Tips on leveling a lathe.

[ddickey]

You should take a spring pass.

 

[D.sebens]

You should take a spring pass.

.002 at slowest feed. No spring pass marks when cranking handle back.

I did on retracting the carriage. Plus the dimension was smaller on the tailstock side. That cannot be caused by deflection as that end was unsupported.

 

[D.sebens]

May of made some progress! I previously maxed out adjusting the cabinet to the point of looking like it will rip itself apart. I decided to shim the rear tailstock end casting again. I’ve got it down to 4 thou at 16” and 1 thou at 8”. Clearly it’s not linear from bed wear I’m guessing. I’m going to let it sit like this over night and see if it changes. If it doesn’t should I see if I can add a little more shim to it and bring it closer?

Fingers crossed that the bench can hold back the shim I put in and it doesn’t stress back

Also I’m having trouble visualizing it will the 4 thou difference cause the part to be 4 thou smaller or 8?

 

[verbotenwhisky]

May of made some progress! I previously maxed out adjusting the cabinet to the point of looking like it will rip itself apart. I decided to shim the rear tailstock end casting again. I’ve got it down to 4 thou at 16” and 1 thou at 8”. Clearly it’s not linear from bed wear I’m guessing. I’m going to let it sit like this over night and see if it changes. If it doesn’t should I see if I can add a little more shim to it and bring it closer?

Fingers crossed that the bench can hold back the shim I put in and it doesn’t stress back

Also I’m having trouble visualizing it will the 4 thou difference cause the part to be 4 thou smaller or 8?

this would be my strategy. If it is twisted it takes time, you cannot rush, if you believe it is twisted support the head stock end off the table a half inch or so and place another support the same thickness under the front or the back based on which direction it is twisted then ad a little extra weight on the opposing side. Check it daily, it will relax gradually provided it wasn't twisted when is was originally machined and doing it this way will insure it stays flat.

 

[D.sebens]

I think we made it boys! I’ve got it down to 1.25thou in 8” and .0002” at 16. I would love to bring 8” in at the expense of 16 going out but to get the 8” measurements in it would really throw out the 16”. I can change the end quite a bit with minimal change at 8”.

 

[RJSakowski]

I think we made it boys! I’ve got it down to 1.25thou in 8” and .0002” at 16. I would love to bring 8” in at the expense of 16 going out but to get the 8” measurements in it would really throw out the 16”. I can change the end quite a bit with minimal change at 8”.

Is the taper in the same direction at both 8" and 16"? If so, tweak it a little more to flip the direction of the taper at 16". you should be reducing the taper at 8", hopefully to under .001" while not introducing a terrible taper at 16".

Knowing what your lathe is actually doing, you can make some dead reckoning corrections while turning to get your desired results.

 

[D.sebens]

Is the taper in the same direction at both 8" and 16"? If so, tweak it a little more to flip the direction of the taper at 16". you should be reducing the taper at 8", hopefully to under .001" while not introducing a terrible taper at 16".

Knowing what your lathe is actually doing, you can make some dead reckoning corrections while turning to get your desired results.

Unfortunately its a hump. 0.00 to -1.25 to +.0002

 

[Old Iron Guy]

Hey all I am trying to level out my sheldon lathe. I dont have a precision level, I figured I would use my construction level to get things close after placing the lathe. I am "close now" but I have ran into a problem, I cant seem to get the twist out. I mounted a level on two flat surfaces of the carriage and moved it from end to end. I am using the adjusting nuts on the cabinet(steel desk construction not cast) and have gotten to the point that I need to raise the rear leg at the tailstock but the front leg at the tailstock is currently floating. I dont have a test bar or even a long piece of metal to play with to see but I did turn a patch on a bar about 1.5 inches long and got 4 thou of taper. It tapers smaller towards the tailstock end. This was done unsupported with a 1-2 thou cut at the slowest feed. This is repeatable on cuts.

This leads me to believe either the front tailstock adjuster needs to lower or the rear needs to raise but it is currently floating on the front. I suppose I could bolt down the front leg to the floor? Right now I left it floating so maybe it wants to relax itself down while resting.

Most of what I do isnt "high precision" its always as precise as it needs to be. But 4 thou seems like a lot. I have a ram from a hydraulic cylinder that is in good shape but scrap. I am thinking of measuring it to see how consistant the size is and if reliable chucking it up as a test bar equalize runout along the bar and use a indicator on the carriage to sweep it and see if the issue is just bed wear near the front or real twist. I have basically confirmed it is bed twist as I cant get it level but figure I would get more info.

So after all that mess of words I am hoping for some advice on how to level it more and if my thought on the test cut matches my thoughts on which leg needs to move which way.

Few things come to mind. Some may not apply since they are already checked.
-Tailstock offset should first be checked so it matches the headstock. 4 thou taper in 1.5" seems a lot to solve by levelling.
-a welded sheet metal stand is almost certainly not a flat plane on top and also twists and sags as you try levelling it. The heavier headstock end is likely also bowing it. Get it close to level first before trying to level the lathe. A solid wood top helps IMHO. Serious steel is better.
-How long has the lathe sat there unlevelled. Iron settles over time if under stress. I level once, wait a month and try again. Always changes after any move.
-Most lathes have some bed wear, even with hardened beds. Mostly on the front V. You could twist the bed to get rid of it but sometimes it's ok to just work with it.
- You can chuck the hydraulic ram but remember to spin it in the chuck and average the readings. Chuck jaws have error. Collets are not perfect. Spindle may have slight twist relative to lathe bed. Hydraulic ram may have a slight curve. Tailstock may not be parallel to bed. Everything has errors.
- last thing. Once you start cutting, the pressure of the cutter changes the geometry. You can go crazy chasing things. Better to do your best and just start working.

 

[D.sebens]

Coming back to this. The twist came back. I decided that I should go at this more methodical and less brute force.

When I first was setting it up I didn’t have a precision level. I do now. I loosened up the foot pads and adjusted it level cross way and mostly level long ways. More importantly they are basically matching at the headstock and tailstock. I set the level on a large flat spot on my compound for the measurements. They were all done without moving the level from its rest.

Re-measuring alignment at this point using the RDM (rotating the spindle and averaging the readings) setup. I was vertically low at the tail stock end of the bar and too close to the tool holder at the tail stock. Thinking about this I decided that if I shim the right front of the head stock it would work in the right direction for both. On my lathe the headstock indexes on the rear v way so the front is flat.

It ended up needing shimmed .016” at the headstock right front bolt. This put me .00025(ish) difference on my taper measurement at 8” with the 8” end being larger and .001” taper measure at 16” with 16” end being smaller.

These are measured with a dial so it should be double that what cutting. I think the shift in direction between 8” and 16” is bed wear.

The top (vertical) measurements are basically bang on for 0” and 8” and .005” high at 16”. I decided to leave it at this as this is probably bed wear.

As 8” should be cutting too large in respect to headstock and tailstock I think I’ve reached my optimum setup. Especially as the 0” and 8” are basically bang on.

After doing that head stock tweaking the spindle was noticeably harder to rotate. This had me a bit worried but I loosened the tops of the spindle bearing caps and retightened and it now rotates freely.

I think I’m getting it dialed in now. I’ll have to do some test cuts to verify but by the method it should be close.