Setting up the lathe accurately

[VicParticipant @vic]

Does this help?

http://igor.chudov.com/manuals/Rollies-Dads-Method-of-Lathe-Alignment.pdf

[Anonymous]

As far as I'm aware the M300 does not have the facility to adjust the headstock. I was advised to never even think about mucking about with the bolts that held the lathe to the cabinet, as these were shimmed at the factory so that the lathe was as good as it could be.

When I say clean the gap piece I mean the following. First wipe over with a cloth. Then a wipe over with a cloth soaked in solvent, then dried. Every bolt hole and especially the taper pin hole are cleaned through. Then wiped over with a cloth and solvent again. The lathe bed has the same treatment. The taper pin hole is cleaned with a cloth pulled through, and each tapped hole has a tap run through it to clear any dirt. It's a bit of a palaver but you need very little to push the saddle out of line. If the bed is worn evenly by a thou or two it doesn't make much difference to the turned diameter, but if the gap piece is tilted by a thou or two on one side then it will make a big difference.

Regards,

Andrew

[alan frostParticipant @alanfrost17805]

Dave,if you're going to go for it (nothing wrong with that) but its time consuming and may be diappointing then after paying due care to the gap piece and after "levelling" the lathe (usually a tricky and time consuming job with an engineer's level). I usually have a smoke while the bubble settles. First you should run the lathe for the recommended hour to remove oil film clearances in bearings. You don't need a dial gauge with graduations finer than 0.01 mm "finer graduations lead to misapprehension regarding the accuracy of the gauge which is at best 0.0025mm " .

Then you are aiming for (lathes up to 8 inch centre height) 0,005mm true running at the spindle end and 0.015 mm at a distance from the chuck of 300mm). This is almost certainly the best your lathe was aimed to meet when new. Of course it could have been inside this of possibly outside as Harrison would have been aware that no factory employing skilled turners was likely to need this accuracy or check it..

For what its worth I know a skilled tool maker (retired) with two largish old Colchesters which like most peoples lathes are just plonked down (I suspect amateurs like me worry far more about levelling than professionals). He produces work of immaculate quality and fit and tells me, and I believe him, that he can work to a "tenth" provided the lathes are "loose" i.e,. have some backlash " can't work to that on a tight lathe ,Alan"

Anyway best of luck. Alan

The

[alan frostParticipant @alanfrost17805]

PS. Of course should you achieve this it will only apply probably for the same temperature and almost certainly won't apply the moment you take even a light cut. You'll also need the right preload on the bearings .God knows what this is for a twenty year old lathe (younger than most of mine).

Edited By alan frost on 16/11/2014 13:43:53

[alan frostParticipant @alanfrost17805]

Just in case you're tempted and taking note of Andrews post above ,first para, and to quote Mr. Schlesinger "Dismantling is always detrimental to the machine". Schlesinger may not have been God but he knew enough to have his standards adopted by practically all machine tool manufacturers in every country. He was of course also aware that certain manufacturers could make by taking great care machines very well inside his standards in which case the manufacturer usually proudly and rightly specified this.

My own "best" lathe had a run out of 40 millionths when new but I certainly won't be measuring what it is now or trying to achieve it. Regards.

[Dave CParticipant @davec87625]

Thanks Alan

I took note of Andrews post and once again removed the gap bed and thoroughly cleaned and replaced it and all of the bolts and bolt holes etc. I then re sharpened a tool and ran some further test cuts taking light cuts.

I now get repeatable results of .4 to .5 of a thou run out over the six in length.

I am more than happy with that now thankfully as I didn't really want to start messing with the headstock or tailstock bolts for fear of making things worse, Also I had been advised to not mess with the assembly unless absolutely necessary.

As you say I may get different results on a heavier cut due to preloading etc.

Results may change over time but I am now happy that I am starting out with good results and feel all the better for doing the tests. The saddle only runs onto the gap bed by about an inch during the test cuts but this must have been enough to throw things out a little. Anyway it seems sorted now.

Many thanks to everybody for the help and emailed advice.

[Phil WhitleyParticipant @philwhitley94135]

RESULT!!

Phil.

[alan frostParticipant @alanfrost17805]

Well done.Its very easy to become obsessed with unneeded accuracy and then to become disappointed in one's lathe. After all you probably have a quite fine lathe (they were certainly well respected when manufactured ) but you would n't necessarily feel that if becoming too finicky with accuracy. My guess would be that it is more important to get a good fit on parts which probably depends more on other factors than fine tuning of the lathe. Even the judges at Model exhibitions can't tell if your parts are within small fractions of a mm in absolute size but they can evaluate fit.

I have a Mikrokronator (or whats its face) that has divisions of I think from memory 5 millionths (and no backlash unlike a micrometer) , a very high quality toolroom lathe, and a balance that weighs to a millionth of a gram (I think -have n't had it long) but only because I am interested in metrology. I do most of my work on an Emco V10P which was probably manufactured to Schlesinger limits ( heaven alone knows what it is now ). I'm not a model builder altho I admire them , and a lot of the time,in fact most of the time, anything within 10 thou. will do ,probably a great deal less.

Regards Alan

 

 

Edited By alan frost on 16/11/2014 17:46:05

[alan frostParticipant @alanfrost17805]

Apologies to anyone interested in metrology who looked up mikrokronator. It has of course nothing to do with time unless travelling near the speed of light and the "whats its face" is Microkator,but the divisions are ).000005 inches. One can extrapolate of course (there's no Vernier ).. I used to work in an industry where even this would have been far too sloppy and eventually very short wavelengths of "light" had to be used.

I think I remember (talking of metrology) the tube ads. by Shackles and B——–ks that showed a piece of extremely fine tubing by their closest rivals with a piece of their finest tubing threaded down the bore. I would like to think it was a sloppy fit.

A similar story which I think is true concerns the attempts by the `Japanese to break into the lucrative American veneer knife business. The Japanese were repeatedly told the tolerance was zero but persisted and came up with a knife that was extremely fine but still did n't cut it. It would be nice to say that the Americans returned the Japanese knife with the edge split exactly down the middle but thats how untrue legends start.

In those days American and British know how was the tops.

 

 

Edited By alan frost on 16/11/2014 19:57:12

[Phil WhitleyParticipant @philwhitley94135]

Come on Alan, you are teasing us now, what industry did you work in?

Phil

[alan frostParticipant @alanfrost17805]

Phil, not at all,I worked in semiconductors, integrated circuits which at the time I retired was just going sub-micron. They are well beyond that now (X ray lithography and probably something better than that now ) and so far manage to keep Moore's law running true. I did say I worked in it first as a deign engineer and then in sales . I never worked in the factory and altho no doubt we had guys in the factory who had to consider this accuracy to keep control of the process , the main task of ensuring this accuracy was down to the companies manufacturing the production equipment, not us as a semiconductor company. As you probably know the whole very large and extremely heavy factory for processes this accurate had to be mounted on shock absorbers (traffic vibrations within a good distance threw things out) and the air had to be filtered almost to molecular level (to exaggerate a bit).One of the worst things about the industry was that a very very expensive factory only had a few years of life before the geometry moved down another order and a better one had to be built. The more chips you got to a wafer the cheaper you could sell 'em.

It would be interesting to know where they are at now as I retired from the industry 25 years ago. Incidentally the products we made to a precision exceeding any jewellery including watches by many orders often sold for a few pence (to industry, more to the man in the street ).

We did make pick and place machines for surface mount that could under ideal conditions place a million components an hour very very precisely. Heaven knows what they can run at now and to what precision.

I would think that there must be a few industries where they work to this precision (optical instruments comes to mind ) but usually ,and certainly in the case of semiconductor manufacture it bears little relationship to measuring turned diameters or mechanical accuracy. I'm happy on the days I get something to half a thou.

Regards Alan

PS added. If you want a good ,very large sound proof workshop Hyundai built a state of the art factory in Dunfermline which due to a market turn down never got used. Don't know if its still there though

 

 

 

Edited By alan frost on 17/11/2014 00:55:10

[Author]

Posts