Myford ML10 headstock bearing play

[Dave HicksonParticipant @davehickson80498]

Hi, I am completely new to model engineering and have no previous experience of machining or engineering. I have acquired an old Myford ML10 (not a Speed 10). I also bought a 0.8mm DTI gauge from Ebay for £20. The scant knowledge I have comes from watching some Youtube videos and some internet research.

I appear to have some play in the headstock bearings. If I place the DTI against the headstock itself and apply firm manual pressure to the headstock then the gauge does not deflect at all. If I place the gauge against the spindle and apply firm manual pressure to the spindle then the gauge will deflect 1 division (0.01mm).

My question is this: is this level of deflection something that needs to be fixed? I can get a large chuck to register a similar movement of around 0.03mm, presumably because I'm measuring further away from the headstock? Could I get this deflection down to zero by refurbishing the headstock?

I seem to be able to take fine cuts of 0.001" in Aluminium rod, but completely lose reference once I remove the rod from the chuck. I would like to buy some collets to give repeatability, but am I wasting my time due to the play in the bearings?

Many thanks for your help.

[Dave HicksonParticipant @davehickson80498]

[Thor 🇳🇴Participant @thor]

Hi Dave,

Welcome to the forum. If you remove the rod from the chuck and then rechuck in a different position you shouldn't expect it to run perfectly true unless you have an expensive chuck. I would expect a collet chuck to give better repeatability.

If your lathe is an early version with a hardened spindle running directly in a split bearing you might try to remove a shim. I don't think 0.01mm is much so I don't think you need to remedy this just now. Try and make a few things and see of the lathe can produce the work you need and to the tolerances you want. Have you checked here?

Thor

 

Edited By Thor on 29/12/2017 17:54:51

[Jim GuthrieParticipant @jimguthrie82658]

Posted by Dave Hickson on 29/12/2017 10:18:50:

 

I appear to have some play in the headstock bearings. If I place the DTI against the headstock itself and apply firm manual pressure to the headstock then the gauge does not deflect at all. If I place the gauge against the spindle and apply firm manual pressure to the spindle then the gauge will deflect 1 division (0.01mm).

 

Dave,

0.01mm is 4/10 of a thou which is not too bad for an old lathe. I suspect that you are shifting the oil layer around. . I've got an original ML10 and I suspect that I would get a similar reading. You can tighten up the head bearings by taking out the clamp bolts and removing the spacing and removing one layer from the spacer in the clamp slit then reassemble and re-tighten to get the bearing tighter. But I suspect that to get a free running spindle you might have to leave a bit of clearance – maybe .01mm.

I note that you also quote thous somewhere else in your message and I guess that you might be suffering from what I suffer from when bouncing between metric and imperial – i.e. I expect precision to have at least three decimal places after the point and I often forget that 0.01mm is less than half of 0.001".

Jim.

Edited By Jim Guthrie on 30/12/2017 09:20:16

[Mick B1Participant @mickb1]

I'd agree with the above. I'm sure my Speed 10 was at least as bad, and I made 2 working engines on it. Don't worry about it till you can't achieve the precision you need for what you're trying to make.

What comes out yer shed is more important than what you got in there.

[Dave HicksonParticipant @davehickson80498]

Hi Thor, Jim & Mick,

Thanks very much for the welcome and info, all very helpful indeed! I will leave the bearings alone for now and just start machining and see how things go. I also plan to buy an MT2 collet holder and some ER collets to explore repeatability.

Thanks again,

Dave

[HopperParticipant @hopper]

Posted by Dave Hickson on 29/12/2017 10:18:50:

The scant knowledge I have comes from watching some Youtube videos and some internet research.

You need to get yourself at least one good book on using the lathe. "The Amateur's Lathe" by LH Sparey is my favourite. There is an overwhelming amount of bumf on the internet that such old basic books cut straight through.

If you have only .01mm movement in your spindle, you are doing well. I would leave the headstock bearings well alone.

As far as repeatability goes, do you have a four jaw chuck? If so, use that. It's not hard. But really, it is most often a matter of planning your steps to machine a job so it is all done in one set-up. Sparey's book and others like it have lots of tips on that kind of thing.

[Mike CrossfieldParticipant @mikecrossfield92481]

Dave,

If you are buying a collet chuck for your ML10 you might want to consider one which screws onto the spindle register, rather than a MT2 mount. The MT2 version is fine for holding milling cutters etc, but very limiting in terms of the length of stock material you can hold.

[Dave HicksonParticipant @davehickson80498]

Hi Hopper & Mike,

Again, all sounds like great advice. I do have a 4-jaw which came with the machine, and I will definitely go for the spindle register collet holder.

Many thanks,

Dave

[Jim GuthrieParticipant @jimguthrie82658]

Dave,

Watch out for direct spindle mounting collet holders. I tried to acquire an ER25 one some years ago and I couldn't get one with less that 0.005" runout after two swaps with the retailer. I finally finished up getting a faceplate mounting holder and a threaded backplate from (the original) Myfords and doing my own machining and fitting.

Jim.

[SillyOldDufferModerator @sillyoldduffer]

Assuming you're stuck with a 3-jaw:

Trick one: I can't believe it's an original idea but I use a DTI whenever I put work back in the chuck. Put the part back in the chuck and apply the DTI to the machined part. Look at the error while turning the chuck by hand. Slacken the jaws and rotate the part relative to to the chuck and remeasure. Repeat until the error is as close to zero as you can manage.

Trick two: If you know in advance that the work is going to be moved, mark it and the chuck with a felt-tip so you can put if back where it came from. It's still worth checking with a DTI.

Trick three: plan the cutting sequence to avoid moving the part in the first place! (When I first started metalwork I often moved work. After many painful disappointments, I've learned to think hard about doing the job with the smallest possible number of moves. Without solid references, re-acquiring register is troublesome and error-prone. On complicated jobs it may be worth making a jig or fixture just to provide repeatable references.)

Dave

[Steve WithnellParticipant @stevewithnell34426]

The runout on my first lathe was nearly 0.1" and would still make perfectly good parts! Really bad runout will still deliver round parts From a novice point of view (which I will always be at this game), knowing how your machines perform is more important than seeking perfection (which you will never achieve) in them.

Don't buy a collet chuck to make the runout go away – most turned parts are produced in either a 3 or 4 jaw chuck with only a few specialist jobs needing a collet chuck. I have all three – but 75% of jobs are done in the three jaw. 20% are maybe done in the 4-Jaw. A few are done on the face plate. Doesn't leave a lot of jobs that NEED a collet chuck.

I'm definitely with Jim and Hopper on this one.

A more important check might be to see if it is actually turning parallel, rather than taper and work on reducing any taper. This is easy to do – just stick 100mm of 25mm steel bar in the chuck and take a light skim over it's length and see it the start and end diameters are the same. If not – someone in the group who knows the ML10 better than me will advise next steps!

Happy New Year

Steve

[Author]

Posts