Oh boy, you guys are gonna hate me.

[RafaelMarujo]

Thanks for responding so quickly. I hope the experts can help me understand more . For what it’s worth , on my Clausing 5914 if I want to hit close tolerances I found i have to put a dial indicator on the cross slide when I get close to finish size . Again I have no idea why , but my cross slide screw is worn and nut might be too .

True I might have to try that, but I doubt it would work 100%, because my surface finish is also pretty rough... I'll assemble the lathe like it is, take some cuts and document everything so you guys can see

 

[vtcnc]

Thanks for responding so quickly. I hope the experts can help me understand more . For what it’s worth , on my Clausing 5914 if I want to hit close tolerances I found i have to put a dial indicator on the cross slide when I get close to finish size . Again I have no idea why , but my cross slide screw is worn and nut might be too .

A better search term for you to research is "backlash".

"Slop", "chatter", "poor finish" are the symptoms we are feeling/detecting while using the machine. Backlash in the leadscrew/nut pair, excessive wear in the ways (sliding contacts) of the machine and poor setups will combine to create these symptoms.

Here is a quick video.

@RafaelMarujo - I would make an earnest effort to level the lathe and try to remove any twist that may be present in your setup. There are some leveling procedures around here somewhere. I'll see if I can dig it up. Or you can do a search in the meantime. Once you have the machine level, then you can begin to narrow down where additional problems may exist in your machine.

 

[RafaelMarujo]

A better search term for you to research is "backlash".

"Slop", "chatter", "poor finish" are the symptoms we are feeling/detecting while using the machine. Backlash in the leadscrew/nut pair, excessive wear in the ways (sliding contacts) of the machine and poor setups will combine to create these symptoms.

Here is a quick video.

@RafaelMarujo - I would make an earnest effort to level the lathe and try to remove any twist that may be present in your setup. There are some leveling procedures around here somewhere. I'll see if I can dig it up. Or you can do a search in the meantime. Once you have the machine level, then you can begin to narrow down where additional problems may exist in your machine.

I do know about backlash, i always remove backlash and try to tighten unused axis... In my teachers lathe i do the same and everything hits correctly to the hundreds of a mm.
I didn't yet level the lathe perfectly because it's not in its final place, but I will try to level it better when I reassemble it

 

[jwmelvin]

You seem to have a problem with repeatability. And it seems that the saddle does not properly fit the bed ways. That lack of proper fit could show up as an uncertain position for the saddle in the x-axis (cross-slide direction).

It was suggested earlier to determine the bearing surfaces for the saddle on the bed ways. Given the extreme angle that the saddle seems to take, and the apparent nonrepeatability, this should be your prime objective. The saddle needs to have a consistent interface with the bed ways. It also needs as large a contact area as possible (distributed across the saddle extents). Once that is achieved, you will have good stiffness and good averaging of any way nonlineararity. Resurfacing the saddle as Richard mentions would achieve the broad contact that you need. But determining your current contact surfaces may shed some light on the issue.

Edit: also, if you can get a more sensitive level, you’d want to measure across the bed ways without the carriage supporting the level (e.g. on 123 blocks) to see if you can get the ways themselves straight without any possible influence of the saddle-bedway fit.

 

[jwmelvin]

I do agree with others that leveling (untwisting) the bed ways seems important before determining saddle-bed fit. That’s why I mentioned using a level on the bed without the saddle.

Even once the bed is as straight as possible, I’d want to consider the saddle-bed fit at multiple positions along the z-axis, as bed wear will certainly affect the bearing surfaces.

 

[RafaelMarujo]

You seem to have a problem with repeatability. And it seems that the saddle does not properly fit the bed ways. That lack of proper fit could show up as an uncertain position for the saddle in the x-axis (cross-slide direction).

It was suggested earlier to determine the bearing surfaces for the saddle on the bed ways. Given the extreme angle that the saddle seems to take, and the apparent nonrepeatability, this should be your prime objective. The saddle needs to have a consistent interface with the bed ways. It also needs as large a contact area as possible (distributed across the saddle extents). Once that is achieved, you will have good stiffness and good averaging of any way nonlineararity. Resurfacing the saddle as Richard mentions would achieve the broad contact that you need. But determining your current contact surfaces may shed some light on the issue.

Edit: also, if you can get a more sensitive level, you’d want to measure across the bed ways without the carriage supporting the level (e.g. on 123 blocks) to see if you can get the ways themselves straight without any possible influence of the saddle-bedway fit.

That's a good idea I will try to check the saddle bed fit with some paint before I assemble it...
The lathe might not be 100% level,, but it's 99% and I wouldn't think that would give the error it's giving... I definitely expect to have some parasitic error forever, but I think it's way too much right now...
Now I'm starting to Think the error wasn't that much and I was overreacting, I hope not

 

[vtcnc]

Now I'm starting to Think the error wasn't that much and I was overreacting, I hope not

I hope you were overreacting! If there really is that much error, there is something seriously wrong.

 

[RafaelMarujo]

I hope you were overreacting! If there really is that much error, there is something seriously wrong.

I'm willing to bet there's something wrong with it, there's a reason this was a 500€ lathe, specially a 2m between centers one, it's worth more in scrap

 

[Richard King 2]

Over here in the USA there is a simple solution to worn acme screws and nuts. We can buy off the shelf pre-made ones. Then we cut the old screw portion off and keep the area where the front part of the screw that fits into the saddle. then bore a hole in the front part, turn the new screw down to the bored hole to a press fit and cross drill install a taper pin or silver solder them together. Mill the acme nut to fit like the old one. http://www.greenbaymfgco.com/ACME-threaded-rods.php

 

[RafaelMarujo]

Over here in the USA there is a simple solution to worn acme screws and nuts. We can buy off the shelf pre-made ones. Then we cut the old screw portion off and keep the area where the front part of the screw that fits into the saddle. then bore a hole in the front part, turn the new screw down to the bored hole to a press fit and cross drill install a taper pin or silver solder them together. Mill the acme nut to fit like the old one. http://www.greenbaymfgco.com/ACME-threaded-rods.php

I would say that the trapezoidal thread isn't worn out that much honestly