Trying To Align My Lathe

[BigSpike]

I checked the lathe & it is level across at the chuck end & at the tail stock. Since the bed is solid, I have a 1-2° slope to the tailstock to allow the bed to drain when using flood coolant.

I went & bought some 2" aluminum round. I will turn both ends of the test bar equally & bore a couple of 2" sections to press on each end, then I can try the collar test.
I will check back in a day or two when I get this done. Thanks all for the help

 

[Tozguy]

I'm just sharing the "standard practice" as I've read it in multiple books. You're not taking a heavy cut for the test. That would completely defeat the purpose. Assuming you already know you're close you could do it with a vertical shear bit. The tailstock and headstock will be the deflection points since you're running on at least the tailstock center and only barely cutting the collars right where they are well-supported.

Obviously it will work either way. The idea, to me anyway, is just to minimize the work of rechecking it later on (such as after setting up a taper, for instance), so you will be more likely to do it since you know it will be fast and easy. Set it up once and you never have to touch the center section of the test bar again.

The twin collar test shown in the video above is NOT run between centers. After this test has been run and the lathe corrected if necessary for twist in the ways, there is seldom a need to revisit it. This test is basic and should be run even if the best level in the world tells you all is OK.

The second twin collar test, run between centers to reset the tailstock after turning a taper, is another matter. It is only reliable if there is no twist in the ways. The OP has already done the equivalent of this but he found a contradiction between the test result and a sweep with a dial indicator. There is something else going on from the look of it.

 

[Uglydog]

I checked the lathe & it is level across at the chuck end & at the tail stock. Since the bed is solid, I have a 1-2° slope to the tailstock to allow the bed to drain when using flood coolant.

Thanks for responding to the level in plane question. Sloping shouldn't be a problem as long as the ways are level with each other.
Took me a long time to understand this concept.

Are the flats and ways worn unevenly? Please, let me ask this differently.
When we check for level we often put the levels or parallels on the highest spot. Are all the flats and ways at the same height?

Also, I'm curious if the height of tailstock is correct relative to the headstock. In effect creating a Y axis, which lathes presumably don't have.

Is this phenomena an acute change in results, has it been gradual, or is this lathe new to you?

I don't pretend to have the answer here. But, I'm fascinated by machine tool geometry and therefore by the puzzle.

Daryl
MN

 

[BigSpike]

the lathe is practically new, in other words I bought it new & have hardly used it
setting a height gauge on the ways tells me that the tail stock at a couple of points is the same height above the ways as the center in the chuck
I will tackle installing the hubs on the test bar this weekend

 

[intjonmiller]

Like I said, I'm just repeating the advice I have learned from multiple sources.

From Rudy Kouhoupt: "The test bar that we're making here is really a tool in its own right. It's a useful accessory that you will protect and put away and you will always have it to use for your own machine, and if a good friend of yours down the street tells you 'I can't get my lathe lined up' you can take it down and do it for him in a matter of a few minutes and he'll probably be your friend forever."

That's roughly 15 minutes into his video on Fundamentals of Lathe Operation. I just went through it again to make sure I was remembering it correctly. And he does turn it between centers. Which makes sense because that is the most accurate way to work. Everything else is a compromise to some extent.

Here's the funny thing: I had forgotten that his test bar is FAR simpler. He preps the ends of his 10" long 1" round stock by holding it in his 3-jaw and steady rest, facing and center drilling each end. Then he turns it between centers and turns down "10 to 15 thousandths" off of the tailstock end. And that's it. That's his entire test bar. Now he removes the driving mechanism and replaces the test bar (machined end at the tailstock now) with just the centers and mounts a dial indicator in his compound. He zeroes the indicator on that machined section, then reverses the test bar and measures at the tailstock. Now he can adjust the tailstock with a live reading of the results. When it reads well he again reverses the test bar to check it all over again.

No further machining of the bar is needed. Brilliant!

I can't believe I had forgotten that. I saw that video before I read the two collar test in multiple texts.

The two tests are not identical, though. They have overlap in function, but the two collar test will help diagnose other issues, as were described in the initial post here, that Rudy's test bar won't reveal since it is only machined once on one end.

 

[GadgetBuilder]

There's a simple way to find horizontal tailstock offset using a DTI in the chuck and a dead center in the tailstock, as described here: http://www.gadgetbuilder.com/Lathe_Align.html#Tailstock_RDM_Align (scroll up one paragraph). This will get it within a thou or so pretty easily.

If you're enthused and want to get closer than a thou and/or want to measure vertical offset then a shop made test bar and a different approach is described at the above link. The test bar I prefer is made from a line printer shaft - these are accurately ground to diameter so all that is needed is to center drill the ends; length is arbitrary, 6 to 10 inches is good with a diameter of 1/2" or so. It is worth fussing to get the holes as precisely on center as possible -- a collet chuck or a 4 jaw or a 3 jaw with aluminum foil shims... whatever it takes to get the work precisely centered for drilling. I used a graver to "catch the center" since I've found this gets the starting divot exactly on center and then a gentle approach with a sharp center drill locates into the divot nicely. However, if either or both center holes aren't precisely on center you can use the averaging method known as RDM to achieve the desired measurement accuracy, see the description linked above. Despite my best effort I was about a half thou off with my center holes so I used RDM to complete the measurement.

John

 

[JLDavid]

ABom79's video of the 2 collar test really cleared up my confusion on the 2 collar test. Was having hell wrapping my head around that. So my next question is at what point do you move the headstock or do you just tweak the ways to get it cutting straight without a taper? This is all new to me so it may take me a while to understand it all. Thanks.

 

[JLDavid]

So how big of a piece of round stock would you use in a g4003g lathe? 2" big enough? Maybe a foot long?

 

[BWS Arms]

I use 2.5" OD 6061 round bar with 10" sticking out from the chuck jaws. You want a large OD so that it wont sag.

 

[JLDavid]

I have a piece of 2 1/4" aluminum heavy walled tubing that I found under a cabinet yesterday. Think that will work? Walls are .446" thick.