Precision Level or Precision Frame Level

[This posting has been removed]

Lathes don't have to be 'level but they do need to be supported in such a way as to not impose a twist to the bed.

In order to do this each of the supporting points need to sit on a plane just as the bed did when the ways were ground. The plane need not be horizontally orientated but if it is then a level can be used on the ways in order to check this. In order to eliminate twist using a level the supporting plane only needs to be 'horizontal' across the ways, the longitudinal or spindle axis doesn't really matter.

If I could comment on the subject of bolting down or not as I see it and using the Myford as a 'test subject' there are a couple of points.

The lathe bed was originally ground flat and true by bolting upside down and grinding the feet level. The bed was then flipped the right way up and bolted down to the bed of the grinding machine and the ways brought true, by definition in relation to the feet.

When the rest of the lathe is assembled which includes having a large motorising 'lump' bolted over the back of the headstock this imposes a twisting force as much of the weight is transferred to the rear foot under the headstock. Correct bolting to a suitable stand counteracts this force and further stiffens the bed casting against cutting forces. It will also add mass to some degree which will help with vibration. This is a good enough reason as any for bolting down in my book. The main thing though is without some positive means of adjusting he support arrangements, in my case levelling jack feet there is no way to actively bring the lathe into 'perfect' alignment.

You would be surprised at how 'flexible' small lathes are. The Super 7 bed is quite a sizeable cast iron lump but just try taking a finishing whilst leaning in the headstock and then take the same cut standing clear and you will see a fraction go missing from the diameter.

I appologise to all who know all this already but I hope this helps those who are still figuring it out.

regards Martin

Edited By Martin Kyte on 01/05/2019 09:19:09

Hi all

After reading many of the contributions to this thread I am thinking of buying a level to use for "leveling" my lathe, and also for setting work up on my mill.

There has been much input on this thread about the sensitivity needed for these tasks and I find myself confused by all the different "Units" quoted by manufacturers and advertisers.

There seem to be units based on angles such as degrees, minutes and seconds of arc, and also degrees and decimal fractions of a degree. Then there are units based on offsets, and here there are many different lengths of baseline and some use Imperial measures and some use Metric units.

So how sensitive does the level need to be? and can the answer be quoted in many units so that levels from different sources can be compared? It is almost as if we could do with a conversion table, anyone know of one?

Thanks to anyone who can clarify this.

Regards

Rod

Rod

The Starrett Master Precision Level sensitivity is quoted as 10 seconds of arc per division. This corresponds to a slope of 0.0005 inches (half a thou) per foot or 0.04 mm (40 microns) per metre.

In my vernacular a total pain in the butt to use as the slightest touch takes the bubble out of view.

My old gunners clinometer, picture about the middle of page 2, is slightly less sensitive at 30 seconds of arc per division, 0.0015 inches (one and a half thou) per foot or 0.120 mm (120 microns) per metre. The divisions are approximately 2 mm apart so its easy to estimate 1/4 of a division. Being so small the clinometer can easily be reversed end for end so the effective sensitivity is doubled. In practice it runs a Starrett Master Precision level so close that any difference is more likely to be operator error than real.

The huge practical advantage of my clinometer that it has a screw adjuster calibrated in minutes of arc over a 5° range. Keeping the bubble in sight for any rational error is easy and tracking changes in level due to adjustments is almost trivial.

By far the easiest way to level a machine. If you can find one grab it and don't loan it out.

There are other styles of sensitive clinometer, most significantly less easy to use for this purpose, but all far less frustrating than a master precision level.

Clive

Edited By Clive Foster on 02/05/2019 21:53:44

.

It may be useful to note that, in this context, 'levelling' is the process of using a level

[ rather like 'milling' is the process of using a mill ]

The term originates with surveying, where a level is used to quantify slopes, not to flatten them.

MichaelG.

And once you have got your lathe perfectly level, you should then do a turning test by taking a light cut over a piece of 1" diameter bar sticking four to six inches out of the chuck, with no centre in place. Any variation in diameter between the two ends of the bar can be adjusted out by shimming or raising either the front or rear "foot" of the bed at the tailstock end. See the Myford users manual for details on this. (Available free online as PDF).

In fact, might as well forget the whole level thing and just do this test. Especially with even a slightly worn lathe that is not going to have a perfectly flat bed surface to rest a level on anyway. 

At the end of the day, all that matters is that the lathe turns parallel, which this test establishes better than any other. In fact, it's the only test that establishes this.

Edited By Hopper on 03/05/2019 07:10:04

Hopper is spot on. Levelling to the ‘n’th degree is clearly time wasted if the lathe then needs ‘unlevelling’ to make it cut a uniform diameter.

Hi, yes the term "level" is a bit misleading in this context, the lathe will turn just the same upside down (just ask them down under) if fact it will be the same in a vertical position. The point being that the bed shears are on the same plain as each other, from the headstock to the tailstock, whether the lathe itself is tilting forwards or backwards or sideways. If you are using a level, all you need to observe is that the bubble reads the same at each end of the lathe bed when placed across the shears and the bubble doesn't even have to be within the central two lines. The turning test will tell you if it is correct.

Regards Nick.

Coplanar ( in the same plane) is probably a better description of how each end should be.

Bloomin' swot

Neil

Hi JB

I’ve just spent the last week sorting out my Harrison T300 lathe...I tried using the Rollie Dads method but found the accumulation of errors (taper adaptor, test bar etc) difficult to resolve, so went for levelling the lathe. Having levelled the lathe, I did a test piece and had an error of 1/1000” over about 6”. Tweaking the relevant lathe foot, now gives me no error over 6”

If you would like a recommendation: the level I used was a Moore And Wright Machinists Spirit Level. 12”/300mm. I found it perfect for the job required. My level was second hand, from the well known auction site. I’ve just checked and there is one currently available (buy it now £79.99)

Hope this helps

Tony

Hi Clive

Many thanks for your prompt and comprehensive reply to my question.

I just have to find one like yours now!

Even if I can't find an adjustable one, I now have an idea of the sensitivity I am looking for.

Thanks again.

Rod

If I didn't already have mine I'd be seriously tempted to make something functionally similar. Start by strapping one of the short, inexpensive, import bar levels to a sine bar style pivoting plate on a flat base. Add a vertical screw to adjust the pivot. For quick and dirty I'd probably have proper full circle pivot pins at each end of the bar. Then drill and tap the one at the moving end for the screw. Put another pivot in the base at the moving end and drill part way through to form a location for the end of the screw. 5° movement should be plenty. Being able to go an extra couple of degrees below horizontal.

Proper calibration like mine is handy but only necessary if you are using it to measure things.

For levelling the screw adjuster is just there to keep the bubble in view whilst you get close. Finish off by switching end for end until the bubble stays in the same, or near enough the same, place. Always remember that the objective is to get the thing bolted down, or standing on its own feet if its a bigger beast, with no extraneous stress on the bed. The level just tells you when making an adjustment or tightening a bolt is straining things.

100 mm long levels seem to be in the £30 range so could be sat on a pivoting bar from 4 to 6 inches long. Close to 1° per turn is 10 tpi for 6", 11.5 tpi for 5" and 14 tpi for 4" between pivot and screw. I'd probably go for 11 tpi on 5.22" between pivots as being very close as one of my lathes has 11 tpi on the screw cutting box. Or if you are metric 2 mm pitch on 115 mm between pivot and screw is very close to 1°.

If you do go that way don't be tempted by the 0.02 mm / m ones. Unless you can stop breathing!

Clive