ML10 - ready for the boneyard?

Got it. Here's what you need to do, in addition to taking about .001" of movement out of the headstock spindle by simple adjustment.

.

< pedantry alert >

The text that I posted is from the ML10 manual, but your link is to the ML7 manual.

There is, I believe, a significant difference in the torsional stiffness of the two beds; and the ML10 is not so amenable to adjustment ... The ML10 instructions only reference 'without distortion' [suggesting that the ML10, as shipped, was expected to be 'level' in the free-free condition].

MichaelG.

Maybe so. Maybe not. Nobody knows until they try. We are, after all, talking about a very minor taper on a test piece here. My point is, basic engineering trouble-shooting, eliminate the simple stuff first before even considering a $1,000 major engineering project that is way beyond the scope of a beginner who has not yet learned the basics of using a lathe.

Exactly .. but you claimed readings to 0.001mm with it!

In a-level chemistry I was expected to subdivide a 0.1ml reading to 0.01ml by eye with reasonable accuracy, but I wouldn't have staked my life on the results!

Neil

Neil, you are right and the point is well taken. I was only trying to do the best with what I had at hand. And Hopper was quite correct in simply rounding off my measurements.

Neil, I did as you suggested, chucked that bar of 14mm silver steel in the lathe and measured it up and down its length, every 1.5 cm.

My measurements did NOT vary, the micrometer reading out 13.985 mm at every position. Note: I'm adding on that last '5' because the reading was always between 13.98 and 13.99 on the dial.

Then went digging around in a forgotten tool chest under the bench and found an old Lufkin micrometer (model 1641V) that is calibrated in ten thousandths (had been the property of a friend's dad who was a machinist in Milwaukee). The Lufkin gave me a reading from that silver steel bar of 0.5505 inches ( = 13.9827 mm )

The readings from the two micrometers are very close to one another... and as far as I'm concerned, the micrometers agree.

I then re-measured the free machining steel test bar I had turned. The Lufkin micrometer gave me almost exactly the same readings as I had obtained earlier (but in inch measure), showing a total deviation along the test turning - low to high - of 0.0022 inches

So for whatever it is worth, my measurements seem to be valid... disappointing, but valid.

Sounds as though you have talked yourself into wanting a new lathe, so why not do it and save yourself some time?

From the numbers and description, the bed appears to be worn slightly at the chuck end. The amount of error wouldn't worry me much, but its seriousness depends on what the lathe is used for.

Quite a few lathe errors can be compensated for by the operator. Apart from being on the small size, I can't think of anything I do that couldn't be done on this slightly worn ML10 unless something else is also wrong. Admittedly much of what I do doesn't require everything to within tight tolerances.

Ignatz - what are you intending to do with your lathe that needs it to be spot on? Have you tried to make a real part and failed? The symptoms might suggest a work-around.

Dave

I was hoping to move a bit beyond the simple down-and-dirty garage maintenance stuff... perhaps into small engines if the spirit moves me.

This whole thread started because my attempts at turning a shaft between centers rendered such disappointing results. Some of the issues that dogged me have been partially resolved thanks to all the helpful comments. Others need still be addressed.

I might have made more progress with my tests and adjustments if it hadn't been for the rather nasty heat wave we have (all) been suffering from.

Anyway, as I stated in a previous post in this thread, I intend to continue to use the lathe as it stands until I can either arrange for a regrind or purchase a (larger) lathe in better condition.

So far, the theory of a worn bed is purely that, a theory. Based on pure assumption, which although a good guide for the direction to look in, is not in itself a reliable guide to what actually is.

Bed wear is easily measured in at least two ways.

1. Set the carriage at the unworn tailstock of the bed and tighten up the gib adjusting screws until the carriage is slightly tight to move up and down the bed by hand. Then move the carriage down to the suspected worn spot near the headstock end. See if the carriage becomes easier to move and if it has noticeable side to side shake or movement possible. Try inserting a feeler gauge between the carriage and the angled surface of the dovetail. With a feeler gauge you should be able to measure how much wear is on the bed. Most of the wear is usually on the rearward-facing surface as the tool is pushed toward the operator during cutting use.

The other way is to measure across two small cylinders, say 25mm long each, places on the dovetails as in the below scenario. Again, compare between the unworn and worn sections of the bed. Diameter of the cylinders must be such that contact with the dovetail surface is somewhere around the middle as in the drawings.

However, even if there is bed wear, a certain amount of it can be compensated for on most lathes by shimming the lathe mounting feet or even adjusting headstock alignment.

IN their old brochure about their reconditioning service for the ML7, Myford says more than .003" of wear on the bed in the horizontal direction is time for a regrind. The implication being that up to .003" could be compensated for by  twisting the bed via shimming the feet. (Note that they allow up to

Edited By Hopper on 09/08/2018 01:09:47

Edited By Hopper on 09/08/2018 01:12:25

Edited By Hopper on 09/08/2018 01:15:02

Hopper,

I have run the first test you suggested.

I ran the carriage down to the tailstock end, snugged up the gib screws and then brought it back towards the headstock. The movement of the carriage became progressively easier (read looser) towards the middle of the bed way until, very close to the headstock, it became even tighter than at the tailstock end, very difficult to move.

Testing with a bit of shim stock between gib and dovetail proved inconclusive due to limitations, both in my ability as well as the variety shim material thicknesses available to me.

The second test you suggest - measuring across bar material in the dovetail groove will not serve me well here.

The design of the bed of the ML 7, having flat ways with squared off sides means that the wear of the bed on the top is not linked to wear of the sides and/or gib.

On the other hand, on the ML 10 the dovetail design of the bed way means that wear either in the dovetail against which the carriage and gib bear or on the top of the bed itself will both have an effect on the back and forth carriage movement.

You can see this from my little illustration below.

The dovetail test in the image you attached to your last post indicated a reading across test bars placed down in the grooves formed between dovetail and bed.

That would be thus mean reading across bars riding in the grooves a/b and d/e

Unfortunately, the carriage does not really bear on those grooves, instead referencing itself to the corners formed by b/c and c/d

If we wear away the top of the bed way (exaggerated here for clarity) the measurement across the rods in the grooves previously indicated will still remain the same (for all intents and purposes), but the actual cross-section of the bed as experienced by the carriage will have been reduced. This is presented by the lowest image where I have superimposed the worn cross-section of the bed way over the original.

So, at this point my quest is pretty much at an end...

If I can find someone locally who can do the necessary regrind of bed way and carriage at an affordable price, all well and good.

Failing that, I will continue to use the lathe for what it is until such time as I choose for another.

I am also intending to lap down the headstock shim to see if I can remove some of the play in the front journal. I will try to take appropriate photos and document the procedure (putting it into a new thread) as it might prove instructive for someone else in the future.

Thanks again to everyone for your tips and suggestions.

For the amount of wear that seems to exist in the bed, it wouldn't surprise me at all if a quick session on the shears, with a well sharpened scraper sorted out your problem.

I'd use a straight edge and some 'blue' to gradually take off enough material to let the carriage move back and forth sweetly. It would be hard to remove too much material without being reckless.

I didn't catch how much play is in the spindle, but remember that once it's rotating, an oil film will build up within the bearing and remove quite a large amount of play.

Martin.

Martin(blowlamp),

Your suggestion is well meant, but...

I think that once I had...

• Obtained a sufficiently long, precision machinist's straight edge or granite surface block (at great cost)
• Paid money for scraping tools and bluing
• Taught myself to scrape to any reasonable degree of proficiency (doesn't happen overnight)
• Disassembled the lathe and...
• Spent God knows how long actually attempting what you suggest (whether I succeeded or not)

... that I would probably wish I had spent the money for a regrind or a brand new lathe.

Edited By Ignatz on 10/08/2018 20:23:32

Correct, my suggestion is well meant - I've used it on numerous occasions to bring equipment back into shape.

1. Use the side of the Cross Slide as your straight edge if you can't find anythig else. I don't know why you think you need a granite surface plate.
2. Borrow a scraper from someone or make one from an old file. A tin of Stuarts Blue.
3. Take a look on Youtube, but in a nutshell, if you can push a scraper diagonally through the centre of the blued high spots and remove a reasonable amount of material, then you can scrape. It doesn't need to look pretty when done.You can check as you go by moving the carriage over treated areas.
4. Remove the tailstock, move the carriage out of the way and tilt the lathe to a good working angle.
5. You've been testing and posting about this since the end of July - a day or so scraping will be time well spent.

I have to say that as a newcomer to model engineering, this thread has been very enlightening. I've looked at any amount of old lathes on Ebay, and now I'm thinking 'Do I really want to enter this minefield?' as most of them will likely have at least the issues that the OP has, and probably many more. I would like to rebuild an old machine at some stage, but I'll need a newer or new machine to do it with I think...

I have to admire the OP's tenacity in checking exactly what the wear is on his machine, but I think the time may have come to either suck it up, or splash the cash. If you ditch the old Myford, then by all means leave it somewhere I can find it!

Good look whatever the outcome, and do show us either the results of the re-grind, or the new lathe.

Edited By RevStew on 11/08/2018 09:54:39

The only thing this thread demonstrates is that it is difficult to make meaningful measurements that reveal one, and only one, error. The readings on the test cuts generally get larger towards the tailstock. If the bed was worn near the headstock, so the saddle dropped, you would expect the readings to get larger towards the headstock.

Personally I'd be making parts, not looking to blow money on an alternative.

Andrew

+1 on that. ^^^^

And bear in mind that if you buy a brand new low-cost hobby lathe, they have been known to arrive with their own problems that need a bit of "fettling" before use. It's not unheard of for some of them to turn more of a taper than this ancient example if the bed has not been levelled or the feet shimmed to get it turning straight, or if the headstock has been bumped in transit etc.. I've personally seen one that came with a .007" bow in the bed but that's probably an extreme example (hopefully!)

Edited By Hopper on 11/08/2018 11:46:07

Martin,

You do have a point. Layout blue is cheap and, yes, I could make scrapers out of old files. Still, I wouldn't feel comfortable without at least a long, precision ground straight edge as the minimum reference tool.... something like the Veritas precision straight edge **LINK**

Even then, considering that I have absolutely zero experience with scraping I wouldn't have confidence in myself to necessarily improve the situation, that for a whole bunch of reasons.

So I'm not thinking to try that at the moment. Nevertheless, it does remain an option. Thank you.