Please. Having access to better measuring equipment is nice, but that's because it allows one to be *more sloppy* when meeting a given specification, not because it's absolutely required by the laws of physics. If one has to maintain a tolerance of .001", then it can be done by either:
- using measuring equipment that is only reliable to .0004", and making sure that the measurement is within .0006";
or
- using measuring equipment that is reliable to .0001", and making sure that the measurement is within .0009".
In either case the numbers add to .001". But what a production engineer sees is that if he makes choice #1, he may be saving a few hundred dollars, in getting a cheaper measuring tool, but he may have to rework the entire production machinery (perhaps costing millions of dollars) so as to get the measured errors into a narrower range, without actually improving the delivered accuracy. This is generally a no-brainer: spend the extra money, and get the better measuring tool.
Why stop at a factor of ten? Well, for one thing, it's a nice round number: easy to remember, and easy to use. One can, for instance, add up the possible errors of the chain of calibration quite easily:
10% + 1% + .1% + ... = 11.111...%.Also, beyond that point the returns diminish very rapidly: if a factor of
20 were used, instead of 10, then the required accuracy of the measuring instrument in the above example would be doubled, but the allowable production tolerances would just move from .0009 to .00095 -- only a 6% improvement.So the standard factor of ten makes sense: it's about what most people would choose anyway (if they are among the class of people who pay attention to industrial standards); and standardizing puts everybody on the same wavelength.
On the other hand, for an individual doing a one-off project, improving measured tolerances by being more careful might cost a few hours; and the income from a few hours of work might easily be less than the cost of the more accurate instrument.