Master precision Machinist level

[verbotenwhisky]

The story of my life

At the risk of reiterating a previous point, You would do well to have a decent Machinist level, I have the one you asked about, it is a good tool; but it will have you chasing rabbits, its too sensitive, if I had it to do over I would have bought a .001 in 10 or 12 inches, not the .0002 in 10". All you really need to do is level the ways to match on either end, if it isn't level with the world your okay.

 

[RaisedByWolves]

Im very new to machine work
I often ask myself
$130 for a new precision level .0002” accuracy? Can I find a better Used unit for the same costs?

Much of it comes down to cost/need/want and budget

I have very little budget for new tools or equipment so I make do.

I have a 1990s smithy 1220 and its not capable of anything great in my hands. I called upon a friends dad, retired machinist who worked at the long closed ship yard. I knew he had levels and set up large lathes.
When he came to inspect my little lathe he laughed at me me about “leveling” my lathe.
He says its in a shed on floating on a crushed stone foundation ,
Along with the way rails being so short. Your going to fight deflection in this small lathe more than any leveling will help. He installed his test rod set up some indicators and said as it sits it will deliver the accuracy the machine claims IF you can.

I did pick up one of those small staretts for short money. I have not used it yet….

I had to go back to this as two things stand out.

1: The guy you had over is smart enough to understand that houses "Float" on the surface of the earth.

2: Deflection. Your work holding (Collet vs 3 jaw chuck) and stickout will be the greatest factors you are likely to run into and not understand.

It takes a huge amount of force to make a cutter cut aluminum or steel. Just try to hold a tool bit in your hand and make a cut.

Second think, don't do that, there will be blood.

But you get the picture, it takes tons of force to make metal cut metal and every action has an equal and opposite reaction.

Wonky chuck jaws will induce taper in your part, this is why we have collets, for times when you need more accuracy.

And stickout is teh debbil.

 

[Dabbler]

At the risk of reiterating a previous point, You would do well to have a decent Machinist level, I have the one you asked about, it is a good tool; but it will have you chasing rabbits, its too sensitive, if I had it to do over I would have bought a .001 in 10 or 12 inches, not the .0002 in 10". All you really need to do is level the ways to match on either end, if it isn't level with the world your okay.

I do own a .0002 per foot level, but never use it. I own 2 .003 per foot machinists levels I use all the time.

 

[Nesse1]

I own 2 .003 per foot machinists levels I use all the time.

Any specific brand/model recommendations?

Most of the levels I see are either the .0002"/10" type, or more like .005"/10". I'd like to find a quality tool, not crazy expensive, that falls in the middle.

 

[RJSakowski]

BTW, who is Rollies dad?

I don't know if this is in jest but for the benefit for newcomers to the forum who may not have been introduced to the Rollie's Dad's Method, here is the full scoop.

Rollie's Dad was a machinists who came up with a brilliant alternate for checking taper on a a lathe.

The method consists of chucking a round bar in the lathe. The requirements for the bar are that it be round, i.e., uniform in diameter as you rotate the bar, and that it be uniform in diameter across its length (although there is a workaround if the second condition isn't met). The bar doesn't have to be straight nor does any runout in the mounting matter.. The bar is mounted in the lathe and a dial indicator is mounted on the cross slide so it contacts the bar at a point horizontal to the spindle axis.

The principle of operation is this. When the chuck and bar are rotated, the dial indicator reading will change depending upon the amount of runout present. If you are making readings close to the chuck and adjust the cross slide so the maximum and minimum reading are equally plus and minus, a zero reading is located exactly half the diameter of the bar from the true spindle axis.

Now if you move the carriage towards the tailstock, without changing the cross slide, and again determine the maximum and minimum readings, the average of those readings will be the distance the dial indicator has moved relative to the spindle axis. Since the dial indicator itself hasn't changed, that distance is he amount the cross slide has moved relative to the spindle. Divide that distanc3e by the carriage traverse and you have the taper..

There are several important advantages of using Rollie's Dad's Method (aka RDM) rather than cutting a test bar.

1. You don't need to do any machining so there is no need to make a new skin cut after each adjustment.
2. There is no need for an expensive test bar
3. There is no need to adjust for zero runout
4. The method can easily be adapted to set up the tailstock
5. It can be used to set the tailstock for intentionally cutting a taper
6. It can also be used to measure alignment in the vertical plane
7. An adaptation of the RDM can be used to vertical travel on a mill or to align the axis of an RT when used in the vertical position.

 

[verbotenwhisky]

I do own a .0002 per foot level, but never use it. I own 2 .003 per foot machinists levels I use all the time.

When I bought mine I had ordered the lathe and it is my first lathe, I have owned several mills, and I kept reading articles which discussed accuracy and bed induced tapers so I thought the more accurate the better.... Now I know, I have chased the rabbit.... I looked at a used Starrett the was .002 in 12", in hindsight, that is the one I should have purchased. My bed had a twist in it when I received it, It took a few months, 150 pounds of lead and a great deal of patients to get it straight so I was glad I had a machinist level.

 

[RJSakowski]

Any specific brand/model recommendations?

Most of the levels I see are either the .0002"/10" type, or more like .005"/10". I'd like to find a quality tool, not crazy expensive, that falls in the middle.

Levels are relatively simple tools. The primary defining characteristic is their sensitivity. A level is self proving, meaning that it is possible to calibrate a level without any external device. If you place a level on a reasonably level surface so that the reading is in range when thew level is in one orientation and when it is rotated 180º horizontally, the average of the readings will tell you a.) how far out of level the surface is and b.) fow far out of calibration the level is.

For example, if the level is reading 2 divisions to the right in one orientation and remains 2 divisions to the right when the level is rotated 180º, then the level in calibrated correctly and the surface is out of level by 2 divisions OTOH, if the level reading moves 2 divions to the left when the level is rotated, the surface is level but the level is out of calibration by 2 divisions.

The are two subtle additional characteristics. Precision levels generally have a less sensitive cross vial. The reason is that there is a slight interaction caused by a non level condition in the dirction at right angles to that being measured. Ideally, the surface will be level to the degree of sensitivity of the cross vial. The second concern is the bottom surface of the level. Yhis surface should ideally be perfectly flat. I f there are high or low spots on the surface and the level isn't returned to the same contact points with subsequent measurements, errors can be introduced. This is a concern for lower cost levels. It isn't a deal breaker in my book as this can be corrected or worked around.

There are differing opinions as to what sensitivity one should get in a level. My preference is to go with a more sensitive level. The reason being that you need you see movement in order to detect a deviation. An analogy would be you can't determine a time trial to a hundredth of a second if you are using a wristwatch that only has a second hand. However, the more sensitive the level is the longer it takes to settle and the more likely it will be affected by external disturbances, passing vehicles, thermals, etc. To that end, the sensitivity of the Starrett 199 series levels at .0005"/ft. is about right for me. My much lower cost level has a sensitivity of .0005"/10". The level linked in the original post has a higher sensitivity at .0002"/10" and that may be too finicky for some users. The Starrett 98 series levels with their sensitivity of .005"/ft might suffice if one is careful to note the 10x smaller movement of the bubble.

 

[Dabbler]

Any specific brand/model recommendations?

I was glad I had a machinist level.

I own Moore and Wright machinist levels. They are the ELS (6 inch) and ELM (12 inc) sizes, with a resolution of .0035 inches per foot, you can read to half that with practice.

When they come up here, they go for a reasonable price. The 6" cost me 60$ and the 12" cost me 40$.

I use them for quick set ups (but an digital angle box is faster), for setting up lathes, and generally level the beds on all my machine tools.

 

[Nesse1]

I've found a VIS 8", which is .0005/10", and an EDA 10" at .001/10" or .0005/10" for about the same price. The VIS looks like a higher grade unit. It's a little more worn looking than the EDA, which looks like new. I'm quite certain that the .001" unit is more than accurate enough for my needs, but...

 

[Dabbler]

I've found a VIS 8", which is .0005/10", and an EDA 10" at .001/10" or .0005/10" for about the same price.

The one thou will be noticeably easier to use. Heavier weight might be given to condition.