ML10 - ready for the boneyard?

If I had an old lathe that was within .00069" of alignment horizontally over 26cm (10 inches) I'd consider myself blessed. Hell, if I could even accurately measure anything to within .00001" in my home workshop I'd consider myself blessed.

Vertical alignment is nowhere near so critical. Until you get down to very very small diameters, the surface of the job where it meets the tool is virtually vertical so the tool being a tad higher or lower makes very very little difference to the turned diameter of the job. GH Thomas discussed this at length in one of his books, I forget which one.

I'd leave things well alone until I took a fine test cut over a four to six inch length of 1" bar and see how it measures up.

Yes, you will get some droop over 26cm of 14mm bar. Quite a bit in fact. ISTR in Connelly's MC Tool Reconditioning book he says a piece of 1" bar 6" long will droop by .0004". I would imagine almost double the length at almost half the diameter would have considerably more.

Re oil film "sag", when making these types of precision measurements, it is customary to run the lathe for 20 minutes at medium to high speed first, to get things up to operating temperature and conditions.

Let us know how your turning test on a piece of 1" bar measures up. Will be real interesting to see how it compares with the static test already done. Theoretically, based on the deviation of .00069" already measured, it should be pretty good.

from your figures I think you have conclusively proved there is nothing much wrong with your lathe! You would be hard put to find anything affordable in the hobby market giving much better results than < half a thou out at 6 inches, whether you think it is wear or alignment causing the issues.

Try to bear in mind also that the spindle also needs some freedom to move, so I don't think there is excessive bearing play either- less than a thou 10 inches out? come on!!! how much better do you expect it to be?

Time to stop chasing your tail and start having fun making something useful.

OK... I'll find some time later on this evening or early tomorrow to run a test piece for parallel.

Just ran my first cylinder turning test using 25mm free machining stock.

I chucked it up nice and tight in the 3-jaw, skimmed off until I was turning clean everywhere and finished up with a final pass, one thousand deep using some cutting oil.

The measurements with my micrometer were taken along a 9cm length, starting 1.5 cm away from the end of the chuck jaws. Each measurement was taken 1.5 cm further on towards the tail end of the lathe.

The readings were:

24.3125mm -- 24.335mm -- 24.353mm -- 24.36mm -- 24.363mm -- 24.355mm -- 24.351mm

This gives me a cylinder whose diameter varies by 0.0505 mm ( = 0.00199 inch ) over a length of four inches. But, in truth, the major portion of that variation is occuring over only half that length.

Hopper suggests that 0.001 inch of difference over a length of six inches would be good for an old machine, so these results don't seem so very good to me.

Not only is there more variation than desired, but of course, most of it is within 4 cm from the end of the chuck... right where one would most expect it on an older machine... and right where one does the majority of one's work.

Edited By Ignatz on 07/08/2018 16:14:35

... Oh yes, I forgot to mention that even though the turned tolerances of my test piece is disappointing, the surface finish of that piece of free machining stock didn't look so bad.

What are you using to measure down to .0001mm with? You really can't measure that accurately in most home shop environments without temperature control and long experience at using such sensitive instrumentation. Normally, micrometers and dial indicators with a resolution of .01mm is used for this type of work. More sensitive instruments cause problems due to user error, mounting flex etc.

Working to the usual hundreths of a mm, your readings would be something like

24.31 -- 24.33 -- 24.35 -- 24.36 -- 24.36 -- 24.36 -- 24.35

So for us fossils who still work in imperial, that works out to a variation of .04mm or .0016", which is pretty good for a lathe that has not yet had the bed mountings shimmed.

What you need to do next is put a thin shim under one of the lathe's mounting feet at the tailstock end. To correct for a test piece that is larger at the tailstock end like yours is, you will need to put a shim under the FRONT of the lathe mounting at the tailstock end. (The lathe mounting where it is bolted to the bench top, not where the lathe stand is bolted to the floor.). Thickness of the shim is by trial and error. You might start with .005" thickness. Beer can material is traditional! Cut a square about 40mm square and then cut a slot into one side of it about 12mm wide, so the piece ends up a U shape that you can slide under the mounting foot with teh U surrounding the mounting bolt. Then tighten down the bolts and take another test cut.

Another point is that it's best to use a very sharp high-speed-steel toolbit for these tests. Using carbide insert tooling on such fine cuts can provide false reading as the insert can tend to rub rather than cut cleanly.

Also, you might try holding the test piece in the four jaw chuck if you have one. It holds more firmly than a three jaw and is usually less likely to be bell-mouthed slightly, which can skew readings toward larger at the tailstock end.

For a full description of the bed-mounting shimming process, see page 13 of the Myford ML7 User's Manual. It is freely available for download on the internet as it is long out of print and copyright is no longer enforced by the holders. More information here **LINK**

Edited By Hopper on 08/08/2018 05:25:00

Hopper,

I'm using a regular metric micrometer, divided into the usual hundred gradations per millimeter. The extra digits are my estimates of the 'between' position of the scale around the reading mark. I agree with your rounded off figures.

The cuts were indeed taken using a very sharp HSS cutter. As I mentioned, the actual surface finish on the steel was rather nicer than on 'scrap bin' stock.

After running this test I got my eyes down low and up close and personal with the underside of the front dovetail on the bed way. It didn't look so good. Evidence of long, deep scratches there, doubtless the work of cutting swarf between gib and dovetail. Under the finger it feels rather rough, certainly in comparison to how baby smooth the dovetail feels at the (unused) tail end of the lathe.

Yes, I could try some shimming under the bed, and although it might help that portion of the cut out away from the headstock, I have my doubts about the area closest to the chuck. One would be asking for rather a fair amount of twist - and just along that short section of bed - to correct for the difference. I'm really thinking that this is a case of accumulated wear over time.

Yes, there will be wear over time. No doubt about it.

But still worth reading the User's Manual on the shimming procedure recommended by Myford, and tested with a four to six-inch test piece so it must influence performance up close to the chuck, which I agree seems surprising but is the way it works.

Plus tightening up the headstock bearing may take up some of the taper turning tendency.

It's worth a try.

Otherwise, just start using the lathe and see how it works in practice. A few seconds with some emery tape on the job at full rpm will remove one or two thou of taper on critical jobs.

.

Ignatz,

Despite Neil's reply, I was quite serious about the advantage of a replaceable bed-plate ... and I think your current experiences demonstrate the point nicely.

I don't know the detailed dimensions of each profile, but I think you will find that they are sufficiently similar to demonstrate that the arrangement would have been practical; and may even be feasible as a not-inexpensive retrofit.

If the existing dovetail edges were machined off the ML10 bed; a top plate, in the Hardinge style, could be fitted: This would increase the centre height of the Myford, but retain its working profile. This might be worth doing on your Myford, but it obviously depends on how good the rest of the components are, and whether you are on good terms with a grinder !!

MichaelG.

.

For inspiration: http://www.babinmachine.com/index.php?HARDINGEDV59

Edited By Michael Gilligan on 08/08/2018 08:25:49

Following Myford's standard procedure for shimming the bed feet to get the lathe cutting straight would be a lot easier.

Myford's ML10 instructions:

So I'm left with a few choices.

Tighten up the main journal a wee bit and try to shim the bed a bit. Couldn't hurt and might help. At that point I would just continue to use the lathe as it is until a better option (new or used) presents itself.

Totally agree that a replaceable bedplate would have been just the ticket for the ML10, most especially given the fairly straightforward profile. But that is all in hindsight as Myford chose not to take that route. And, yes, one could elect to have that sort of thing done, but this demands access to larger machine tools which I do not have. Paying for same - let alone finding someone with the necessay expertise for such an adventure - would seem to be prohibitive.

Regrinding the bed is always possible. Pursuant to that I got in touch with Myford Solutions in Holland. The bare minimum would be to true up the bed way with a surface grinder, clean up the dovetails on a Bridgeport and similarly treat the carriage and the gib. The price for this is at least €500 and that is not even allowing for the cost of shipping the bed and carriage up and back (since I don't have a car) . If I went further and had the man do both the cross slide and tool post slide it would bring the cost up above €750. Pretty high considering that the feed screws and nuts would still be doubtful... and again, that is a hefty chunk of the cost of a new lathe.

I'm hoping to perhaps find someone in my own area (Gent, Belgium) to do the work, but we don't enjoy the same healthy engineering hobbyist community as up in the UK so my search may be both long and in vain.

Because you'd presumably have to dimension the superimposed bedplate to suit the existing saddle assembly, I can't see that you'd effectively achieve the first half of the highlighted sentence - unless you considerably reworked saddle and crossslide.

And the second half - retaining the working profile - wouldn't be much of an advantage. The Myford 10s are already steering close to the minimum limits of a geometry that gives adequate support for general use. Both bed and crossslide are narrow when working with the larger diameters the 10 series can accommodate.

I'd agree with other posters that there doesn't seem to be a helluva lot wrong with the machine as it stands. But using it as is with a bit of ingenuity to get around the issues to make things you want is one thing; embarking on a substantial project for limited benefit quite another.

Unless of course you do it because you like the work for its own sake - that justifies almost anything in a hobby like this!

.

Mick,

I think we are at cross-purposes ... sorry if I wasn't sufficiently explicit.

When I wrote 'If the existing dovetail edges were machined off' I meant exactly that ... i.e. the top of of the Myford casting would now have vertical edges and be of reduced width ... allowing the new plate to have the same profile as the original.

MichaelG.

May I suggest you get a piece of accurately ground silver steel bar, chuck it, and measure its diameter at 1.5cm intervals to 0.0001".

You will get a series of varying readings, unless something odd is going on.

As the bar shoudl be of consistent diameter along the full length this should give you a clearer idea of how accurate you measurements are. For example, if you are consistent down to a thou or half a thou.

Be brutally honest and don't try to get the same reading each time. A hand-held micrometer is NOT actually the easiest of ways to measure accurately, much easier to put the mike in a stand and bring the bar to it, for example.

Only once you are confident of how small a deviation you can accurately measure can you really make a judgement on your lathe.

Once close to alignment, there's a simple test - get a ball race with a bore of, say, 0.500", or anything with a decent clean bore to it. See if you can turn a bar to a shake-free sliding fit in the bore along its length. If you can, then the lathe is a good 'un and well set up.

Neil, the resolution of the micrometer I have at hand is 0.01mm (discounting an eyeballed .005 in-between marks) which only equates to 0.00039 inch, this being much coarser than your suggested level of accuracy.

Nevertheless, I shall do as you suggest with that length of 14mm silver steel rod I used for the 'Rollie's Dad's' test.

Edited By Ignatz on 08/08/2018 10:51:11

Get yourself a scraper and have a go at the bed dovetails.

Martin.

That is what I thought you meant. So your new plate has dovetailed external sides to match the Myford original, vertical internal faces to fit the modified casting, and enough horizontal thickness on top to maintain its profile.

But presumably you're going to fit the headstock, tailstock and saddle to the new plate in the same or similar manner to the way they were fitted to the original bed, so I can't see how you'd change the centre height or anything else in the basic geometry. And unless you took the bed top surface down far enough to maintain the same height as the original, wouldn't you run into the leadscrew bracket issues in earlier postings?

Edited By Mick B1 on 08/08/2018 11:03:49

WTH? Scrape the bed, or regrind the bed, or remachine and fit a plate because the initial turning test is about half a thou more taper than what would be ideal, before the bed has been properly "levelled" by shimming under the feet as specified by Myford in their set-up procedure? (On the page following the one Michael kindly posted above. Sorry I don't have the tech to copy and post the needed page but I have posted a link to it above.)

Sledghammers and mosquitoes come to mind.

Edited By Hopper on 08/08/2018 11:14:56

.

Further apologies, Mick ... I was about to edit my last post to include reference to the leadscrew, when the roofer arrived at the house to discuss some work.

As for the centre height : Yes another example of bad wording on my part ... The centre height above the bench is increased, but the working centre height within the machine is not.

It's all totally academic in the real world of Ignatz, so let's not labour it any further.

MichaelG.