Thread: On the LEVEL.

Hi Michael,
The Starrett 98 series levels are only graduated .005" per foot, about .4mm per meter. I have a second hand 98-6 and although it is good for "rough" leveling, the CTC 8" makes it look pretty ordinary at .02mm per meter. I have played with both and if i level something as good as i can with the starrett, i put the CTC one on and its clear it is not REALLY level.
The CTC failed a spin test but it only took 4? rounds of scrapping to get it flat. I have never actually checked the Starrett.

Im no authority but : If you browse the UK ebay site you will see many older levels eg Moore & Wright and Rabone Most of the sellers will post to Australia if you ask them . The smaller 6" levels are Ok to post, but the larger ones are too heavy .Ihope that helps the prices are good and cheap if you have the time to browse

I got my 98-8 from this forum.

The CTC site is very interesting, even for a wood turner

Originally Posted by Ueee

The CTC failed a spin test but it only took 4? rounds of scrapping to get it flat. I have never actually checked the Starrett.

In real terms how bad was the bottom of the level? I suppose as with all things that are supposed to be accurate we all want them perfect and I will just have to buy one and see how it performs.
It won't add a lot to the freight bill.

I took one pass out of the middle of the level before the first blue and it didn't spin after that.....so it was pretty darn close.....maybe a Ray or 2 out
I think with a decent builders level to "rough" with and you can go straight to the CTC.

It makes me think.....all the furniture, kitchens etc that i have installed "level" were not really level.

Originally Posted by localele

Alright guys which is the nicest level to have stashed in a drawer for posterity? Is the Starret 98-8 a better buy than a CTC 8 inch level? While the Starret is almost twice the price I am more interested in the USE of these levels around the workshop. Will they both need to be checked and tuned up as far as their bases go or will they be useable out of the box? I don't really consider $125 a lot for an accurate tool which should last for generations.

It depends what you want it for.

Most machinist levels (like the Starrett 98) are in the ballpark accuracy of around 0.3mm/m. That makes them quite useful for many tasks, including the levelling of bench lathe beds.

Precision levels are about 10 times more accurate, good for levelling long lathe beds and the like. But the high precision also makes them far too sensitive to be useful for most everyday tasks.

Think of it like the difference between a touch dial indicator of 0.01mm resolution and of 0.002mm resolution. In practice, the latter is useless for most everyday jobs like aligning a vise, because it only has a travel of 0.2mm and because its indicator fluctuates around surface imperfections and flex of the setup. So it only gets used maybe once every two years, and spends the rest of its life in the box. It is exactly the same with a precision level. It resolves wear and tiny dents in the ways of a lathe bed. And if used elsewhere the bubble flips always to one or the other side. Only buy a high resolution DTI or a precision level if you really know that you need it. Else it may just be a dead investment that gathers dust (unless of course, if one was to collect tools for reasons other than to use them, but then one would probably not look at a Chinese level ). Chris

Edit: the quality of the vial in the level is very important, especially if it is a sensitive precision level. You get waht you pay for. Think of it again of the analogy with a DTI: a good and not too old DTI has a smooth and crisp needle movement. The needle of a cheap DTI does not react as crisp and accurately, same as the needle of a good but old DTI with gunked up oil in the gears.

It is an interesting conundrum. My use is just for setting up machines to a better tolerance rather than using it to level cupboards.Having a height adjustable workbench I do use a level to set it up when it is adjusted.I checked it and the 1200mm Stabila level states an accuracy of .5mm/m.While this is close to the stated accuracy of the Starrett it only has two lines engraved on the bubble.A bit long for machine setups as well.
I would have imagined a machinist precision level to be quite a bit more accurate than a builders level even though they perform a similar duty.

The Conundrum:
The CTC is 21 times more accurate/sensitive but for 1/2 the price.
The Starrett looks nice and has the heft of a piece of cast iron (never a bad thing).
Either one may require a tune up to be really effective.(is this a sign of the times).

Getting a bit off my own topic here but did machinists from 50 years ago expect to have to tune up a precision instrument they had just bought?

Originally Posted by localele

It is an interesting conundrum. My use is just for setting up machines to a better tolerance rather than using it to level cupboards.Having a height adjustable workbench I do use a level to set it up when it is adjusted.I checked it and the 1200mm Stabila level states an accuracy of .5mm/m.While this is close to the stated accuracy of the Starrett it only has two lines engraved on the bubble.A bit long for machine setups as well.
I would have imagined a machinist precision level to be quite a bit more accurate than a builders level even though they perform a similar duty.

The Conundrum:
The CTC is 21 times more accurate/sensitive but for 1/2 the price.
The Starrett looks nice and has the heft of a piece of cast iron (never a bad thing).
Either one may require a tune up to be really effective.(is this a sign of the times).

Getting a bit off my own topic here but did machinists from 50 years ago expect to have to tune up a precision instrument they had just bought?

- The Stabila builder levels have no adjustment for the vial. They are guaranteed for life to be "better than 0.5mm/m accurate". What they mean is that when you turn the level over side by side, the bubble will be centered and show no more variation than if you tilted it by 0.5mm/m.

- A typical machinist level of 0.005"/ft (0.4mm/m) accuracy is much more sensitive, about 8 to 10 times more sensitive then a Stabila builder level.

An experiment: put your 1200mm long Stabila on a surface and level it as good as you can, maybe putting shims at both ends 1m apart. Now put a typical machinist level of about 0.4mm/m accuracy on top of it. Measure how many mm you have to lift one end of this assembly, to cause each bubble to move sideways by 2mm. I just did this experiment (Stabila model 87 1200mm long vs. a little 4" Rabone), and it is a real eye opener the Rabone is 8x more sensitive .

I am no expert but I believe this is so, because the glass vial inside of a machinist level is precision ground and polished to a very fine surface finish, and it is filled with Ether or similar substance with very low viscosity and surface tension. In contrast, the Stabila builder level uses a low cost block vial made from plastic, and if you look at it close-up, you can very clearly see some concentric machining marks left on the inside surface. Mind you, I am not saying Stabila's are bad, I have several and am very happy with them - for building tasks. What I am saying is, that the definition for accuracy is obviously different between a carpenter level and a machinist level.

Some manufacturer links with good information:

Spirit Level Sensitivity | Spririt Level Accuracy | Degrees Minutes Seconds Converter

WYLER AG, WYLER AG is the leading manufacturer of precision spirit levels, Cross spirit levels, Circular spirit levels, clinometer, inclinometer, inclination measuring instruments, tilt sensor (download the pdf catalogue "spirit levels" from this website)



You are after a level to setup machines. But do you really need a high precision level accurate enough to show you the seasonal movements of your concrete garage floor (which moves with the humidity of the soil underneath)? The seasonal movements with temperature and humidity of a wooden bench can be observed with a normal 0.5mm/m machinist level.

Let's assume you need a level to level the bed of your lathe to make sure it is not twisted. A 0.5mm/m machinist level will let you do that, as the bubble can be read down to an accuracy of 0.1mm/m. The final adjustment is not done with a level. It is done by a test cut, you adjust the lathe levelling screws or shims such as to obtain a taper free cut. Regardless of what a super precision level may tell you. After all, the end result you are after is a taper free cut, not a perfectly accurate level reading. Chris

If you can get hold of "Model Engineer's Workshop" issue 77, October 2001: On page 38 is an article "On the level". Among other information, with instructions on how to lap your very own glass vial. Very informative, even if you do not want to make your own machinist level. Chris

0.02mm/m accuracy is way overkill... In fact I doubt those vials are ground to be that accurate anyway....

0.05mm/m (10 arc seconds) is quite good enough

A starrett 199 is 0.04mm/m plus has 1/8" graduations, which actually makes it more sensitive.

Originally Posted by cba_melbourne

The final adjustment is not done with a level. It is done by a test cut, you adjust the lathe levelling screws or shims such as to obtain a taper free cut. Regardless of what a super precision level may tell you. After all, the end result you are after is a taper free cut, not a perfectly accurate level reading.

Hi Chris,
Not so sure about that.
Wouldn't it be fair to say if the bed is level* yet the lathe is cutting a taper, the problem isn't the lathe bed?....Now you may choose to fix the problem by twisting the bed, but that's not the real problem.

Now if everything else is correct and the operator knows what he is doing**, taking a test cut should level* the bed


HI Rishard,

Originally Posted by .RC.

0.02mm/m accuracy is way overkill... In fact I doubt those vials are ground to be that accurate anyway....

Accurate in the absolute sense? I doubt it(though I checked mine and it seems pretty good). But do you need it in a level? Arent you after sensitivity and repeatability?

Originally Posted by .RC.

A starrett 199 is 0.04mm/m plus has 1/8" graduations, which actually makes it more sensitive.

??

Stuart

*untwisted on the same plane if you like.

**If you're talking about chasing the last half a 1/10 thats fine but I always picture someone with a bad tool turning 20mm bar 8" out of the chuck then twisting the bed to get it to cut straight.

Sorry for the mixed units, trying to keep both camps happy

Originally Posted by .RC.

A starrett 199 is 0.04mm/m plus has 1/8" graduations, which actually makes it more sensitive.

That takes us into a new league of level with a Starrett 199Z coming out of the catalogue at $814.

Does anyone have an idea of the system that a manufacturer like Hercus would have used to set their lathes up in the factory when new?

Originally Posted by Stustoys

Accurate in the absolute sense? I doubt it(though I checked mine and it seems pretty good). But do you need it in a level? Arent you after sensitivity and repeatability?

??

Well probably should have used sensitivity..... I once used a borrowed polish made 0.02mm/m graduation level..... It was an absolute nightmare to use... Even looking at it moved the bubble.. Having owned one of the generic precision levels even though it was supposed to be the same sensitivity, it reacted differently to the polish Kimex one...

Accuracy would be how good the vial has been ground... I am having trouble wondering how a $50 chinese one can compare to a $1000 Wyler one..

Of course using to level you are using it as a comparator comparing one section to another and getting them both the same....

Originally Posted by .RC.

Accuracy would be how good the vial has been ground... I am having trouble wondering how a $50 chinese one can compare to a $1000 Wyler one..

I have enough trouble trying to equate a $50 with a $125 level.Maybe it will end up being both to try and answer it for myself.
Therein lies the problem of what level of accuracy to aim for. Given that it is to compare the sections of lathe bed and if the divisions were too coarse there is always the option of a bit of masking tape on the vial.Not meaning to demean the debate but I looked back through some photos from one of the scraping classes to check on how to do a lathe bed and found that as a method used. I actually think I would start with getting the lathe bed level at the headstock end and trying to replicate that at the tailstock before finishing with a test cut or five. Yes I do know you can have a lathe on a boat but if I sit a beer on my lathe bed I like it to sit up straight.