How round and straight is the CRS and 4140? I've never found CRS to be round or straight.
I don't understand the plus and minus readings.? Are you zeroing on the low spot?
At this point this is what I would do:
Chuck up a piece of 2" aluminum with about 8" of stick out and take a skim cut to turn it true.
Measure each end of the rod for taper. (If there is significant taper that is another issue.)
Mark the jaws 1,2 &3.
Mark a line on jaw 1 and a corresponding mark on the rod.
Flip the rod and index it to jaw 1, then measure for run out.
Mark the chuck at the low spot and label #1, zero the indicator and the cross slide.
Record the measurements for each jaw, near the jaw and at the end of the turned part of the rod.
Re-index the rod to jaw 2, measure and record with the same indicator and cross slide zero as jaw 1. If the low spot moves mark it on the chuck and label.
Repeat for jaw 3.
Another option would be to make a set of soft jaws, pre-load the jaws and bore them to fit a turned or ground rod and do a similar measurement test as above. This would take the jaw clamping surfaces out of the mix and isolate it to the scroll/jaw teeth interface.
How well to the jaws fit in the slots of the jaw body?
