I am lead to believe Topcon GPS users may only calibrate for elevation on three, or is it four, control points to establish a localised level datum and that the result is an inclined plane of best fit between the control points. I am also lead to believe the option to maintain a "constant adjustment" is not available as it is with Trimble stuff. What happens if one or two of the points is in error vertically or if these "one or two" points were not officially issued as "control"? Answer..."it" hits the fan.
Surely Topcon provides the option to create an horizontal datum plane by adopting only one point for level and other points would be used as checks only. I sure hope it does if I ever have to use Topcon.
Can anyone enlighten me while I do a bit of a clean up!
Replies
As a basic rule of plane geometry - you should always hit at least 4 points when calibrating. Otherwise the routine being used will or could show a perfect fit of the three points to the vertical plane. By using four or more points you should get a residual error report for each point being a couple hundreds off the "best fit" vertical plane.
Having not used Topcon GPS ( I've only used Trimble or Ashtech/Sokkia gear) I can only speculate as to the actual process. But, I would always hit 4 or more points depending on the size and shape of the area I was working in.
Hope this helps
Some interesting replies, thank you all. I've got the basic rules down pat, after 50 odd years it's just difficult to come to grips with the notion of level datum plane not being horizontal over a small area i.,e., the RL of the datum varies depending on where you are. In the past I have established a site level based on a particular bench mark and checked it against a number of other marks, deciding whether I'm good to go after that. Kevin D. Allen in your Trimble equipment do you calibrate using a "constant adjustment" or an "inclined plane"?
I always used "constant" - I could never get over the idea of an "inclined" plane being level
I can't comment about a "constant adjustment". It sounds cool but might require a bit of hocus pocus.
Any localization (gpe calibration) goes this way.
1 control point allows one to translate the project
2 control points allows one to translate, orient and scale
3 control points does #2 and allows a tilt trend and supplies redundant horizontal data
4 control points does #3 and supplies redundant data all around
more control points add to redundance and eases finding poorly fitting control.
JAC