Starrett 199 calibration

[BVH]

I set out to play with my Ebay Starrett 199. Ultimately, I wanted to level my workbench.

1. Based on what I've read here and on the label on the Starrett box, I got out my almost 15" square piece of 3/4 polished absolute granite and set it on the garage floor until I found a place where it did not rock. Using a 2' carpenter level and an almost new 8" aluminum level, I verified that the granite was close to level, but not completely.

2. I set the 199 on the granite, marked its' position and noted the exact position of the bubble. I'll call this measurement direction "A" - the Starrett label facing me. As viewed from the top, the left edge of the bubble was two divisions and a hair to the left of the large right-hand division line. I reversed the level into the exact same position and noted a significant difference. I'll call this measurement direction "B" - the label upside down.

3. In position "B", I adjusted the screw to start to bring the bubble closer to the prior position. I repeated comparing the bubble in positions "A" and "B" - only adjusting in position "B", until the bubble was as close as possible to the same relative position in both directions after letting it sit for a minute each direction. All in all, I think the screw ended up within 1/4 turn of where it was before I made any adjustments. The screw had a calibration sticker over it when I received it, no date.

4. I then began to adjust my bench in the X and Y directions. I made slow changes, taking lots of measurements with the level. Each time I remeasured with the level, I had the label facing me. I finally got the bench very close to level in both axis.

5. For fun, I made sure the bench was level, left to right using the level in position "A". I then turned the level 180 degrees to position "B" and put it in the same spot. To my surprise, the bubble was almost bottomed out to the left. I turned it back to position "A", and the bench showed level.

6. I then went back to my granite and took measurements both directions. In position "A", the left edge of the bubble was 2 divisions beyond the right-hand large division line. Pretty much where it was before. I turned it to position "B" and the bubble completely shifted to the left, almost mirroring the position.

I'm sure I'm doing something fundamentally wrong but I have not been able to figure out what it is. I think the laws of physics govern this calibration procedure so why isn't it working for me?

 

[gadget_lover]

It sounds like you did it right.




Daniel

 

[BVH]

I don't know how accurate the small, cross vial is. If I place the level measuring my bench left to right and it's perfectly level and then turn the level 90 degrees in the same relative position, the cross vial does not show exactly level. The bubble is shifted about 1/16" or about 20% out of center. Should it not agree with the main vial? This has got to be telling me I'm not calibrated?

 

[precisionworks]

For starters, calibrating a precision level is done with a surface plate Laboratory Grade AA or better. An 18"x18" AA plate will have flatness lapped to +/- .000025" (25 millionths). My guess is that your granite flat is nowhere close to that.

Examine all contact areas of the base & stone out any burrs.

Clamp a parallel to one edge of the plate (it should be as long as the level). Clamp a stop bar 90° to the parallel so the level repeats to the same position each time.

Using your level, bring the surface plate to dead level.

Place level on plate against both the parallel & the end stop & wait 15 seconds for the bubble to settle. Note reading, turn level 180° and repeat.

 

[BVH]

Barry, I realize that I won't be completely precise without the plate but I thought I would be somewhat close and have repeatability which I did not. I think I may have found my problem. When I began calibration, the bubble was far over to the left - well beyond 3 divisions, probably 10, where the left end of the bubble was not visible. Although I did not detect that the bubble had reached its' end of travel in the vial, (I did check for additional leftward movement, there was) I believe it was to the side far enough that there was some amount of compression/distortion of the bubble taking place from being close to end of travel.

I redid the calibration procedure, this time ensuring that when I started, the bubble was only 2 divisions out of center. After 5 or 6 "spins" of the level and some adjusting, I had the bubble ending at the same spot in the vial in both directions, repeated 3 times. I then leveled the bench. It now shows level with the level both ways. When I turn the level 90 degrees in the same relative spot, the cross-vial is perfectly centered.

EDIT: Full ISO calibration was quoted at $176.00, Non-ISO with data is $128 and Non ISO, no data is $102. Is anything less than ISO worth having done?

 

[precisionworks]

Is anything less than ISO worth having done?

Yes if you work in an ISO certified factory or shop where everything has to have a cert. Level is not nearly as important as repeatability since you don't care whether the lathe bed is level or not. For lathe use you just need to be certain that both ends of the bed are tilted exactly the same amount and in the same direction so there is no twist from end to end.

What we do when adjusting a lathe bed might better be termed coplanar axial alignment.

 

[BVH]

....Level is not nearly as important as repeatability since you don't care whether the lathe bed is level or not. For lathe use you just need to be certain that both ends of the bed are tilted exactly the same amount and in the same direction so there is no twist from end to end.

What we do when adjusting a lathe bed might better be termed coplanar axial alignment.

That answers a nagging question - "what does being level in and of itself have to do with bed twisting?" and finally turns on the proverbial "light bulb". The key here is "coplanar axial alignment". Thank you very much for clearing this "level mystery" up!