View Full Version : Lathe alignment/setup--what's good enough?
07-12-2007, 03:20 PM
I recently reassembled an 11" Rockwell lathe after a move to the basement. I basically just leveled the cabinet with a carpenter's level then assembled the bed/headstock and everything. I backed the built in adjusters all the way out and tightened the mounting bolts. I cut 2 test bars. One was 3/4" dia and about 5" long. The other was aboout 1 1/4" dia and 12" long. Neither one varies more than .001" in diameter end to end. The 3/4" bar was just held in the 3 jaw. The larger bar was held in the 3 jaw and live center from tailstock. The previous owner did have some adjustment done using the built in risers. I guess I'm wondering how tight you guys have your lathes aligned? Here's a pic of the bigger test bar...
07-12-2007, 04:02 PM
I think .001" is pretty good over those distances. You're probably not going to do any better than that. Check again in a couple weeks, and see if anything has settled. I'd be happy with the results you're getting, now.
Oh, and, I'm sorry about your brother, too. Hope things are getting better.
Sounds like you got her set up pretty close. Do you know anyone with a precision level you could borrow?
"Good enough" can only be defined by your needs.
07-12-2007, 10:08 PM
Yep, I'm pretty happy with the results. I'm also somewhat sceptical--seems fishy that it could be that easy. Dumb luck? I cut the 2 different lenght/diameter test bars just in case there were offsetting errors with the chuck/tailstock/bed twist etc. I guess it proves out. I still will try and check it with a proper level though...
07-12-2007, 10:48 PM
For the unsupported test, a 1-1/2" to 2" diameter aluminum or mild steel bar would limit any deflection, along with a good sharp tool and small nose radius.
The long bar test is more of a check for tailstock alignment (if it is adjustable). Keep it as a test bar by turning a protruding band at each end with matching diameters (by offsetting the tailstock). Each time you change tailstock offset and want to re-zero it, use that bar and a dial test indicator. Den
07-12-2007, 11:00 PM
The short one is pretty much the only one that really matters for anything other than aligning your tail stock.
You should definitely be able to get the one between centers to WELL under 0.001 in 12" unless your bed is really twisted or unevenly worn (which would show up as variable deviation along the length, not an even taper). But don't even fool with that till you get the head stock (if you want to fool with it at all that is).
0.001 in 5" isn't really good at all by some accounts, plenty "good enough" by others. But it's also not that simple. Read recent posts here (and elsewhere) by searcing on things like "RDM" (Rollie's Dad's Methdod), "test bar", "lathe alignment" and so forth. Lots of opinions and controvercy, make your own decion based on ability and need. But my worn old 11" turned 6" of of 2.5(?) low carbon steel to considerably less than 0.001 deviation in 6".
Edit: Looks like nheng beat me to it.
07-12-2007, 11:12 PM
I used Rollie's Dad's Method which I alternatively named the Average Wobble Test when I assembled my Logan. It took 3 or 4 attempts but somehow the headstock finally landed straight. Cleanliness maybe? I recall stoning the mating surfaces at the end and that may have done it. I do remember .001 18" out on the dead center and just a few tenths 8" out unsupported.
Got a 10" piece in there on the dead center right now I'm turning various dia's on in sections for different sleeves I need. It's all good and straight.