Lathe alignment and cross slide play issues

I have just acquired a new lathe, a WM250 type. It seems to be very robust and reasonably accurate, however I have two questions/issues that I was hoping someone could comment on.

Issue 1: Excessive play in the cross slide lead-screw. By pushing the cross slide back and forth I can move it ~0.7-.8mm. Is there any way to limit this play a little? It is not a major problem, mostly irritating.

Issue 2: Bed alignment with head-stock. I have carefully skimmed an unsupported 100mm length of 25mm free turning steel and measured the diameter at both ends. There is a difference of 0.02mm (smaller at tail) between the diameters. Is this something to worry about? Reading Harold Halls lathework leeds me to believe that this 10x more than acceptable. How would one go about figuring out what needs adjustment, and what are the adjustment possibilities on this type of chinese lathe?

Regards,

Geir

Edited By Geir Ertzaas on 17/02/2013 16:23:19

Npt familiar with this model but on most of them, if you remove the handle, you will find a nut which controls the play on the thrust bearing. It sounds as if this has come loose. On my Chinese mill I had to drill and tap this nut for a brass locking screw.

Can't help with headstock adjustment.

Russell

Geir, I think my lathe is a smaller version of yours. To adjust the free play on the cross slide feedscrew, I just tighten up the Nyloc-type nut which holds the handle on, to close up any gap between the rear face of the dial collar and the front face of the bracket behind it. If yours is the same, it's a very easy adjustment to make.

Lining up the headstock so that the spindle is parallel with the ways is rather more complicated. If your lathe is a bigger version of mine, the headstock will have two Vee grooves on its underside, which locate it on the same prisms on the top of the bed which locate the saddle and tailstock. I think re-aligning the headstock would mean shimming or scraping those Vee grooves, reassembling, taking a test cut, disassembling to try different shims or do more scraping, re-assembling, and so on. The procedure is described, but with only one Vee groove to deal with, on GadgetBuilder's website, here. He relies on "Rollie's Dad's Method" to check alignment, but that is not foolproof. For example, it can't distinguish between a badly aligned headstock and a warped bed.

For the sake of 0.01mm over 100mm, I wouldn't bother in case things were made worse!

I hesitate to suggest it, but as your spindle seems to point ever so slightly to your side of the tailstock, you might try a light tap with a soft hammer on the front of the headstock, at the pulley/changewheels end.

Andy

Sorry; forgot to add the LINK to GadgetBuilder's site.

Andy

Geir,

I think the amount of taper that you have measured along the axis is acceptable. The most likely cause is deflection of the the bar by the tool as the cut is being made. This could be the bar bending but is more likely to be caused by the bar deflecting in the chuck jaws. If you put the test bar in the chuck and press on it do you see a deflection with a dti? Mis-alignment could also be caused by twisting of the bed which may depend on how it is fastened to the bench/stand. However, as I said, I wouldn't worry about it.

cheers,

Rod

Are you sure your reading is ten times more that what Harold says as 0.002mm would be 0.000079"

How much backlash is there in the handwheel, there are adjusting screws in the nut under the crossslide which may want tweaking. Also check the actual nut is not loose, there should be an allen head screw that is probably hidden by your topslide that holds the nut to the underside of the cross slide.

J

Thanks for the quick replies, they are very appreciated. I am tending towards not worrying overmuch about this right now, especially as the procedures for "fixing" could easily do more harm than good. However I am still curios as to how one determines the most likely source of the misalignment.

Rod:

I agree that deflection might be an issue, however I tried to take very light skimming cuts (dust) to eliminate this as problem. However I see that there is still a good probability that I am still measuring deflection. I did use a carbide insert cutter, and I from what I am reading these are not ideal for these kind of light cuts, however the finish was excellent as far as I could see (feel).

The way I understand it the test bar must be unsupported and that deflection must be handled by using an appropriately thick bar and take light cuts.

The lathe is set on a very sturdy wooden bench that has been additionally reinforced to provide a stable level surface. The lathe is fastened by two bolts, one at the headstock and one at the tail end.

Jason:

In the "lathework" book, HH states: "With repeated machining and checking, one should be aiming at a virtually error free result, certainly no more than 0.002mm on diameter over 100mm length". I must agree that I thought this a little on the extreme side as well, as any dust or minute deflection would make this virtually impossible to measure and maintain.

The handwheel seems to be nice and tight, I will look for the nut you mention.

Edited By Geir Ertzaas on 17/02/2013 18:22:45

Edited By Geir Ertzaas on 17/02/2013 18:23:18

Geir

I cannot totally agree with Andy as making adjustments to the headstock mounting are a very last resort and certainly not for an error of only 0.02mm

For most normal day to day use of the lathe 0.02mm over 100mm is quite good, Even so, to minimise an error of that size the method is to make minor adjustments to the mounting feet, typically, if the test piece is large at the outer end then raise the front foot at the tailstock end and if small the rear foot.

In my book though you will see I was attempting to set up for making a cylindrical square for which 0.002mm would be required, ideally. If your lathe then has jacking screws at each mounting then aiming for 0.002mm is not that extreme as very minor adjustments are easily made. However, if using shims then the task can be tedious and something more like 0.005mm is perhaps more realistic.

I realise that stating 0.002mm in my book is being rather precise as few could measure that difference with certainty. I feel I should have said "with no more than a just noticeable difference with the micrometer readings"

If you want more detail on setting up a lathe to turn parallel then see, as I have said in another thread going at the moment, my website here http://www.homews.co.uk/page309.html

Harold

PS I went for tea half way through writting this should have appeared before Jason's item

Edited By Harold Hall 1 on 17/02/2013 18:27:08

**LINK**

The authors suggests that "personally they would be satisfied with 1 thou in 4 inches". As the error is doubles when you measure diameter, you are getting a better result than that.

Good advice on testing and setting up at thaht link.

Neil

**LINK**

The authors suggests that "personally they would be satisfied with 1 thou in 4 inches". As the error is doubles when you measure diameter, you are getting a better result than that.

Good advice on testing and setting up at that link.

Neil

I did conclude by saying that I wouldn't bother, for fear of making things worse!

Andy

Jason:

Your advice helped fix the lead-screw play - the nut was completely loose.

Harold:

Thanks for the helping clear this up - 0.002mm is probably good advice when making cylidrical squares. I was thinking of attempting to make some, but I can see that this may have to wait a little. The lathe is mounted on a wooden bench and inducing a twist will, as you mention in your article, require something substantial to push against. The squares will have to wait for a new bench/frame.

Is twisting the bed always a good fix for minor alignment errors? The chinese lathe I have is fixed to the bench using three screws that run centrally to the bed. One at the tail (as depicted), and two at the head-stock. Inducing a twist will involve shimming on either left or right side of the screw. The "box" structure of the bed frame will probably be quite resilient to twisting without applying considerable force, and I wonder how effective shimming will be? It may be that actually re-aligning the head-stock, as per Andys comments, is the most viable option?

Regards,

Geir

Hi Geir,

I have taken the liberty of using your photo and edit it a bit, hope that is OK.

Underneath the lathe you bolt pieces of flat steel - say 12 to 15mm thick (red in pic.) and drill holes at front and rear end for height adjustment bolts. Harold Hall has description of how this is done on Myford lathes. As you say a sturdy bench is required. The adjustment of the bed (twisting) you need to make will probably be very small.

Regards

Thor

I have some basic lathe knowledge, and enough engineering background to see when I am getting out of my depth . I will let this matter rest for a while and proceed in the following order: make lots of simple stuff, make even more simple stuff and then sometime in the far future build something awesome

Only if the bar is deflecting, if the tool is being pushed off the work then it will be smaller at the end, a bit like how a boring bar gets pushed off the work - less cut as it gets nearer the chuck.

J

Edited By JasonB on 18/02/2013 10:01:52

Chris,

Well spotted. Jenkins: Read the question boy!

Rod

Issue 1: Excessive play in the cross slide lead-screw. By pushing the cross slide back and forth I can move it ~0.7-.8mm. Is there any way to limit this play a little? It is not a major problem, mostly irritating.

-You can eliminate the issue by making a digital cross slide with a 10 dollar vernier

Issue 2: Bed alignment with head-stock. I have carefully skimmed an unsupported 100mm length of 25mm free turning steel a... etc

-Get into the habit of using the tailstock to support work whenever you can

Wooden bench. Give it a while to settle under the weight while keeping an eye on the humidity. And is the floor concrete or wooden boards like an interior sprung matress? If you are into woodworking you might find it easier to make an adjustable foot for the bench. You only need one to impart a twist.

Don't forget to lock the cross slide when taking the test cut.

Do you have a dial indicator? Try setting it on the end of the bar and pushing down on each corner of the bench and see if it moves.

Don't even think of moving the headstock. There was a guy on here last year moaning like mad about his lathe who did that then after a few weeks found it was something trivial he could have fixed on day one if he had listened ot the advice.

Cross slide play is annoying as it can affect repeatability. It probably can be adjusted out - and note some lathes have a cross slide nut that can be adjusted (and therefore can come loose). Patience and common sense should help track this down. If its a worn feed nut, replace the nut. If a new nut is slack than the problems is more of an annoyance than a problem, as the backslash should not noticeably affect repeatability. Do check that a loose cross-slide gibb isn't creating the impression of a slack adjustment elsewhere.

As for the parallelism, 0.02mm in 100mm isn't a bad result for a beginner.

It's a mistake to test something, look at the result and say "is this good enough?"

What you should do is think "How good do I need this to be?" and then test and see if it comes up to your requirements.

In practice something like turning a long shaft parallel will depend more on issues like the material, depth of cut, tool sharpness anbd tailsstockl support and alignment as much as anything else. As Michael says you can get perfect results on an old mangle if you apply the right techniques and patience.

Harold's test square is about as demanding a piece as you can imagine. A long shaft possibly only needs to be dead to size at the two pojnts where it goes through the bearing, for example.

So don't fret about chasing 'tenths of thous' get on with making something and building your skills. by the time your lathe starts to limit your results, you will probably know just how to fix it anyway.

Joey