Starrett 199 precision level concave like crazy - what's going on?

I recently purchased the level from a local carpenter. From words of that guy, he bought the level brand new and used only a couple of times. Visual appearance confirms that - the thing looks like it's new, not a single scratch. However, when I checked it against a surface plate, the blue was only on very edges of the level. I measured the low spot with a dial indicator and it shows 0.001" hole in the middle, which seems like a lot to me. The surface plate was made and calibrated a few days ago, so I don't doubt it's accuracy. I also checked the level multiple times in different locations on the surface plate and results were identical.

The surface of the bottom of the level has very nice looking, consistent scraping marks. I believe they are original from the factory. I don't see any signs of excessive wear there, at least visually.

Is it supposed to be like that?

Thanks!

They are intentionally scraped hollow in the middle so that you will always get good bearing on the ends even on surfaces that may be a bit convex or otherwise questionable. I

nygavin said:

They are intentionally scraped hollow in the middle so that you will always get good bearing on the ends even on surfaces that may be a bit convex or otherwise questionable. I

Click to expand...

I've read the same, but if that print is good I'd consider that going too far. I'd want the outer 1/4 or so of each side to have a clean, uniform bearing, not like what we're seeing.

Well, it could be worse- which would be bluing in the center and high on both ends.

You do want contact to be greater on the ends than in the middle, if there is ANY contact in the middle.
(Many precision levels are made with no base section in the middle).
The area of contact should be even across the width of the level, not greatest in the center as it appears to be at both ends.

If this was my level, I would rescrape it.

Having said that, I assume you did not take the level out of a hot car ( or other warm area) and set it on a cool granite plate, causing it to warp concave. If it did not sit for several hours out of its box at the same room temp as the surface plate I would recheck it when it has had plenty of time for the temp to even out.

Don't mess with it. It is exactly the way it should be.

steve-l said:

Don't mess with it. It is exactly the way it should be.

Click to expand...

I don't agree. Read this thread, many either want the level fully flat, or the outer 30% flat and the center relieved by ~13u. This level (again, if the print is good) is too far out and irregular in its contact.

http://www.practicalmachinist.com/v...ction/scraping-precision-master-level-342871/

Thanks everyone for the replies,

I blued the level many times with identical result - it should not be a flaw in blueing technic. I scraped a few other parts using the same process and results were very consistent.

The main concern about concave level is: imagine I'm removing twist on a bed of my mini-lathe. It's 2 times narrower, than the level and the level sits perpendicular to the ways. I'm moving the level from one end of the bed to another. Since, I can't put the level in the exact location every time, contact points will shift back and forth along the concave part of the level and I will be getting erroneous readings. Also, I'd like to sometimes use the level as a small handy straight edge (is it a good idea?).

The Force is not with me yet. Once it comes - I will re-scrape the level. What bothers me more - why did it end up like this? Could cast iron bend/twist just because of sitting on a shelf for a few years?

J_R_Thiele said:

Having said that, I assume you did not take the level out of a hot car ( or other warm area) and set it on a cool granite plate, causing it to warp concave. If it did not sit for several hours out of its box at the same room temp as the surface plate I would recheck it when it has had plenty of time for the temp to even out.

Click to expand...

The level was outside of the box a few hours before blueing and it was room temperature. I only handled it by the plastic top to not warm it up. Just in case, will re-check this tomorrow.

Milland said:

I don't agree. Read this thread

Click to expand...

Thanks for the link!

Frigzy said:

Thanks everyone for the replies,

I blued the level many times with identical result - it should not be a flaw in blueing technic. I scraped a few other parts using the same process and results were very consistent.

The main concern about concave level is: imagine I'm removing twist on a bed of my mini-lathe. It's 2 times narrower, than the level and the level sits perpendicular to the ways. I'm moving the level from one end of the bed to another. Since, I can't put the level in the exact location every time, contact points will shift back and forth along the concave part of the level and I will be getting erroneous readings. Also, I'd like to sometimes use the level as a small handy straight edge (is it a good idea?).

The Force is not with me yet. Once it comes - I will re-scrape the level. What bothers me more - why did it end up like this? Could cast iron bend/twist just because of sitting on a shelf for a few years?


The level was outside of the box a few hours before blueing and it was room temperature. I only handled it by the plastic top to not warm it up. Just in case, will re-check this tomorrow.


Thanks for the link!

Click to expand...

I do not agree with scraping these levels. I did not agree with the thread in November when I read it then. The level works perfectly for its intended purpose. You will never, in your life, use this level enough to wear it out. Further, any measurements you make on your mini lathe bed will not be meaningfully and significantly different either way, so why chance ruining a perfectly serviceable tool?

J_R_Thiele said:

If this was my level, I would rescrape it.

Click to expand...

agreed. I can'tt quote the OEM specs chapter and verse but after decades of working with these things and listening to others remarks on them my strong sense is the the centre relief is not supposed to be anywhere near as much as .001" and its supposed to bear on the two sides properly up to the centre relieved section, not just on the extremes. A level that far out is rather useless imo as in practice its not always used on the extremes of the surface; you need to be able to rely on it if its bearing on anything but the relieved section.

No idea how it got that way. My first reaction is that scraping didn't look very Starrett like, but I've got several of these and just looked at two: they are all different so that might not mean much. Not coming up with anything other than its been rescraped or it went through Starrett's QC on a really bad day. I would post it in the scraping forum, for anyone scraping this is a basic tool so there are lots guys experienced with them, you might get some other ideas there

edit. on a bit of searching and from accounts by Strarrett employees reported here, the relief is suppose to be .0002" and comes from a Federal spec

Frigzy

You can use the level as is to remove the twist from the lathe.
The dip in the center is .001, the level is 15 in long, so the change of .001 occurs over 7.5 inches. If the change is linear (unlikely) that is .00013 per inch. If you are off by a tenth of an inch in placing the level at the same contact point you have an error of .000013- which you will never notice.

One approach is to put marks on the sides of the level, and use those when placing the level.

Another option is to use rubber bands or masking tape to hold a piece of metal with parallel ends to the bottom of the level. Put dowel pins in contact with the parallel ends. To remove twist, the actual reading of the level does not matter- it just needs to be the same. If you want to know the actual measurement, you need to measure the distance between the dowel pins. The Starrett level graduations are .0005 in 12 inches. If you have the level supported at the very ends, one level division indicates a height difference of .000625 (15/12 x .0005) If you are using the level on a 30 inch straightedge supported at the ends each division equals an actual height difference of .00125 (30/12 x .0005). On the mini-lathe, if the dowel pins are 4 inches apart, the actual height difference is .00017 ( 4/12 x .0005- .0001666). If it is leveled with a 20 pitch bolt, each revolution of the nut is .050 inches. Turning the nut by only 1 degree moves it .00014. You may want to start with a less sensitive level. As you are only trying to remove twist, shimming the level to get it in a readable range would save a lot of time.

Frigzy said:

The Force is not with me yet. Once it comes - I will re-scrape the level. What bothers me more - why did it end up like this? Could cast iron bend/twist just because of sitting on a shelf for a few years?

Click to expand...

I scraped a smaller less precise starrett machinist's level. It was twisted and rather unflat in no real identifiable way but didn't look worn or mishandled. I figured it was just the iron relaxing over a period of years