Ok, so I set up my edge technology indicator, took the back lash out and dial it to zero. Moved my cross slide to 0.10 and I get like .095 on my indicator. I had my compound set at 30 degrees before. So I set it to zero. Is it normal to lose a half thou?
Looking at your measurements, it looks like you lost five thousandths, and not a half of a thousandth? Can you confirm that?
Loosing a half of a thousandth over 100 thousandths, until you're very familiar with the machine, I'd call that as not a big deal at this point. There's too many things that this could have come from.
Loosing five thousandths over 100 thousandths, that would say to me that there's an error somewhere. Backlash introduced during the measurement and/or not accounted for, indicator setup "anomoly", wear in a lead screw, etc. Something's not right, but if it's a fixed value, all in all, it's something that you may or may not want to address one day, but it shouldn't really be standing in your way.
How much are you missing your intended numbers by? Are you measuring a "first cut"? To do this, you really don't want to measure a first cut. If you take (for example) 0.015 inches off of the diameter on a first pass, I wouldn't expect to see the diameter change by a dialed in amount. The second pass, cutting the same 0.015 inches should be pretty close to that. A third pass should pretty much have you dialed right in. There will pretty much always be some tangible degree of disagreement in actual depth of cut when you change the load on a cutting tool. Setting yourself up for two or three "equal" cuts as you approach the final dimension (or as many cuts as you want), will help greatly with hitting exact numbers.
So, because you are getting an indicator movement that is pretty close to actual tool movement, I'm going to assume that you have a radius dial on your cross slide. It doesn't matter which one you have, you're going to have to "math" sometimes either way. If there were truly a best way to do it, there wouldn't be options.... So your dial will measures the ACTUAL depth of cut, or the actual distance the tool bit moves. Because the work piece rotates, you are taking the cut off of BOTH sides. So the diameter reduction will be two times what you dial in. I'm picking out conservative numbers here, they don't matter, it's just to make the point. Maybe your lathe makes very accurate passes at one tenth of these numbers, or maybe it makes accurate passes at ten times these numbers..... The process matters more than the numbers.
So if you cut a test piece to "whatever" starting dimension- Literally, just put "something" in the chuck, take a cut off of the outside for some workable distance you need to work with and measure accurately (Maybe a half inch or inch long cut on a piece of round stock? Whatever/. Take a first cut and measure it accurately.... You can set a "test target" at 0.060 inches smaller in diameter. That will mean, because you're on a radius dial, you must take 0.030 depth of cut (off of each side) in order to hit your test dimension. So take that in three passes of 0.010 inches. DIal in the first ten thousandths. Because you didn't care how deep the first "cleanup" cut was, you've changed the load on the lathe. That second cut (the first careful cut) might take off 0.008 thousandths (0.016 on diameter), or it might take off 0.012 thousandths. (0.024 on diameter) Then dial in your next cut. If you didn't quite cut enough, So if you meant to take ten thousandths, and you took 12 thousandths, your next 0.010 adjustment, maybe you dial in off 9 or 8t thousandths. (eighteen or sixteen thousandths on diameter) That "should" pretty much put you where you want to be, and you should be pretty darned close to takint the third, final pass at the "desired" 0.010 deppth of cut. Because you've set up "early", you've balanced those cutting loads, and the dimension should pretty much come together.
It does take some practice, and you will probably not hit dimensions to a thousandth of an inch on your first tries at it. But once you figure out how the machine is going to react, it gets a lot easier. And the workpiece too. How rigidly the part is held and how rigid the workpiece is will also make a difference in "final numbers". So you never really entirely get away from this, although in most cases you can probably get down to two "equalish" finish passes once you know the lathe a little better. You always have to check (measure) the actual part, and "react" with the dials to some degree. And the best "bullseye" you can get on a final dimension will come with an actual, tangible cut. Sneaking up on a dimension a couple of thousandths at a time is the hard way at best, and depending on your lathe, your tooling, and your experience level, usually just leads to tears. It's much better if the last pass makes real chips, and not dust or hairs.
