Pocket Micrometer

Starrett claims a spring that limits force and removes backlash. Write down why a measurement that varies with grip strength is worthless on a production line.

Measure one foil sheet with the micrometer, closing until you just feel contact. Record it.

Measure the same spot again, squeezing firmly. Record. The two numbers differ — on the same sheet, with the same instrument.

Subtract. That difference is pure operator error, and it is often larger than the tolerance you are trying to hold.

Three people, one sheet, no instruction on how hard to squeeze. Record all three. This is the 1880s workshop.

Now repeat with everyone closing on the ratchet until it slips. The spread collapses. That is Starrett's spring, in your hand.

Wind a short coil from steel wire around the brass rod. Compress it — force rises with compression. A spring converts a squeeze into a known force.

Cut two card discs, press them together with the coil between, and turn one. Past a certain torque they slip. That is a friction thimble.

Close the micrometer, then reverse the thimble slightly. There is a dead zone before the spindle moves. That is backlash, the spring's other target.

Measure the foil five times, always closing from open. The spread shrinks. Consistent approach beats a better instrument.

Measure a 0.20 mm feeler leaf. It is a manufactured standard, so it tells you whether your instrument is honest.

Close the micrometer on nothing and read. If it does not read zero, every measurement you have taken today carries that offset.

Measure the brass rod at three points along its length. A single measurement of a real object is almost always a lie about the whole.

US 399,167, “Micrometer Caliper-Square”, Laroy S. Starrett of Athol, Massachusetts, filed 1 February 1888, granted 5 March 1889. The patent describes a micrometer nut of 25 divisions working against fortieth-inch beam graduations to read thousandths.

Starrett was a Maine farm boy whose first patent was a meat chopper. He founded the L. S. Starrett Company in Athol in 1880 and held roughly a hundred patents; his combination square, a separate blueprint in this library, is the other tool of his that every workshop still owns.

Why constant force is the real subject. Palmer's screw made the thousandth readable. It did not make it reproducible, because the reading depends on how hard the jaw is pressed — and soft materials compress measurably. Starrett's spring removes the operator's hand from the result. That is the difference between an instrument a craftsman can use and an instrument an industry can use, and it is why this patent belongs beside Palmer's rather than inside it.

The pattern repeats. Every instrument in this batch eventually meets the same problem — the human operating it — and every good answer is the same: remove the judgement. The ratchet, the reversal test, the go/no-go gauge and the projected shadow of the optical comparator are four versions of one idea.