Beginners guide to lathe alignment and leveling

[choochoo_balooParticipant @choochoo_baloo]

Hello all,

First of all I am a newcomer to metalworking so please be patient with me!

After an off axis tailstock drilled and tapped hole, this made me think that the tailstock is not properly aligned with the head stock. After deliberating whether to tackle the more involved 2 collar turning method, I decided it safer to pay a bit more and buy a ready ground 2MT test bar (centre drilled both ends too). Photo of setup attached. If somone can address the following that would much appreciated:

1. Am I correct that it can be used for all 3 alignments: headstock, bed twist, tailstock alignment?

2. If so i thought that taper mating at each end would independently align them by running a toolpost clamped dial test along it. Which leads me to…

3. There is a strange periodic swinging of the dial over about a thou every ~1/16 inch or traverse. I can only think that the dial (0.0001" precsion) is actually picking up the helical cutting marks?

4. Videos I've watched always show the bar between centres. But why does it have a taper?

5. Please talk me through all the necessary steps to align the tailstock at the very least!

Thanks in advance.

Edited By choochoo_baloo on 19/12/2016 00:36:08

Edited By choochoo_baloo on 19/12/2016 00:36:50

[choochoo_balooParticipant @choochoo_baloo]

[Thor 🇳🇴Participant @thor]

I am no expert, but when setting up my lathes I have followed Harold Hall's advice. Since I don't have a MT test bar I first used a crude spirit level to get the bed as level as possible, then used a bar with two collars.

First I mounted the bar in the 3-jaw (no tailstock support) and adjusted the levelling feet to get bed and headstock aligned, you should be able to do that using your MT test bar mounted in the headstock MT (no support at the tailstock end) and the dial.

When the bed and headstock were aligned I mounted the testbar between centres and took a light cut and adjusted the tailstock. Since your MT test bar has centre holes at both ends you should be able to use that.

This procedure may take some time, but you get a lathe that turns accurate.

Thor

Edited By Thor on 19/12/2016 05:30:52

[Mark Simpson 1Participant @marksimpson1]

There's plenty written about actually checking alignment, can I just suggest you check the simplest stuff first…

Is the morse taper bore clean and not dented/bruised and check the headstock morse taper bush for dents and burrs (if it's got one)

All you need do is put a stripe of Black felt pen on the male tapers and gently turn in the bore… If the pattern is not consistent then sort this first…. The bores of my South Bend and Colchester student were both marked by people (before me) jamming in centres and collets without being properly cleaned… Very gentle work with an oil stone or taper reamer made a huge difference

It's worth doing this with your new test bar too, mine came beautifully ground but with a subtle burr at the small end….

If your test bar won't show pretty much zero error at the heastock end the rest is never going to be right.

Good Luck!

[HopperParticipant @hopper]

Aligning the headstock and tailstock to the bed axis is the last step in setting up your lathe, not the first.

Before getting out the test bars and dial indicators, you need to make sure your headstock bearings are set to the correct clearance and have virtually no detectable movement of the spindle. Easiest way to do that is to grab a piece of 1" bar a bout a foot long in the 3 jaw chuck. Put a dial indicator mounted on the bed (not the carriage) with the plunger touching on the top of the spindle collar just behind the chuck. Yank the end of the bar up and down and see how much the dial indicator moves. Anything more than half a thou is cause for concern. Repeat the procedure with the dial indicator plunger on the side of the spindle and yank the bar back and forth. Observe the reading. Same tolerance applies. If you have too much movement here, adjustment of the bearings is required, as per the manual. (It involves removing shims between the two halves, blueing and scraping etc.)

BTW, a dial indicator reading in tenths of a thou can be an albatross around your neck for this kind of work. As you have found it can be too sensitive and read surface finish rather than true movement. A clock graduated in half thou graduations is ideal. One thou increments is the common useage for home shop and will work OK for our purposes.

Once you have ascertained that the headstock bearings are tight but free running, it is time to strip clean and reassemble the carriage, cross slide and top slide, making sure all burrs and dirt are removed and that gib strips are all good and flat and engaging over a wide area.

Then you are ready to do the headstock and tail stock alignment. There are various ways of doing this. Some start by levelling the bed to get any twist out. But on a 40+ year old lathe like yours appears to be, probably not a lot of poing, because you will probably end up twisting the bed out of level to get it to cut true anyhow. So up to you if you want to spend a day with an engineer's level getting the bed level length ways and then cross ways at several points. Avoid laying the level on the worn section of the bed , ie the first quarter nearest the chuck.

Then you are ready to check alignment of the headstock to the lathe bed. This is done by putting the tapered end o f your shiny new test bar into the taper socket in the end of the lathe spindle. MAKE SURE all is immaculately clean and free from dirt and rag lint and burrs. No tailstock centre used in this test. Mount your dial indicator on the (locked) cross slide and move the carriage back and forth along the bed, with the plunger touching first the side of the test bar, then the top. This will tell you how far out of alignment the headstock is with the bed. Adjustment can be affected by adjusting the mounting feet of the lathe to twist the bed, as per the Myford manual. In extreme cases of bad alignment, you may have to loosen the headstock bolts and use the two jacking screws to align the headstock. It is a good idea before doing this to rotate the lathe spindle with the dial indicator bearing on the end and see it is rotating concentrically. If not, rotate the spindle until you are at the halfway mark in the swing of the clock hand. This will give you an average reading.

Aligning the tailstock is done last.

Put the test bar between centres, the dial indicator on the carriage and move carriage back and forth along the parallel section of the bar, with the plunger first on the side then the top of the bar. This will tell you how far out of alignment the tailstock is. Adjust tailstock where it mounts to the tailstock base to get zero offset in the horizontal plane. The vertical plane is not so critical. Many lathes are made with the tailstock a thou or two higher than the headstock spindle to allow for wear, sag etc.

You can also put the test bar taper in the tailstock socket and run your carriage mounted dial indicator along it with plunger first on the side then the top. This tells you if the tailstock quill is parallel to the lathe bed axis. Best to do this test before setting the offset.

In all probability, if you had a problem centre drilling, you may just need to do the last tests to align the tailstock where some previous owner has fiddled with it and left it offset a bit.

Either way, you are going to have to do a final test turning piece to make sure of the end result under real world conditions with working loads on all old worn parts. Turn a length of bright mild steel bar held in the three jaw with no tailstock centre and measure for parallel. This tells you the headstock alignment. Then turn a test piece between centres and measure for parallel. This tells you tailstock alignment. Test piece size is best about 1" diameter by 6" long mild steel, with a fine finishing cut. Rough turning down the midlde section of the piece to create the traditional two collars speeds things up a bit: not so long to wait for fine cut all along the full length.

There are other ways of doing all these things but this way uses your test bar you bought and should do the job.

Have fun!

 

 

Edited By Hopper on 19/12/2016 06:40:39

[Brian HParticipant @brianh50089]

I found this very interesting because it's something I've never done!

My lathe is a Mk3 Boxford AUD and I wonder if anyone can tell me what effect the factory made cabinet has and how any alignments are carried out.

Brian

[AjohnwParticipant @ajohnw51620]

Just adding to Hopper's comment. Check the top of the test bar first by running the dti across the top of it with the cross slide at several points. This should be done first really. The error should be very low.

Then the side can be done by moving the saddle along. Any height variation in the first test will cause some degree of error in this one.

The procedure should be done in the same order when you align the tailstock. However a point to note. Lathes for some reason are usually made so that the tail stock centre is marginally high. Not much just a little. I suspect this is done for final alignment when the bed may be twisted to correct any remaining small errors when work is between centres.

If there is height from bed variation put your maths head on and work out what this would mean to the reading along the side – if you like.

How small should the errors be. This should give you some idea – on a brand new lathe

The fact that this is for a boxford doesn't really matter. It shows the limits. When and if you get to the 2 ring turning test this also gives some idea of how much error they expect to correct over a shorter distance via straining the bed. Not much at all in other words.

A lathe that has seen real use wont meat these figures. Rather than straining the bed to try and correct it a better idea is to do some between centre turning with the centres a sensible distance apart and adjust the tailstock to remove any taper that the lathe turns. The lathe will turn slightly "oval" work when this is done but it's usually by a rather small degree that will be difficult to detect even with a mic. For oval it might be better to read not round. This should also help when work is held in a chuck plus support from a centre in the tailstock. A centre drill will produce a central hole even when it's just cutting on one side. This is why they don't look like drills and tiny ones can be easy to break. When a hole is drilled afterwards the hole made by the centre pulls the drill in unless it's a larger size of drill.

If things are out and the bearings are tight and ok it's probably best to shim the head to get the best results from the 2 ring turning test and to correct gross errors shown via the test bar when it's just in the morse socket, A distance of 6" / 150mm is usually used for this sort of thing as most work falls within that distance. The main problem when this is being done due to wear is that it may no longer be possible to bring the tailstock in. With this sort of problem the best first option where it can be done is to change the headstock bearings. That should then just leave bed wear problems. Once that gets to a level of a couple of thou or more up by the headstock the bed should be reground really.

John

–

[AjohnwParticipant @ajohnw51620]

Having mixed the Xmas cake and got it into the oven time for a bit more. Bit late this year but it usually lasts until March, April or even May.

I don't think levelling was mentioned. Somebody needs a cut and paste prepared spiel on this or maybe Niel could put a link on the forum pages in huge letters that goes to a short explanation of what it actually means.

Most people think it means get the lathe level, No problem but it doesn't really need to be level and an ordinary spirit level would be fine for that. It might matter if pumped coolant is used to ensure that the tray the lathe sits in does drain away.

It''s actually a method of checking a lathe bed for twist. It could be used on other parts of machines as well. It needs a very sensitive level. The saddle would usually be positioned at one end of the bed. The level placed onto it and then zero'd. Then the saddle is moved along the bed watching for bubble movement. If it moves the bed is twisted. I've also heard of it being placed on a parallel across the bed and that sort of thing.

The vials aren't easy to come by. BW electronics used to sell ones that were ex theodolite usually with military markings. I'm not sure how well these compare with the real thing. Some of the digital ones have a vial on the top. The one I have is more sensitive than the 0.1 degree digital reading but there isn't much bubble movement for those sorts of angles. As it's an analogue device though tiny movement can be seen easily.

John

–

[choochoo_balooParticipant @choochoo_baloo]

Thanks all, particularly Hopper, for the guidance.

I shall re-post if I get any further problems!!

[Author]

Posts