Selecting an Engineer's level

If you buy a secondhand level, it can be checked for accuracy, and corrected, by taking a reading facing in one direction. When turned through 180 degrees the readings should be the same. If they are not, the level should be adjusted until the readings at 180 degrees are the same.

Without a level, a lathe can be adjusted to eliminate twist by using the method advocated by Myford, and described in Ian Bradlewy's book, "The Amateur's Workshop" on pages 27 and 28.

The Myford 7 Series Manual, again by Ian Bradley, describes the method on pages 41 and 42

In "The Amateur's Lathe" L H Sparey, on pages 35 and 36, describes, (not so clearly, in my view ) how to level a lathe using a D T I.

Howard

A lot of discussion on how to fix it but as yet the amount the lathe is "out" is unknown. I might suggest that if "it isn't cutting true" and can be seen without measuring, then something more fundamental might be askew. "Levelling" or better "Untwisting" is normally something employed to correct tapers measured in thousandths of an inch not something you can see with the naked eye.

John

The aim of the op is to align the spindle, bed and tailstock. Levelling the bed can be part of the process but is just one tool that can be used in the process. We should not keep saying levelling a lathe when we mean aligning the various parts of a lathe to give the required relative movements. Calling the process levelling makes some people think you need a level to do it and that that is all that is required.

Martin C

.

I was trying to keep out of this, Martin ... but I do feel the need to re-state my understanding of the semantics:

Levelling in this context [as in surveying] is the process of using a level to compare heights.

... The ambiguity comes from the assumption that it must mean setting something horizontal.

MichaelG.

Really what we are actually talking about is adjusting the mountings so that there is no twist in the bed, so that the lathe turns parallel.

My ML7 was very sensitive to the torque applied to the holding down fixings.

Its successor, a BL12 - 24, on the same bench, seems to be most affected by the two fasteners, of the six, beneath the chuck.

The Raglans probably overcome the problem by having only one fixing, under the Tailstock, so that it effectively has a three point retention.

Aligning the Tailstock with the Headstock should produce parallel turning between centres.

ASSUMING that the Headstock and Tailstock centres are at the same height!

Howard

I suspect, unless your lathe stand is extremely rigid, shimming the feet may not correct any twist. Many lathe beds are quite likely more rigid than the wooden benches they're sitting on. As well as some of the cheap sheet steel stands available from the Far East.

THIS!

+1.

Even 50 quid for something nice and beautifully made is surely better than paying double for something that looks a bit kack with no brand name or engineering history.

I wish the 70 year old lathe I use had less than 0.025" wear in its ways, but nevertheless, it can still cut parallel, and has never been levelled by me in the last ten years.

I wish the 70 year old lathe I use had less than 0.025" wear in its ways, but nevertheless, it can still cut parallel, and has never been levelled by me in the last ten years.

This assumes that the bed is twisted as a permenant feature. This is not the case. You are not trying to correct a twisted bed but correct an unevenly supported lathe.

The lathe should be 'twist free' when unstressed. It's the uneven mount that introduces the 'twist'. By definition when in the stress free 'untwisted ' situation the mount is imposing no 'torque to the bed but merely evenly supporting the feet. Thus a light wooded bed can correctly support a lathe in it's proper configuration to turn parallel.

This does not however mean that sturdy mounts are not better than flimsy mounts, they are.

regrds Martin

Quite right Michael. Add that to the fact that the most expedient way to to check for twist is the ensure that the ways are actually level both ends. They need not be and it might not even be possible, but why complicate things?

The Colchester alignment check sheet calls it a 'transverse bed level' measurement. Seems like a pretty intuitive description.