Starting From Zero

Writing ’00 instead of ’99 reminds me of the importance of zeroing out the measuring instrument or gage before starting to make a measurement. Zeroing sets a reference point from which all subsequent measurements are made. If a gage has been allowed to drift from zero, it will introduce error into the measurement process.

Why do gages shift their zero point? There are probably as many reasons as there are types of gages. But the top reasons include wear, temperature effects, loose gaging components and dirt. It’s important to check zero often so that you feel comfortable with the measurement process.

Basic instruments like micrometers or calipers use their scales as the reference. With a vernier micrometer, verifying the reference point is straightforward. Close the contacts together and read the vernier scale. The scale should indicate zero. If it doesn’t, you can be sure that every measurement from then on is going to be off by the amount indicated on the scale.

There are two things you can do to correct the problem. The lazy man’s way out is to add or subtract this offset with every measurement. This is a trap that may cause a lot of grief the first time you forget to apply the offset.

The best thing to do is correct this disaster waiting to happen by adjusting the micrometer to make it read zero with the contacts closed. Most micrometers—both friction and ratchet drive types—provide instructions to adjust this zero point.

Similarly, a vernier or dial caliper can be checked by bringing the contacts together and holding the jaws up to a light. You should not see light passing through the jaw surfaces. Look for gaps or taper conditions that indicate a worn jaw. If the jaw passes inspection, check for the zero readings. On a vernier caliper you will need to read the lines, while on a dial caliper the indicator should read zero. Both can be adjusted to read zero.

Digital hand tools are easy to zero. Close the jaws, and press the zero button. That’s all there is to it.

Comparison type measurement hand tools such as snap gages, gage stands, bore gages, and so on, may be a little trickier, but they also need to be zeroed regularly. The process is slightly different, but the end result is the same. With this type of gage, the zero point is actually a reference dimension to which dimensions on the parts will be compared. An ID, for example, will be shown to be greater or less than the zero (reference dimension).

For dial indicators, the method is to mechanically adjust the dial indicator on the master so that it is in its midrange, and lock it firmly into place. Then loosen the bezel clamp and turn the dial so that the indicator hand lines up with the zero on the dial.

Something else to remember is that setting to zero this once does not end the process. Take the master out and replace it a number of times in the gage, and check for zero again. A bit of dirt may have been introduced in the initial setting and repeating the process a number of times will help instill confidence in the setup. Usually you want to have a dial indicator reading repeat its zero setting to within a half of a grad or so, or a digital readout should be to within one count. This varies a little depending on the gage and the resolution, but in any case we are assuming that the gage has already been checked for repeatability performance.

You should also know that there are instances where you may want to set your gage to a value other than zero. This makes it possible to correct for a known error in the master or to use a different size master for measuring the part. For example, if the master is +0.0002" larger than the nominal dimension for the part, you would set the dial on the indicator to +0.0002" instead of zero. Now if you have a perfect part, the gage would read 0.0" when the part is measured. Electronic dial indicators and amplifiers, in addition to zeroing buttons, usually have master deviation functions to do the same type of correction.

Zeroing the gage is the very foundation of good measurement practice, but we know from experience that most gages are not zeroed often enough. If too many measurements have been made or too much time has elapsed since the gage has been zeroed, measurements will all be biased by zero shift. An extreme solution would be to zero, measure and zero for every part. On the other hand, zeroing once a day is probably too little. Generally speaking, once an hour is just about right, but the application itself should dictate the zeroing frequency.