What's new
What's new

New guy looking for lathe alignment help


Jan 9, 2024
We spent two days leveling and aligning the head-stock on a 1986 Sunmax 1440 lathe. I have a 2 3/8" diameter bar that is 12" long, in a 4 jaw chuck with a live center in tail-stock.

I have indicating the bar along its length (.002" end to end), after removing stock from the bar, both ends measuring within .001" on the diameters, but the center of the bar is about .008" bigger.

Any thoughts?

(made corrections)
Last edited:
How did you align the headstock/tailstock initially ? The way you describe your operation one element could be fighting the other and causing the barreling. if you look on the internet you will find a guy called “ Schlesinger “ came up with the correct methods we still use for aligning simple machine tools a hundred years ago. There are copies of his handbook freely available on the internet. They are your friend.

I don’t want to depress you but I’d expect to align a lathe like that in a morning.

Regards Tyrone
That lathe is 38 years old. It has outlived its toolroom accuracy days. If you have to have better, I expect a way grinder is in your future. Your results are exactly what I expect with a lathe that has spent most of its time as a chucker. Considerable wear next to the chuck.

There is an outside chance that the headstock is pointing to the rear, and your tailstock is pulling the free end of the bar into place. What results do you get with a bar half that length held only in the chuck?
Leveled the bed with a Starret level on the cross slide. Checked x and y directions. Moved saddle from one end to the other, and all was good.

Made a "dumbell" shaped bar as shown on many internet sites.

Zeroed bar at the end without the tail-stock. Was within .002"

Made an initial cut and the bar was .050 bigger on the tail end than the chuck end.

Tried moving the front right foot pad but this lathe is stiff. No bed twist. So I angled the head-stock using the adjuster screws located between the rails at the far end of the head stock.

Using an indicator on the tail end of the bar, I "rotated" the head-stock, repeated, took another cut, until both ends measured the same.
Your tailstock is out of adjustment. It needs to be moved over toward the operator by about .025" for a start. Try another cut and see how close it is and adjust again if necessary.
You are never going to get it zero at each end on a older lathe as you have there.
Remove the tailstock and cut it. Look up "The 2 collar Test" on YouTube as Adam Booth shows how I taught him. I suspect thought the bed way close to the chuck is worn low .002 to .004 plus the bottomofthesaddle is worn too. Your machine being Taiwanese the headstock sits on 2 flat ways and rotates on a pin.
When a customer would call me to align the bed to remove the twist because the machine cut a taper. If the ways are worn near the chuck. Front inside way as on lathes that area wears bad because the chips land there. The best cure is to have the bed ground. But many can't afford a $10,000.00 rebuild job.
I would ask the customer if he used the tailstock like your doing or just use the chuck to turn out of the chuck and drill holes with the tailstock?
If that's what you do use the 2 collar turn test and loosen the headstock and turn it on the 2 collars the same. Most of those headstock have a adjuster located on the far left side of the head. Remove the cover or cowling, loosen the cap screws that hold down the head and under the gears you can losen the adjuster and twist the head. Re-tighten everything and test cut it again. I set a mag base on the bed or compound and indicate the shaft. It usually is 1 to 1 of movement. Then after that align the tailstock side using you test bar. You can Google 2 collar test Practical Machinist and you will find dozens of threads on this as we have been discussing this on here for years.
I think you'll need one of those special starrette super-levels. Maybe a laser too. Then you could cut the head off a live chicken and paint the lathe bed with the blood.

No seriously, take the tailstock OFF the lathe, put the tailstock on the floor. Then do a two collar test. Report those results back here.
Sounds like you adjusted all the good out of your lathe already. Best wishes putting it back in.

Tailstocks are adjustable. You align the headstock first, then adjust the tailstock.
I went through this a few years back:

Did you align the headstock with a proper test bar? Also I've never heard of "levelling" a lathe using the carriage, usually you put the level on the ways directly to take out any twist.
“Sounds like you adjusted all the good out of your lathe already. Best wishes putting it back in.”

I don’t understand. I started with .010” taper in 6”. Now I have maybe .001” taper in a foot

. Maybe I wasn’t clear. I used the two collar bar method without the tail-stock

I could not place the level across the ways. They are at different heights.

Like I said by adjusting the head-stock I was able to have both collar diameters within .001”. But now the center two inches of the bar is .008” bigger in diameter.

Going to recut my two collar test and then look at the ways for wear.

Amnything else?
................. Also I've never heard of "levelling" a lathe using the carriage, usually you put the level on the ways directly to take out any twist.
This only works on a new right out of the crate. One that is 38 years old with worn ways, this method is thrown out the window, it doesn't work.
As you mention, put the level on top of the bed at the headstock and the tail end of the bed to take out the twist.
Start here, https://pearl-hifi.com/06_Lit_Archi...s/Schlesinger_Georg/Testing_Machine_Tools.pdf

If the ways are at different elevations, use 123 blocks, gauge blocks, shim stock etc to get a repeatable surface along the bed as your checking it so the levels bubble is roughly centered. In reality the lathe bed doesn't need to be exactly level, if the bubble keeps hitting the same division mark as you move along the bed it's true. Or at least as true as you can get it depending on just how much wear there really is. The ways need to end up parallel without any twist. Slant bed lathes also get checked and adjusted for true parallel surfaces in much the same way.
To get .008" barrel on a 2 3/4" bar the wear would need to be huge or you would need to mount your tool under the job.
Yeah this is getting confusing. I think OP needs to begin at the beginning. I was assuming that the cut test was done while unsupported by the tailstock. Now I'm not so sure.
Ooh I forgot the custom-ground hardened test bar. Test bar, a laser, a taut-wire micrometer setup, an autocolimator, a starrett 199 level and the head off a live chicken. For extra accuracy you need to grind three rocks together to make a surface plate.

Read up on that three collar test. You don't, as you mentioned you did in the original post, support the work using a center in the tailstock.
The problem with the "two collar method" is it only makes the lathe cut true at those two positions.

To properly align a lathe you first need to be sure the ways are parallel (no twist), you do this with a master level along the ways using 1-2-3 blocks as suggested in post 15.

Once the ways are straight you align the headstock to them using a test bar in the spindle and a DTI in the carriage, any wear in the ways and/or saddle will show up at this point.

When you've done those two things your lathe will be as good as it can be and you'll have to work within (or around) it's limitations.
Last edited: