Whether it’s a screw thread or digital micrometer, the instrument’s level of precision depends on two factors: the inherent accuracy of the reference (the screw thread or the digital scale) and process errors.
With a screw micrometer, accuracy relies on the lead of the screw built into the micrometer barrel. Error in this type of micrometer tends to be cumulative and increases with the length of the spindle travel. This is one reason micrometers come in 1-inch (25 mm) measuring ranges. Apart from the difficulty of making long, fine threads, the error generated over the longer lengths may not meet performance requirements.
One way to improve the performance of the measurement is to tune the micrometer to the range where it is most likely to be used. For example, if a 0- to 1-inch (0 to 25 mm) micrometer is to be used on parts toward the largest size, the micrometer could be calibrated and set up so that the optimum accuracy is at some point in its travel other than at its starting point. You could chose the middle to balance any errors at the end points or elsewhere to maximize performance at any particular point of travel.
Aside from the calibration error of the thread, which reflects the accuracy of its movement per rotation, there are two other thread-related errors to be aware of. One is error within the rotation, known as drunken thread, because of slight thread waver over the course of a rotation. The other is slip-stick, or backlash, which is caused by unwanted slop between the mesh of the threads. This is a common cause of reversal errors. As a point of reference, the drunken thread is like profile error on a machined surface, while slip-stick is similar to backlash errors seen in gears on dial indicators.
With electronic micrometers, the thread usually drives a sensing head over a scale or uses a rotary encoder as the displacement indicator. Both can induce errors, but the thread of the barrel remains the largest source of error. An electronic micrometer can remember and correct for such errors, and, in the end, can provide better performance than the interpreted mechanical micrometer.
The process for checking the performance of a micrometer is similar to the process for checking other comparative or scale-based instruments. Gage blocks of known sizes are measured, and deviations from expected values are plotted. Usually the gage blocks are chosen so that the spindle travels for a full or half turn of the screw. A rotation of the screw can be analyzed by taking small increments of measurements around the peaks discovered on the first pass. These increments—maybe ten steps in one revolution—may reveal larger errors or show patterns that were machined into the screw threads.
The other significant cause of errors can be found in the parallelism of the anvils. The precision method for inspecting the condition of the anvils is with an optical flat. Using a monolithic light source, it is generally acceptable to allow two visible bands when assessing individual anvil flatness. For inspecting parallelism, six bands may be observed, the combined total of both sides.
The applied measuring force of the sensing anvil on the part and the reference anvil is the other source of process measuring error. The friction of ratchet drive thimbles reduces the deflection of the micrometer frame, but the condition still exists as a source of error. With about 2 pounds of measuring force, typical frame deflection is roughly 50 microinches, although this is apt to increase on larger micrometers where the rigidity of the frame increases.
Other sources of error can also sneak in. Temperature, dirt and the means by which the operator aligns the gage to the part affect any micrometer’s overall performance.