The Problem With Laser Levels

Have you ever used one? Like every tool, if used properly they can be helpful.

One way of solving the "doubt" about the accuracy is to take a third reading for each set-up, on a point close to the laser level. Any adjustment error on this will be zero because it is so close. (Effectively the whole level run is then a giant three peg test).

Every time the laser is moved on you have a comparison between the old foresight delta (fore minus close-up) and the new backsight delta (old close-up minus old fore). They should be the same except the sign reversal. If they are then all is well; if not then you know what the adjustment error is and can mean the two.

You also have a second check in that the observed difference between alternate points should be the same as the sum of the two sub-sections.

It doesn't need any more walking, just an extra rod.

Whether it's a laser or a total station you still need to remember to keep the sightlines well clear of the ground and not shoot over hot hard surfaces. (Yes, it even gets warm in the UK!)

The ones I've used wouldn't turn on if not properly leveled.

I have used them. I trust them to provide results within specs.

Rotating the instrument by 180 degrees and reshooting the point will increase the accuracy by manually averaging the observations of each face.

Topcon recently came out with an auto leveling robotic robot for the construction industry. Mount on tripod, auto level, place over the mark, auto level. I would like to see this instrument in action, interesting idea.

Like any level, there are two principle sources of error.

First, if it sweeps out a cone instead of a flat plane then balancing sight lengths will be important. A peg test should be run to see if it is good. Refraction and curvature also apply here.

Secondly, if the axis isn't vertical the tilt will throw things off. If you can rotate the device 180 without changing the vertical axis then you can detect the problem, and manual averaging will remove most of that error.

Laser Levels work in generally the same way that Autolevels work, only instead of hanging a prism on wires, it's a motorized laser light. If it's in need of adjustment, the plane of light is tilted, so the error will be different in every direction. That's why the same peg test we use on an Autolevel won't work on a Laser Level. The adjustment on a Laser Level is more like the adjustment for a prism pole--it has to be rotated.

Dave

This is all true and I wouldn't have thought of it.
The divergence of a laser level is quite great as well (if you have ever seen it in the dark). I wouldn't trust a reading to less than 5mm and maybe a lot more over a long distance.

Personally I think they are best used for general construction site levels over distances up to max 100m, not control work.

Robotic total station is a much better way to transfer levels (on your own) if you have one available. C&R compensation as well. As long as both faces are shot the error on a trig levelling loop should be pretty good I would have thought. If you do a loop you'll know anyway.

We called that "double hubbing" when running a line that you were not going to close back.

Read the User's Manual under Calibration

Richard,

Thanks for the link. I think their peg test is not rigorous enough. There should be at least three 90-degree turns instead of just one 180-degree turn.

Dave

> Richard,
>
> Thanks for the link. I think their peg test is not rigorous enough. There should be at least three 90-degree turns instead of just one 180-degree turn.
>
>
>
> Dave

You failed to include calibration of the Y axis...