I have the same problem / concern and would like to know if you got this resolved. And what that resolution might be.
This is the only post that I have ever found that has noticed this situation. I am so tired of reading about US foot, feet and international feet.
years ago our office upgraded to C3D 2010 and found the same problem and resolution you mention (sea level). I refrained from using 2010 and continue to use LDD to import FBK data. Now the office is running 2012 and moving to 2014. The same problem exists. The error gets greater the further away the measurement is from the instrument. I've seen error up to .04 around 600-700 feet from the instrument.
the issue I see is when I mix different types of measurements with different collectors. I start with GPS measurements to establish control points (and to topo with). The GPS control is then utilized with conventional equipment all of which are 5600 robots, using a TSC3, TSC2 and TSCe collectors.
I make field books with Trimble link. I first import the GPS field book which establishes all the control for the 5600 robots. Then I import the conventional equipment data (removing the inputed coordinate values). I import the TSC3, TSC2 and TSCe fbk files. When I compare coordinate within the respective collectors, the TSC3 and TSC2 data is really close. The TSCe data contains the most error.
All the FBK file headers are the same. Prism constants are the same. FBK units are the same.
Remember, I am using 4 collectors on the same job site. If I import each FBK file in an individual drawing, the error from the autocad coordinate relative to the collector coordinate is much much smaller. Darn near perfect. But now I have four autocad files with points in them that need to be combined in to one file for drafting. Then there is the issue with download each collector every couple of days to add more data.
Any input on what you know is greatly appreciated.
Welcome to the C3D Forum.
Need more specifics. Are your data collectors all using the same software and version? Are you using Carlson, Survey Pro, Survey Controller, Access?
I know you're "tired" of reading about the difference between Survey Feet and International Feet, but it's the easiest and first thing to check, in your data collectors as well as your survey database.
Can you provide a table of some points that have differences? Establish a pattern, then you'll find your solution.
From your description, I'm not 100% sure if this is your issue or not. We've noticed that if you re-shoot the GPS control points with the total station that the software will overwrite, or average (depending on settings) the GPS point with the total station point. Our solution has been to import the control from the GPS separately as control points. When the points are defined as control points in the database, subsequent imports won't modify them.
I have not found a solution. I've contacted Trimble and Autodesk and no one seems to have an answer for me. I still process my data in TGO and import linework into C3D. The only work around I found is to export a text file from our TSC2 and I can process the text file in C3D. I prefer TGO processing over C3D because there are more linework code options in TGO.
Can you post one of these suspect FBK files, a csv version of the same points and what the coordinate system projection settings are/should be?
As you said in your original post you are only talking about 'a hundredth to a couple thousandths' (assume thats thousandths of feet?). Is this significant or is it within the stated accuracy of your survey anyway.
As suggested if you use GPS for control, import those as 'control points' so that any total station data won't affect it, or for setup points in a field book ensure they are listed as NEZ at the top of the field book and not NE SS, so that they also come in at the data logger coordinates.
Also as you suggested, the only correction that should resonably be done in Civil3D is the Sea Level correction so that ground distances are reduced to ellipsoid distances. This is one area where slight differences can occur, for example in one test I did I found:
"I did a test to compare the resultant ellipsoid distance of an observation on an NZGD2000 meridonial circuit at elevation 3000m and a measured ground distance of 10000m. My TSC2 gave a distance of 9995.299m, Civil3D 9995.291 (or 0.8ppm), so in essence identical."
If this is the only correction turned on then differences should be negligible.