View Full Version : Newb help with tail stock alignment

05-08-2012, 10:14 AM
I off set my tailstock to turn a long taper this weekend, and I need it back centered. I've adjusted it three times, and will adjust till it's correct. However I realize that I may be chasing my tail too. What is realistic accuracy for a 12x36 chinese lathe (probably makes no difference). I've got a stainless shaft that I'm turning 12.5" and taking measurements on both ends. I've got my .0005" indicator out using for a adjustment guide. With using the witness mark I got it back to .0022" diff.... now it is at ...

My test bar measures 1.1743" at the head stock end, and 1.1757" at the tail stock end. (live center if it matters) The tail stock needs to move toward me (or the tool bit) how much? .0007" ?

05-08-2012, 10:45 AM
I use a DTI as shown below to set the tailstock.



05-08-2012, 01:51 PM
My test bar measures 1.1743" at the head stock end, and 1.1757" at the tail stock end. (live center if it matters) The tail stock needs to move toward me (or the tool bit) how much? .0007" ?

That would be my guess. ... ,0007

I don't know what kind of lathe, nor what kind of tailstock adjustment you have. But I know that with some, e.g. the SouthBend 13" lathes, the adjustment on the tailstock would "creep" slightly between the time you make the adjustment and the time you get it locked down. That can have you chasing your tail around for awhile if you don't accomodate it and let the adjustment "settle in."

I guess "stiction" is the term for it. ...or maybe "hysterisis"

Forrest Addy
05-08-2012, 03:48 PM
Here's what's served me well over the years. Chuck a short remnent from the drop box by the saw. Turn a short diameter the size of the tail stock quill. Get on it with a dial indicator with the contact on a horizontal plane with the axis. Extend the tailstock quill. Without disturbing the indicator run the carriage to the tailstock and contact the extended quill. If the indicator reads zero, good, if not adjust it. Then go back to read a repeat zero on the chucked remnent. If there is a small difference in diameter between rem and quill, compensate for it.

This would also be a good time to check for quill alignment and droop, effect of clamping and releasing the tailstock.

No test bars, no fixed center distances; quick and simple. Takes but a few minutes. Less if you already have a quill sized piece you can dial in.

Jaakko Fagerlund
05-08-2012, 04:48 PM
Quickest way is to just put a dial indicator in the chuck and put it like in the above picture by GadgetBuilder.

The Artful Bodger
05-08-2012, 05:15 PM
Yabut...... I was told that the test bar is the best way as that measures the actual results. Also, moving the tailstock down the bed aways (ha ha :D ) reduces the error that would come from spindle not perfectly aligned with the ways (and we all know nothing is 'perfect').

The same old codger told me that setting the tail stock to gnat's whisker accuracy is only relevant if you will be turning between centres, so between centres is the only way to test.

05-08-2012, 09:58 PM
Here's an interesting method for quick, accurate alignment. It's part of the Tubal Cain series on Youtube. http://www.youtube.com/watch?v=1EZh8SUoA4o

05-08-2012, 11:05 PM
I do about the same as Gadgetbuilder but I don't put the indicator on the taper, I use the side of the shaft. All you have to measure is front and rear.

All this does is get the tailstock roughed in. To get it right you have to set it while turning something between centers because every time you move the tailstock it will change a little.

If your looking for as close as you can get it has to be set with light cuts on the work piece. It may take a few passes to get it cut without a taper but it's not hard to do.

Forrest Addy
05-08-2012, 11:47 PM
Did I miss something in the Tubal Caine vid. Usually he's thorough without being wordy. I didn't hear him mention droop in connection with his indicator and I didn't hear anything about wear in the tailstock's baseplate. The method he demonstrates is excellent and doubles any error present but it seems to presume zero wear and perfect vertical alignment in the tailstock.

05-09-2012, 04:48 AM
I use a DTI as shown below to set the tailstock.




You will or may get "indicator sag" with different readings for a correct setting at top and bottom but probably OK at the front and back.

The Artful Bodger
05-09-2012, 05:58 AM
OK, here is yet another way for consideration!

Put a bit of scrap in the chuck and drill a hole in it, enlarge the hole a little by turning. The hole should be less than the diameter of the centre in your tail stock.

Mount a DTI on the tail stock bearing against the side of the partly extended quil, move the tail stock up to the chuck and extend the quil a little more to cause the centre to enter the hole in the scrap.

If the tail stock is off alignment the off centre pressure will cause a deflection on the DTI.

Retract tailstock quil and adjust offset until deflection is minimal or nil.

05-09-2012, 11:25 AM
any tailstock will have some misalignment of the quill, therefore it can only be aligned accurately for a given extention of the quill. on the other hand, if you know what the quill is doing (i have a chart for several positions), you can correct a taper in the work by changing the extention of the quill (if the length of the work allows it).

Forrest Addy
05-09-2012, 01:05 PM
Here's where we get into the ways and means of tailstock alignment. It's not an easy task but doable for the home shop guy if he pays attention. Look in the archive. Some of us have written lengthy tutorials. Look in the archives.

05-09-2012, 02:25 PM
where are the archives?

05-09-2012, 05:18 PM
Frank Ford's trick:


05-09-2012, 05:41 PM
Search function at the top of page just yielded approx 264 topics using "Tailstock Alignment" as the search term.

Forrest is correct, this has been discussed many times, and to do a proper job, it is not easy. The tolerances in the alignment are a lot tighter than you think, the vertical tol. is +.0005" high in 12" on a test bar stock in the taper, and +- .0005" in 12" horizontal on the same test bar, the test is done with the spindle retracted and extended, but it is a little easier to align than the carriage.

05-09-2012, 06:55 PM
Considering the true nature of the offset at the chuck , using the indicator is a fine way to get it close. By close I mean the chuck has run out.
The ways need to be exact for this to work at the length you are asking.
You got to think of the measurement as the Dia at the chuck and the dia at the talestock. you need to divide by two to get the result of .00035 offset; .0007 for both sides. you will be real lucky to get it straight the entire length of the part.

05-09-2012, 08:17 PM
That's where I'd start. If you really want to be fussy you wouldn't want to use the chuck at all as it will have it's own error. My personal preference is that of turning a long piece to verify the alignment. The other thing to consider is how the movement is set up on your tailstock. On the lathes that I've had it was simply opposing set srews that were used so getting them both tight at the same time while the tailstock is in the right location is a bit of an art.

I saw a really great post here last week from a fellow over in England who was using a boring head in his tailstock rather than moving the tailstock itself. I thought that was one helluva clever thing to do, so much more control on the adjustment for the taper and not screwing up the setting for the tailstock.

The Artful Bodger
05-09-2012, 08:32 PM
Maybe I am doing it wrong?

When I want to turn a long piece I put it in the chuck, centre drill the end and bring up the tail stock.

I turn it as required then part it off from the chuck. After moving the tail stock of course.

If this is a new part of overall reduced diameter I dont see that the chuck runout has any effect?

05-09-2012, 11:37 PM
Methods have been presented in regards to checking the position of the tail stock center. Since you have need of altering the position of the center from time to time, I'm going to question the method you have of making this adjustment. First of all, is it an error prone, sloppy 'tap here a bit, check- tap there, check' type of thing? It's kind of like that on mine. The upper part of my tail stock is shimmed to height and to get the tailstock axis parallel to the spindle axis. There are some set screws to move the upper part front to back, and there's a keyed slot to keep the thing roughly in place. Of course you can rotate the thing in the horizontal plane because it's a bit sloppy- needed of course so you have a way of setting the rotation correctly to align the axis in that direction.

Part of the solution to your problem might be in finding a better way to have it adjustable in one direction without losing alignment in other directions. The tail stock has at least four alignment considerations. In no particular order, although order is important- the ram axis height should be the same as the spindle axis, the ram axis should be parallel to the spindle axis, and it should line up with it vertically and horizontally. It takes both shimming and rotation to get it parallel, then it takes raising or lowering, plus moving front to back to get it in line. Likely the only adjustment you would want to do on any regular basis after that is front to back. What you need to arrive at is a way of making that front/rear adjustment without throwing any of the others off. When you can do that, then you can have a stop in place so you can return the front/rear adjustment to that stop, and then know that you've got it back centered without a lot of testing being required.

My last pearl of wisdom here- unless the ram axis is perfectly parallel to the spindle axis, you cannot move the ram in and out without throwing off the position of the center. You need to achieve this first before you get the height and front/rear adjustments made.

05-10-2012, 01:05 AM
Feel free to tell me I'm an idiot for even suggesting it, but...

Let's say you have a spare Morse taper blank arbor. Let's say the blank end is cylindrical (or you make it so). How well would the tailstock be aligned if you:

Chuck the blank end in a 4-jaw and indicate it in (or use a collet).
Loosen the tailstock adjuster(s) so it's "floating".
Slide the tailstock socket over the protruding arbor taper and ram it firmly home.
Tighten the tailstock adjuster(s) while the taper is thus seated.

Would you end up even close?

05-10-2012, 01:15 AM

I didn't have a specific goal in mind... just as close as possible. I made the adjustments this morning, and turned my 12" test rod again. Both ends were gaged multiple times to verify accuracy. I now have it down to .0006" and am calling it good enough!

I'll be chambering a bbl this weekend, and would obviously like to have it "perfect" but, I think it's close enough for "guv-ment" work.....

Thanks to all who replied. It is always appreciated.

Brock in Arkansas

05-10-2012, 01:34 AM
Presuming that the tapered end didn't wobble, and the head stock was aligned to the ways, then yes, you'd be close. Securing it in that position would be the bigger issue. Things move when you tighten bolts- much will depend on the particular tail stock.

But I have touched on another important thing- the head stock alignment, and the bed alignment. If you look at the whole machine from the standpoint of 'getting it all aligned', you have a fair amount of work to do and many things to consider. Just to check the head stock alignment, you'd probably have to start at the floor. As Forrest suggested, the topic has been covered earlier, so I won't go any deeper into it now.

Jaakko Fagerlund
05-10-2012, 03:56 AM

You will or may get "indicator sag" with different readings for a correct setting at top and bottom but probably OK at the front and back.
If your indicator is that heavy that it sags under its own weight in the situation pictured in the photo, then it is not a good indicator for this purpose.

Forrest Addy
05-10-2012, 01:25 PM
Jaako, before you dismiss indicator sag as unlikely run a few inversion experiments. Gravity is always present and gravity causes anything extended horizonaly - bosoms or solid steel - to droop. Tramming tailstock quills is a classic indicator problem. Whatever sag in presnt doubles in the indicator reading.

The indicator set-up in the earlier post seems absolutely rigid but it ain't at least not perfectly. If you have a apparatus I suggest you make the same set up and record the reading. Then make the same set-up separately where the indicatoe is rigidly clamped to and contacts something rigid like a sort piece of square tube or solid stock a 1 1/2 (40mm) sqare or round x 1 foot (300mm) long, Set a zero then invert the set-up and note the indicator reading. It may be a small reading (too many variables for even a guess) but significant.

Horizontal spindle machines for all their convenience have to contend with gravity. So when you dial the OD of a tailstock quill from the lathe spindle you have to be aware of gravity. That's why I reccomment the turned diameter in the headstock and scan from headstock to quill by running the carriage back and forth (see my earlier post). The indicator holds its orientation to gravity and thus its effect is cancelled

Jaakko Fagerlund
05-10-2012, 01:44 PM
I'm quite aware of sagging, I use a horizontal milling machine / long hole drilling thingie a lot at work and the DTI stups have to be pretty firm from time to time.

BUT, as I said, the situation in that picture is more than rigid enough to get proper readings.

Forrest Addy
05-10-2012, 04:49 PM
Sorry Jaako. I know you know your stuff. I thought I dedected a note of disparegment in your post Quixoticly I chharged.

05-10-2012, 05:16 PM
There, I fixed it for you. Now I don't wanna hear anyone whine about headstock sag because I can't afford the rocket to send this lathe into space. (Machining in 0g! fun fun..)

05-14-2012, 03:16 AM
I can check the tail-stock alignment as well as the head-stock alignment with a 20mm (~3/4") and 150mm (~6") O1 (silver steel) rod test bars and a 0.001" dial indicator (a better indicator will be better but the 0.001" indicator will suffice.

It only takes minutes to make the test bars and less to use them.

I will detail it later.