Aligning a mini-lathe

I meant to leave #1 snug, not tight. Loosen #2&3 and lightly snug them Tap the headstock in the direction you need and tighten#1, then #2 & 3.
 
mikey said:
I meant to leave #1 snug, not tight. Loosen #2&3 and lightly snug them Tap the headstock in the direction you need and tighten#1, then #2 & 3.
Click to expand...
Thanks, that helps.

Since the piece was smaller at the tailstock than at the chuck, I want to rotate the headstock away from me which is counter clock wise, correct?
 
Yes, I think so. Try it and see.
 
Ok, after disassembling a lot, I made an attempt at correcting the horizontal error. Had #1 snug but not full tight, loosened #2 & #3 a tad. Tapped the headstock near #2 to rotate the HS CCW. I tightened #1 first, then #2 then #3.

Took another 0.001 pass. Really didn't believe what I was seeing. (0.7 tenths) So I took another pass of 0.001, reasoning that making another cut shouldn't make any difference, but should remove effects of the headstock adjustment a little. The second set of measurements was better.

If I believe my second set of measurements (avg of 7 measurements at TS, and 8 at the HS) the HS is 1.96044" and the TS is 1.96086". This is a difference of 0.00042", with the TS being larger (still). I've reduced the error by a factor of 2. It took me a while (6 readings each) before I really settled down taking measurements. (On the second set.) Had to get a feel for the right sensation as I rotated the micrometer off the maximum. If it didn't feel right, I looked at the measurement, but rejected it. After a dozen measurements or so the feel seemed to be more consistent.

Too be honest, I'm going to wait a bit. Wait for the iron to relax. Then take another cut and measure again. 4 tenths isn't bad for this class machine, although I'd like it a tiny bit better. Tapping the headstock with my machinist hammer seemed like it was doing nothing. But it took off about 0.0005 of the error! From my former, non-machinist point of view, this would be black magic. Now I know better.

Need the level for the next step.
 
Okay, at least we're making progress. The important thing is that you now understand how to do this. I did want to point out that bolt #1 is your pivot and once you move the head, locking down that pivot first will hold your settings so that tightening bolts 2 and 3 will not move the head.

How much more accurate the alignment needs to be is up to you. You may get to the point where you may not have enough play in the hard points to allow for much more accuracy. At that point, you either have to accept it or take out your file.

This was a good lesson, I think. There are a number of ways to do most everything in this hobby but the ones you trust are the ones you test yourself and prove them to be valid. This might be one of them.

EDIT: Whatever you do, do NOT remove your rod from the chuck. This is a first operation and as long as the chuck is not loosened then you can rely on the results from subsequent cuts. I know you understand this; just making myself feel better by saying it.
 
Last edited:
mikey said:
Okay, at least we're making progress. The important thing is that you now understand how to do this. I did want to point out that bolt #1 is your pivot and once you move the head, locking down that pivot first will hold your settings so that tightening bolts 2 and 3 will not move the head.

How much more accurate the alignment needs to be is up to you. You may get to the point where you may not have enough play in the hard points to allow for much more accuracy. At that point, you either have to accept it or take out your file.

This was a good lesson, I think. There are a number of ways to do most everything in this hobby but the ones you trust are the ones you test yourself and prove them to be valid. This might be one of them.

EDIT: Whatever you do, do NOT remove your rod from the chuck. This is a first operation and as long as the chuck is not loosened then you can rely on the results from subsequent cuts. I know you understand this; just making myself feel better by saying it.
Click to expand...
On the fence on further alignment. One wouldn't think much more movement is possible. However, I thought that initially, and got 5/10ths improvement.

Understood on moving, tightening, un-tightening whatever from the chuck. It's good to restate the obvious.

Tomorrow, I'll take another cut to see if the answer is substantially the same. If it is, then for giggles, I might do the RDM again. Ideally, the numbers should be smaller than what was measured before. Might be a revealing test.

Level is on order. Taking a chance on a less expensive one. If it doesn't work out, I'll get something else.

For the tailstock what do you suggest? I've seen the test with a bar between centers, taking a cut at both ends. Is that what you recommend?
 
WobblyHand said:
For the tailstock what do you suggest? I've seen the test with a bar between centers, taking a cut at both ends. Is that what you recommend?
Click to expand...

No. I have an article I wrote some time back on how to make a tailstock alignment test bar. I'll try to clean it up and post it.
 
I would agree that you don't know for sure how true your lathe is cutting, until it is leveled. I just got in a Starrett Machinist level (today) and it showed me something that surprised me. My lathe rests on a 3/4 inch marble slab, which rests on a wooden workbench. When I took the level to the marble slab it was flat, but the wooden table underneath was "sway backed". i.e. The middle was a notably low point, with the wood surface slanting upwards towards the legs. It made me glad I sprung to get the marble slab for under the lathe. My bullet level I had used before, lacked the resolution to see this.
 
addertooth said:
I would agree that you don't know for sure how true your lathe is cutting, until it is leveled. I just got in a Starrett Machinist level (today) and it showed me something that surprised me. My lathe rests on a 3/4 inch marble slab, which rests on a wooden workbench. When I took the level to the marble slab it was flat, but the wooden table underneath was "sway backed". i.e. The middle was a notably low point, with the wood surface slanting upwards towards the legs. It made me glad I sprung to get the marble slab for under the lathe. My bullet level I had used before, lacked the resolution to see this.
Click to expand...
Wood tops just don't hold up under machinery. They are pretty and all, but they sag. I had a solid core door on saw horses for a while. I had a small drill press on it. I was amazed at how much sag there was when I replaced that set up. Steel stands/benches are so much better.

Carpenter levels are not very sensitive. The level has to move quite a bit before the bubble seems to move. One really needs a much better grade level to measure the tiny differences we are trying to correct.
 
WobblyHand,
I could not agree with you more on the flex of wooden tables. I was just surprised how visible it was with a higher resolution level, it was painfully obvious. I live in a very small town, and purchasing metal stock here is very pricey. For the most part, you are forced to get it from a hardware store (If you are unwilling to take off work and drive 3 hours round-trip to get to the "big city". I had a local guy trying to sell some 1/2 inch steel 24" by 48" plate, but it would be fair to say he wanted a premium price for it, very premium. I may end up having to build an edge up 2x6 bench if a good steel source cannot be found (fingers crossed).
 
You must log in or register to reply here.
Back
Top