On Fri, Nov 27, 2015 at 7:13 AM, Bob Ammerman wrote= : > By definition you will have some sort of jitter when you are on the edge = of > a transition on the pot. The only way I can see around this is to use an > algorithm that knows when you are moving the pot. That's an interesting challenge... A regular pot has a mechanical range of about 270 degrees, split that over a ten bit encoder and you get 270/1024, about 0.25 degree resolution. On a 2 inch (~50mm) knob that's about 0.004 inch (0.11 mm) at the rim of the knob. I wouldn't think most people have the manual dexterity to really get single bit resolution out of that setup. Switch that to a 10 turn pot and you have closer to 0.062" (~1.5mm) per bit at the rim, that seems achievable. Either way, that last "bit" is going to be a slow "tweaky" dial in, so playing with the filter times probably works well with the human factors aspect Two separate low-pass filters with grossly different time constants? If it sees a large value change, employ the short time constant for the next X seconds, then revert to the longer time constant. If the short time constant was on the order of 10mS and the longer was say 100mS..... Slightly more elegant option may be a single filter with the filter time constant tied to the magnitude of the most recent change delta - small change =3D large time - with some sort of math function. A bit of experimentation is warranted here, but I suspect a bounds-limited exponential behavior is where you want to be. But I have to agree with Bob, a digital encoder is a much cleaner solution here. And as someone else eluded to, I doubt the human user can perceive the difference in output of a 10 bit system at single bit resolution. -Denny --=20 http://www.piclist.com/techref/piclist PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist .