Test Pin Diameter: Why the Go Gauge Is Made to Maximum Material

2026-11-17

Every dimensional gauge in a plug or socket test set is cut to a limit, not to a nominal. That single sentence explains most of the arguments about whether a part passed: the go gauge is not 4 mm, it is 4 mm plus nothing minus a very small amount, and the not-go gauge is the opposite.

Get the direction of the tolerance wrong and the gauge still works. It just works in the wrong direction, passing parts that are too small and rejecting parts that are correct, and nothing about the gauge looks wrong.

Which end sits where

GaugeMade toPurposeFailure it detects
Go, for a hole or socketMaximum material of the mating pinMust enterOpening too small
Not-go, for a holeMinimum material of the pinMust not enterOpening too large
Go, for a pinMinimum material of the socketMust pass over the pinPin too large
Not-go, for a pinMaximum material of the socketMust not passPin too small
Wear allowanceA fraction of the product tolerance, not of the gauge toleranceKeeps a worn gauge honest for one intervalSilent drift

The logic is that a gauge represents the worst acceptable counterpart. A socket is tested with the largest pin it must accept, because that is the pin that will stress it. A pin is tested with the smallest socket it must fit, because that is the socket that will reject it.

Why the band is so tight

Gauge tolerances are typically a tenth of the product tolerance or less, and often capped in absolute terms. If a pin dimension carries 0.06 mm of product tolerance, the gauge that checks it may carry 0.005 mm. The reason is not precision for its own sake: the gauge has to be able to say yes or no, and a gauge that consumes a fifth of the product band turns a 0.06 mm window into an argument.

This is also why gauge drawings are revised more often than product drawings. A change in the product tolerance band moves both limit gauges, and the sheet number changes with them.

Checking a gauge without a CMM

  • Zero the micrometer on a setting rod, not on the anvil. Anvil wear shows up as a reading that drifts with the direction of approach.
  • Measure in three planes, 120 degrees apart, at three heights. A pin that has worn oval reads correctly in one plane and small in another.
  • Check the tip radius with a radius gauge or a shadowgraph. The tip is where the wear is, and a flattened tip changes both the diameter and the entry behaviour.
  • Record the temperature. Steel moves about 11 micrometres per metre per kelvin. A gauge measured in a cold workshop in winter and used in a warm inspection room in summer is two different gauges.

The wear direction is not random

A go gauge wears small, because it rubs against the part on every insertion. A not-go gauge wears small too, and that is the dangerous one: a not-go gauge that has worn below its limit will enter a hole it should have been rejected by, and it will do so gradually, so no single part looks wrong.

The practical answer is a scheduled check against the gauge drawing at a fixed number of insertions rather than at a fixed date. Five thousand insertions is a reasonable first interval for a production gauge; a laboratory gauge used a few times a week can go a year. A go no go gauge set with a certificate that states both limits, not just the nominal, is the only kind worth buying, and the site index lists the common pin patterns by standard.

Standards referenced: IEC 61032; IEC 60529.

close
Scan the qr codeClose
the qr code