MT5 Test Bar

[mikey]

@mikey I think we've talked about this before.
Our definitions of leveling differ and I may be wrong what leveling is.
To me leveling is checking to make sure the lathe is level as far as no movement of the bubble from HS to TS and that's it, over, done.

Using a precision level is enough to get your lathe leveled pretty close, I agree. However, to get it leveled to closer tolerances a 2 collar test is usually utilized. When the OD's of the 2 collars is the same then the lathe is level and not until.

 

[mikey]

It just makes no sense to me to unload a new lathe from the trailer and immediately start taking cuts to check for HS alignment.
I'm going to level it first to take undo stresses out of the casting. Probably let it sit for a few days and check again. Then I'll start taking test cuts.

I agree. One should do a rough level before starting but doing a precision alignment is not necessary at first.

You have to start somewhere ...

 

[ddickey]

What's a rough level?

 

[Just for fun]

Okay I won't assume.
I'll say I know he would concur.
@Just for fun Is your lathe the TL 1440 or whatever Precision Matthews calls it.

My lathe is a 1236T from Precision Matthews

I'm going to level it first to take undo stresses out of the casting. Probably let it sit for a few days and check again. Then I'll start taking test cuts.

That's what I have did, kind of. I have never checked the headstock. I did align the tailstock to the headstock.

 

[pontiac428]

My lathe manual says to let the machine settle (under operation) for a week, then re-level weekly until set. It's an 8-point test.

The reason we "prove" our lathes for concentricity and perpendicularity is because we depend on those relations when doing work. The two-collar test (headstock only) is the best way to do this proof. It also gives us the data to know "how much" when things do need to be moved. Anything less than "spec" would yield a taper. We measure things using tenth-thousandth resolution, as if all those numbers to the far right of the decimal have meaning. We should apply the same precision to our headstock/bed alignment, no?

 

[mikey]

What's a rough level?

That's just a quick level with a Starrett 98 used along the axis of the cross slide at both ends. I did this on several lathes I helped to set up but for my own lathes I did not even bother with a preliminary leveling. Just went straight to aligning the headstock and had no issues leveling the lathe later.

 

[ddickey]

Okay I will eat crow and say I was wrong about Mikey agreeing with what I said. I respectfully disagree. A proper leveling with no bubble movement with a precision level before taking any test cuts is the way I'll do it. I can't wrap my head around taking a test cut with a lathe that isn't perfectly level.
I'll ask this, how do you know if you're cutting a taper because of an out of alignment HS or twist in the bed by not having a proper leveling? I'm totally open to discussion if I'm missing something. I greatly respect Mikey's knowledge on this subject.

 

[pontiac428]

I'll ask this, how do you know if you're cutting a taper because of an out of alignment HS or twist in the bed by not having a proper leveling?

Exactly. That's why you eliminate leveling first by leveling the lathe. Using an indicator on the top and front of the test piece will tell you more, but once you've isolated the bed you can focus on the head if needed.

 

[Cadillac]

You can check headstock alignment by facing. It should produce a flat cut or minuscule concave cut. Now your saddle could be off but we will assume it’s perfect. You could also put a indicator on your spindle nose and take readings off your cross slide to give you a idea. To be most accurate you’d take the cross slide off put pins in the dovetail and sweep them along the dovetail. That would take all variable out.

 

[mikey]

Okay I will eat crow and say I was wrong about Mikey agreeing with what I said. I respectfully disagree. A proper leveling with no bubble movement with a precision level before taking any test cuts is the way I'll do it. I can't wrap my head around taking a test cut with a lathe that isn't perfectly level.
I'll ask this, how do you know if you're cutting a taper because of an out of alignment HS or twist in the bed by not having a proper leveling? I'm totally open to discussion if I'm missing something. I greatly respect Mikey's knowledge on this subject.

You and I usually agree about everything and there was nothing wrong about what you said.

Intuitively, you would think that a lathe has to be level to test something so it makes sense, right? What I have found is that my intuition can be wrong so I try to test everything before assuming something. In this case and on my lathes, I have found that a carefully leveled lathe is not necessary to align the headstock with the ways; I merely want the lathe to cut in alignment with the spindle. I have also found that, for me, it is far more fruitful to align the headstock first and level the lathe second. This has been borne out by testing on other lathes I've set up as well as my own.

With that said, I don't think there is a need to be right about anything. Test it one way or another to see what works best for you and do it that way. What matters is that you are satisfied with your results.