Accuracy of a 12" Rabone level?

I have a Rabone 12" Engineers level, which I believe was carried over to Rabone and Chesterman era.

There is no identifying number on it, apart from the Makers name, it is utterly without any information. Usually such a well made level would have its accuracy marked somewhere on it, but not with this one!

So what would its accuracy be? I have used it to check and adjust for zero twist in a Myford lathe bed. I am now wondering if it is good enough for this purpose? Or should I purchase a clinometer for such jobs?

Andrew.

Hi, I was having problems with bed twist last week and used a similar or the same Rabone Chesterman level, the ones with the V on the bottom. It has improved the turning remarkably . A simple matter of slackening off one of the tailstock end foot mounting bolts. I think as long as the bubble is in the same place in 2-3 places  on the bed ,its accurate enough for most .

Mine doesnt have any accuracy markings either.

mines identical to this:

Edited By mark smith 20 on 10/09/2017 14:39:32

Edited By mark smith 20 on 10/09/2017 14:40:19

Andrew, The spirit bubble level irrespective of the bubbles 'Speed' is that it is, 'End for End' accurate. There is no point using it if when you change end for end that it runs off. I would suspect that you have checked it end for end and allowed it to settle for a few minutes and that at no time during your testing that you breathed on the bubble nor touched it. This because the bubble and mountings are temperature sensitive.

The sensitivity (Speed) is maybe rated at 10" or perhaps 20", that is immaterial as your testing will be of making sure the bubble is central reading when settled.

My one adage is, never trust a bubble, it is prone to outside effects!

Clive

Hello Clive,

Thanks for your input. I have indeed checked it end for end and it is accurate in that respect. I was feeling quite good about it until your last sentence "Never trust a bubble"!

So is it accurate enough for the job in hand? Or should I get a clinometer? Interesting that you mention "Speed" I think I understand what you mean by that. However I have seen some Engineers levels that do give an accuracy of so many thou per unit length, or does my memory fail me yet again?!

Mark,

Thanks for your encouraging post. After I had levelled the bed and setup the tailstock, accuracy improved a goodly amount!

Andrew.

The precision levels give some figures

such as

**LINK**

or

**LINK**

Thanks Mark,

The first link gives what looks like a slightly modernised version of my Rabone, at 3.5 thou per foot, I think that will do me.

Thanks,

Andrew.

Thanks Mark,

The first link gives what looks like a slightly modernised version of my Rabone, at 3.5 thou per foot, I think that will do me.

Thanks,

Andrew.

If the precision (not accuracy) of a level isn't marked, you have no idea what it might be from comparison with similar, but maybe not identical, ones.

Set it up so it's about level, then raise one end with a feeler gauge and note the movement of the bubble to work out what its precision is.

You could always try putting a small shim under one end and see how much the bubble moves?

Good idea John, now why didn't I think of that!

Andrew.

Generally speaking; I think it is safer to treat Sprit Levels [in common with Dial Test Indicators] as comparators, rather than measuring devices. ... i.e. assess the difference between two "measurements" by seeing how much shim is needed to return the bubble to to reference position.

MichaelG.

Ideally, when raising one end with feelers or shims, try to do so over a known span, (in your case, I would suggest end to end)so that when the bubble has moved by one divison, you will then know the sensitivity in thous per foot for one division.

If you want to be really finicky, start with both ends resting on equal diameter rollers, shim until bubble shows Zero, and then raise one end by placing shims or feelers between the roller and the level at one end. Ensure that the ends are on a diameter of the rollers.

Howard

You have all seen these yellow bubbles in builders levels well they are made by bending the tube, but precision levels have a bubble that is ground so that the inside is made like a barrel. The radius is about 20 meters for a 10" bubble and the job was usually carried out by one man. They retire and can find nobody able to carry on the job. I would think nowadays that the grinding would be carried out by a NC machine so precise glass tubes can be made.

Andrew dont waste you money on a clinometer, it will give nothing that the level will offer and the other thing it will have a short base compared to a level which has a longer base. as I described and also mentioned by Michael that you sue it as a comparison device and it only tells you that something is not level or not the same in a length.

Trust your level but do not handle it near the glass tube or near it and then errors will not creep in. When I said dont trust a bubble you do not know what has happened to it since you last used it.

Clive

Years ago I had the privilege of meeting a German trained instrument maker (surveying instruments). He had designed and partially built a machine for lapping the barrel shape into ID of level vials. If memory serves he used a metal rod that was deflected to produce the correct radius and lapped the glass against the rod. The machine lapped several vials at once, each on its own rod. I believe he was intending to make 10" vials.

'

That is 10 second vials. 10" of arc per 2mm travel of the bubble.