Lathe Accuracy

[John Lucas 5Participant @johnlucas5]

[John Lucas 5Participant @johnlucas5]

Looking through ads and web sites for lathes suitable for model work. Interestingly the ONLY product I could find with specifications for accuracy is the Taig micro lathe.

Seems odd to me that other manufacturers do not give these numbers. When requested the typical answer is "its good enough". At the model engineers' show in December I asked a representative about the Myford Connoisseur. I was told that they are all different and after paying for the lathe I would get a specification sheet!

Obviously every manufactured item will be different but modern manufacturing will mean all will be with a very narrow range. Surely manufacturers could specify "better than" values.

John

[Nick_GParticipant @nick_g]

.

From what I have read. (so not gospel)

Some import suppliers issue an 'accuracy report' with each machine at the time of delivery. However when these have been tested by owners that actually know how to test a machine these reports often transpire to be figments of imagination that would make Thunderbirds look like a factual documentary.

I should however state that I have no idea if these 'testers' have an axe to grind for some reason and I also understand that different importers have different standards of what they will accept from the manufacturers. (even though the machine looks identical)

I have a Boxford. A precision lathe it is not, nor was ever intended to be. But it is more accurate than I will ever be.!

Nick

[Michael GilliganParticipant @michaelgilligan61133]

John,

I think you will find that Cowells also specify an "accuracy"

The hard truth is that [done properly] it would cost more than their selling price, to test and certify most hobby lathes.

MichaelG.

Edited By Michael Gilligan on 30/01/2015 17:47:20

[John Stevenson 1Participant @johnstevenson1]

The accuracy isn't in the machine.

It's in the bloke on the handles.

[Neil WyattModerator @neilwyatt]

"The user of the modern lathe, however simple or unpretentious it may be, should always remember that the capabilities of his machine, for accurate work, are far greater than that of the best lathes used by the craftsmen of a few decades ago."

Edgar T. Westbury, 1936

[Brian Rice 1Participant @brianrice1]

Posted by Neil Wyatt on 30/01/2015 18:07:32:

"The user of the modern lathe, however simple or unpretentious it may be, should always remember that the capabilities of his machine, for accurate work, are far greater than that of the best lathes used by the craftsmen of a few decades ago."

Edgar T. Westbury, 1936

So true and just look at some of the great items they made with them

[mechman48Participant @mechman48]

If you can get the lathe tailstock set & machine two collars on a test piece to within a 'thou', or less, over 6 inches I would call that 'accurate'.

George

[John Lucas 5Participant @johnlucas5]

Thanks for the responses.

I recently completed restoration of an Atlas 6" (my first and only lathe) and was wondering what was available should I want to upgrade to something heavier duty.

So, I''m assuming anything modern for hobby use is better than my abilities. Perhaps something one up from the cheapest imports.

As for the Atlas I can machine a 10" rod in the 3 jaw chuck and be about 2-3 thou out at the end. With tailstock about 1 thou. Think I will save my money and be content for light work.

[Neil LickfoldParticipant @neillickfold44316]

When looking at a lathe, the most important part is how round does it turn a piece. From what I have seen of these Taig lathes and others, is they can not turn to better than about 0.01mm of being concentric and round. The headstock bearings are just not up to it. On the taper bush like a myford, they can be made to be better than 0.002mm. To get that accuracy with precision taper or angular contact bearings, you are paying nearly the price of the machine, just for the bearings.

Neil

[Neil WyattModerator @neilwyatt]

Neil,

I really can't accept that claim. I have a good quality DTI that reads to 0.0005" which is 0.01mm within a spit.

If I turn two concentric diameters on my mini lathe to a good finish, I am unable to see any deflection whatsoever in the DTI on the undisturbed surfaces.

Comparing the very small taig/peatol to a Myford which weighs about fifteen times as much is hardly fair.

I am sure that most 'hobby' lathes are well capable of meeting 0.01mm roundness (which is Schelsinger's limit for finish turning lathes).

Neil

[Tony Pratt 1Participant @tonypratt1]

Posted by Neil Wyatt on 31/01/2015 09:26:30:

Neil,

I really can't accept that claim. I have a good quality DTI that reads to 0.0005" which is 0.01mm within a spit.

If I turn two concentric diameters on my mini lathe to a good finish, I am unable to see any deflection whatsoever in the DTI on the undisturbed surfaces.

Comparing the very small taig/peatol to a Myford which weighs about fifteen times as much is hardly fair.

I am sure that most 'hobby' lathes are well capable of meeting 0.01mm roundness (which is Schelsinger's limit for finish turning lathes).

Neil

I may be oversimplifying the mechanics of what you are doing, but you have substituted the cutting tool with a DTI and then applied it to the previously turned work.

In this scenario I would expect the DTI to show no deflection at all.

Tony

[Russell EberhardtParticipant @russelleberhardt48058]

It's very difficult to accurately measure roundness. A rough and ready check could be perhaps done with a v-block and a dial gauge.

It's equally difficult to see how a correctly set up hobby lathe could produce anything as far out of round as Neil suggests. Most small hobby lathes use angular contact ball bearings or, for the bigger sizes, tapered roller bearings. Those bearings are made to a much higher accuracy than that.

Russell.

[martin perman 1Participant @martinperman1]

Whilst a lathe should be supplied with a level of accuracy at purchase its not beyond the bounds of your average ME to tinker with his machine to improve it, since I've been on here posting many of you have successfully and displayed the fact, checking ovality of a piece is difficult enough but not impossible but it shouldnt be done with an inaccurate piece of equipment like a three jaw chuck, better its tested in a collet chuck etc.

As the quote ETW said our machines, and I include my Clarke CL500m, are a lot better made in design and accuracy and the ultimate accuracy should be in the hands of the man operating it, how many of us push the capabilities of our machines, for whatever reason, and then complain about the fact that the finish, taper and size is bad when we dont take notice of the forces we have just created have exceeded the design parameters of the machine.

Martin P

[Neil WyattModerator @neilwyatt]

Well i did do a test on my test bar (turned on my lathe) recently, using a micrometer reading to 0.0001" and all the readings were within 0.0002" of each other, I assumed the remaining error was my technique with the mike, not out of roundness.

Neil

P.S. don't mention 'lobate'…

[Anonymous]

Posted by Neil Wyatt on 31/01/2015 14:29:47:

P.S. don't mention 'lobate'…

How about debris aprons? Proof that our esteemed editor is really from Mars?

Andrew

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Neil Wyatt on 31/01/2015 14:29:47:

P.S. don't mention 'lobate'…

.

Wouldn't dream of it … but you already did.

A more appropriate test method might be this:

• Turn your test-piece between centres in the lathe.
• …. [use a dead tailstock centre not a revolving one]
• Lock the spindle, and remove the driver
• The piece will now be between dead centres.
• Mount your DTI in the toolpost
• Check for roundness by rotating the test-piece.

MichaelG.

[Michael GilliganParticipant @michaelgilligan61133]

I know it's way beyond the scope of the original question, but this is interesting.

Note: It's only a sample.

MichaelG.

Edited By Michael Gilligan on 31/01/2015 23:44:12

[nigel jones 5Participant @nigeljones5]

apparently you can lead a horse to water but you cant turn cheese into gold…think about it!

[Mark CParticipant @markc]

Michael,

Unless you true up a soft dead center (in which case you defeat the object of the test) how does that help?

Mark

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Mark C on 01/02/2015 01:51:22:

Michael,

Unless you true up a soft dead center (in which case you defeat the object of the test) how does that help?

Mark

.

Mark,

If you mean the soft dead centre in the headstock spindle … then yes [as a matter of course] it would have been trued in-stu before the work started.

Thinks … Now you have me worried: It may be impossible to avoid the periodic errors of the spindle appearing in the trued-up centre. … Perhaps if that preliminary trueing were done with a hand-held graver, the machine errors would be smoothed-out.

Any thoughts ?

MichaelG.

.

Footnote for anyone new to this discussion: The logic of my suggestion [which Mark is correct to dispute] is that 'between live and dead centres' carries spindle errors into the workpiece. … but 'between dead centres' is the best we can do to maintain concentricity … I hoped that spindle errors could be revealed by turning 'live' and testing 'dead'.

Edited By Michael Gilligan on 01/02/2015 08:42:57

[Clive HartlandParticipant @clivehartland94829]

As long as the work you produce on the Lathe is OK why worry about the accuracy? If it is clattering and banging then worry, but maintenance is always on going so you work and become aware of slack movement so fix it.

As to measuring the detailed accuracy of the lathe then accept what it does and do not worry about things. As long as it turns parallel in length I will accept it. Make to fit is fine as far as I am concerned unless its replacement parts then work to tolerances. Otherwise leave well alone!

Clive

[Michael GilliganParticipant @michaelgilligan61133]

Posted by Clive Hartland on 01/02/2015 08:33:05:

As to measuring the detailed accuracy of the lathe then accept what it does and do not worry about things. As long as it turns parallel in length I will accept it. Make to fit is fine as far as I am concerned …

.

Very fair comment, Clive

For me, it's not really a matter of 'worry' … just fascination.

For concentricity and roundness, I am aware of nothing that can improve upon working between dead centres [viz: the watchmaker's "turns" used with the bow, or the bodger's pole lathe]; but this arrangement is inconvenient, so we tend to rely upon the accuracy of the machine. … I find it interesting to know how far the things can be trusted.

MichaelG.

[blowlampParticipant @blowlamp]

Posted by Michael Gilligan on 01/02/2015 09:03:13:

Posted by Clive Hartland on 01/02/2015 08:33:05:

As to measuring the detailed accuracy of the lathe then accept what it does and do not worry about things. As long as it turns parallel in length I will accept it. Make to fit is fine as far as I am concerned …

.

Very fair comment, Clive

For me, it's not really a matter of 'worry' … just fascination.

For concentricity and roundness, I am aware of nothing that can improve upon working between dead centres [viz: the watchmaker's "turns" used with the bow, or the bodger's pole lathe]; but this arrangement is inconvenient, so we tend to rely upon the accuracy of the machine. … I find it interesting to know how far the things can be trusted.

MichaelG.

Even that isn't a guaranteed way of ensuring roundness.

Myford (Nottingham) used to use a process to raise three small areas on some of their (Super7 only, I think) spindle centres to ensure they were perfectly seated whilst being finish ground.

Martin.

[Michael GilliganParticipant @michaelgilligan61133]

Posted by blowlamp on 01/02/2015 12:15:46:

Even that isn't a guaranteed way of ensuring roundness.

Myford (Nottingham) used to use a process to raise three small areas on some of their (Super7 only, I think) spindle centres to ensure they were perfectly seated whilst being finish ground.

Martin.

.

Very interesting, Martin

It reminds me of a bearing arrangement that was used in some Clocks, and on some 'Unipivot' pick-up arms.

… A cluster of three [static] balls; with a conical point, bearing upon them.

Obviously, the ultimate accuracy of turning between dead centres will be prescribed by the quality of the fit between the male and female centres. … Myford's arrangement sounds optimum.

Thanks for the info.

MichaelG.

[Author]

Posts