Master precision Machinist level

[SouthernChap]

Don’t mean to pressure you on the goofs, just don’t want to see you make the same mistake again.

I think I said in another thread that confession is good for the aspiring machinist soul.

Also, if some other beginner reads about my snafus, they might learn to avoid those mistakes.

Oh BTW how are you finding TouchDRO? It's something I'm considering for the lathe if I can find a scale small enough for the cross slide on this thing.

 

[tim81]

I think I said in another thread that confession is good for the aspiring machinist soul.

Also, if some other beginner reads about my snafus, they might learn to avoid those mistakes.

 

[BWS Arms]

I have a 98-12 and a 98-6 that I use to level the lathes and mill. I have not used the .0005” level in years. The starretts are good to use since you’re going to do a two collar test and adjust based on it.

 

[RaisedByWolves]

Oh BTW how are you finding TouchDRO? It's something I'm considering for the lathe if I can find a scale small enough for the cross slide on this thing.

Best DRO I’ve used, and I’ve used a lot of them.

 

[ddickey]

I have a 98-12 and a 98-6 that I use to level the lathes and mill. I have not used the .0005” level in years. The starretts are good to use since you’re going to do a two collar test and adjust based on it.

If you adjust your lathe to cut equal on the two collar test this does not mean your lathe is leveled. What if your headstock is out?

 

[Dabbler]

If you adjust your lathe to cut equal on the two collar test this does not mean your lathe is leveled. What if your headstock is out?

The process of setting up a lathe in the 'right way' (how I hate that phrase, but forgive me) begins by aligning the head, and then usually leveling the bed. Finally aligning the tailstock; Then and only Then, doing the 2 collar test.

So the 2 collar test and the lathe bed 'leveling' are different steps in the process. I hope this helps.

If you start with the 2 collar test, your are only producing an equal cut for those two places at that distance only, not necessarily over the entire bed. This is why it is important to do the other steps first, to get as close as possible to having the headstock and the tailstock collinear.

 

[SouthernChap]

The process of setting up a lathe in the 'right way' (how I hate that phrase, but forgive me) begins by aligning the head, and then usually leveling the bed. Finally aligning the tailstock; Then and only Then, doing the 2 collar test.

Now, that's useful. I was going to ask just that question about what order to do the steps in.

Once you've levelled the bed, I'm guessing you might have to adjust headstock alignment again, would that be right?

 

[RJSakowski]

Now, that's useful. I was going to ask just that question about what order to do the steps in.

Once you've levelled the bed, I'm guessing you might have to adjust headstock alignment again, would that be right?

Headstock alignment is similar to bore sighting a rifle. You want the spindle axis to be parallel to the ways of the lathe. Unfortunately. the way we determine that condition will use a dial indicator on the carriage and twist in the bed will alter the reading.

There are test bars available which fit the Morse taper socket and project the spindle axis out to a distance where a reasonable measurement of a non-parallel condition can be made. The test bars offer no advantage over using the RDM method for this determination. which is why I use the latter.

There is still leaves the issue of separating out bed twist from headstock misalignment since both have a similar effect. The way I do this is to check cross slide alignment with the spindle axis. If the headstock is properly aligned, a facing cut will be perfectly flat. While this can be done with a facing cut, similar to the two collar test, i prefer a test analogous the the RDM method. I mount a slightly domed boss on my faceplate at a distance that can be reached to both the front and the rear of the spindle axis. Then I mount a test indicator on my cross slid so that I can intersect the center of the boss when the boss is horizontal with the spindle axis. I zero the test indicator. Them I rotate the faceplate 180º and move the cross slide so the test indicator contacts the same spot on the boss. The reading on the indicator tells me how much out of alignment my headstock is.

Because this test is made with the carriage very close to the headstock and it is locked down, bed twist and wear issues are minimized. Bed leveling can be done before or after this test as there is very little interaction. The best strategy would be to do it before and check it after.
Following headstock alignment and leveling the I would address bed twist using either the two collar test or the RDM. I prefer the RDM test as it is simple and fast. If I do use it though, I confirm the alignment with a two collar test as a final.

 

[Dabbler]

Once you've levelled the bed, I'm guessing you might have to adjust headstock alignment again, would that be right?

Not necessarily - what I mean is that if you had *significant* twist in the lathe bed, over the 2" or so of aligning the headstock may have resulted in a distorted reading. Then you would have to realign the headstock, but this should be a very minor adjustment.

@RJSakowski managed to beat me to my reply. As hinted by his answer, the test bar method is not normally the 'go-to' for headstock alignment. I have little faith in these bars. I have heard rumours that a Moore and Wright or Browne and Sharp test bar are great, but I've never seen one. Stay away from offshore ones. I have never heard of a good one, but probably a unicorn does exist.

I agree the RDM method/face cut is a great way to go. This gets you one step closer.

Before you do the 2 collar test you have to align your tailstock first. Your tailstock centre has to be in the center of rotation of the spindle, or you get an erroneous reading.

[update] a for instance: My Leblond toolmaker's lathe tailstock is .003 high on centre, which make it fail the 2 collar test. I compensate for that by using a MT3 boring head and indicating the centre to the rotation before doing the 2 collar test.

 

[RJSakowski]

Before you do the 2 collar test you have to align your tailstock first. Your tailstock centre has to be in the center of rotation of the spindle, or you get an erroneous reading.

I don't use the tailstock when doing the two collar test. The tailstock just adds another variable to the mix. I use a rigid enough bar tha deflection is minimal and make very light skim cuts.

For aligning the tailstock, I mount a round bar between the headstock chuck and the tailstockot that it is important that the bar be fixed center and use the RDM test to adjust for zero (or desired taper). Note that it is important that the bar be fixed at the headstock so it rotates with the spindle. Otherwise and bend in the bar can affect the test.