I just finished my latest calibration after creating a new sled using @Jayster technique shown here…
Seems to be a pretty solid improvement, as now there is about a 1/16" vertical deflection when rotating the sled through 90 degrees with the bit cutting .1 inch deep… the resulting hole is 5/16" high x 1/4" wide. I spent a lot of time with this build, and don’t really know how I could do much better working with wood screws and hand-held tools. Also, I was able to get a pretty accurate measurement of the “rotation radius”, measuring 119 mm from the center of the spindle to the outer edge of the ring. From the outer ring to the bottom of the roller bracket where the chain attaches is 20 mm, so a total of 139 mm, pretty much in the zone that is suggested in the cal routine.
I ran the Chain Calibration routine. I actually ran it twice, as it seemed like the chain may have skipped when it was playing out the first time, and the results were pretty crazy. I like the way it is set up, with the software calculating how much chain needs to be played out. A countdown would be a nice addition to give time to get to the chain/sprocket. I advise keeping tension on the “tail” of the chain, as links sometimes catch on the sprocket teeth. I have run the old “chain-measuring-motor-space” routine more times than I can count, and always eased the chain out in shorter increments, so the links never skipped. Having it all go in one run seems to increase the skipping possibility. Here is the result:
I then zeroed out the fields in the advanced settings, per Bar.
After recalibration, ran the Benchmark, and the results were better than I have seen up to now, although I only used my metric steel rule. Doing the metric conversions from my British units caliper was too much for me today… Here is the chart…
Still short in height, but a real improvement over the last one…
I am excited to see what further improvements are possible when the chain calibration process can be put to use.
Thanks for all your work, @bar !