| ??? 12/31/03 16:39 Read: times |
#61676 - RE: Atomic 16bit read Responding to: ???'s previous message |
Since on all digital measurement an error of 1 digit must be accepted, the variation of 1 digit was o.k.
Oh, I were not allowed that (you should have seen the result). The project was a pattern knitting machine where needles came in on precise counts. With one count off it started making spiderwebs. It is obvious that if you can handle a nominal error, some shortcuts can be taken. Erik Ps if several starts stops, reverses happen in one revolution the error may be more than 1. |



