Hi all
I am a first-time poster here, though have previously trawled the forums for answers to questions.
Firstly, a bit of background. I am an archaeologist who does a bit of surveying and would therefore classify as one of those dangerous creatures who has a little bit of knowledge. Go easy on me!
I am having a pretty basic problem: I keep getting an unacceptable margin of error on my x,y,z co-ords when setting up and orientating to reference. I have tried most, if not all options and tried to eradicate variables, but am still stuck. I know you guys probably canÛªt provide THE solution, but it would be great if you could point me down some avenues that I should explore.
The setup I am using is a Nikon NPR-362 linked to a CAD program (BricsCAD) through an interface program called TheoLT. The total station was calibrated last week by a very reputable technician. I am setting up over and shooting to State Permanent Survey marks that were verified last year.
The first setup I did was shooting from one SPM to another, over a distance of about 130m. The prism (a Nikon prism with 0mm constant) was set up (accurately!) on a tripod/tribrach. I consistently got a plan (x,y) error of 0.050-60mm and 0.020-0.30mm on the z. Thinking it might be a shonky tribrach, I reshot the SPM with somebody holding a pole/prism (CST mini-prism, 0mm constant). Same result. I repeated the arrangements throughout the morning, to no better result.
Later, I set up over two completely different SPMs, which were about 80m apart. Same arrangement with prism on tribrach. The error I got was lower (0.030-35mm on plan and 0.015mm on z), but was still present.
Things I have checked are:
1/ Prism constant is set to 0mm in the machine, to match both the Nikon and CST prisms. As I got increasingly annoyed and stroppy throughout the course of the day I changed the constant in the machine to -30mm and reversed the prisms (their reverse side is -30mm). The result was worse
2/ I checked and re-checked centreing. No problems there
3/ I checked that TheoLT, the interface program, was not altering the results somehow. It had a prism constant option, but this was set to 0mm. There is no other way the data could be modified
4/ I tried to orient the machine without TheoLT (just using the NikonÛªs inbuilt interface). The error WAS better, but still 0.030mm on the plan and about 0.020mm on z. This was between the first set of SPMs, where the plan error had been 0.050-0.60mm.
Help!
I have checked your forums extensively and have a couple of things I will check:
1/ Check the prisms along a baseline to work out Prism Constant (seems to be the most recommended thing to do)
2/ Check the temperature and pressure settings. Would incorrect temp and pressure have introduced such unsatisfactory errors? The day was probably low 20 degrees Celsius, though there was a noticeable heat-shimmer shooting between the two SPMs which were about 130m apart.
3/ Do I need to check the Instrument Constant if the machine was only serviced a week ago?
Am I missing something obvious? Do you have any other suggestions to help me narrow down where the error might be coming from?
Anyway, thanks for taking the time to read this. Any further guidance you can provide would be much appreciated!
weakwilled
" I consistently got a plan (x,y) error of 0.050-60mm and 0.020-0.30mm on the z."
Please clarify. Are these all metric? .30mm or 30mm? What's the ".050"? Inches?
weakwilled15, post: 347380, member: 10822 wrote: Hi all
I am a first-time poster here, though have previously trawled the forums for answers to questions.
Firstly, a bit of background. I am an archaeologist who does a bit of surveying and would therefore classify as one of those dangerous creatures who has a little bit of knowledge. Go easy on me!I am having a pretty basic problem: I keep getting an unacceptable margin of error on my x,y,z co-ords when setting up and orientating to reference. I have tried most, if not all options and tried to eradicate variables, but am still stuck. I know you guys probably canÛªt provide THE solution, but it would be great if you could point me down some avenues that I should explore.
The setup I am using is a Nikon NPR-362 linked to a CAD program (BricsCAD) through an interface program called TheoLT. The total station was calibrated last week by a very reputable technician. I am setting up over and shooting to State Permanent Survey marks that were verified last year.The first setup I did was shooting from one SPM to another, over a distance of about 130m. The prism (a Nikon prism with 0mm constant) was set up (accurately!) on a tripod/tribrach. I consistently got a plan (x,y) error of 0.050-60mm and 0.020-0.30mm on the z. Thinking it might be a shonky tribrach, I reshot the SPM with somebody holding a pole/prism (CST mini-prism, 0mm constant). Same result. I repeated the arrangements throughout the morning, to no better result.
Later, I set up over two completely different SPMs, which were about 80m apart. Same arrangement with prism on tribrach. The error I got was lower (0.030-35mm on plan and 0.015mm on z), but was still present.Things I have checked are:
1/ Prism constant is set to 0mm in the machine, to match both the Nikon and CST prisms. As I got increasingly annoyed and stroppy throughout the course of the day I changed the constant in the machine to -30mm and reversed the prisms (their reverse side is -30mm). The result was worse
2/ I checked and re-checked centreing. No problems there
3/ I checked that TheoLT, the interface program, was not altering the results somehow. It had a prism constant option, but this was set to 0mm. There is no other way the data could be modified
4/ I tried to orient the machine without TheoLT (just using the NikonÛªs inbuilt interface). The error WAS better, but still 0.030mm on the plan and about 0.020mm on z. This was between the first set of SPMs, where the plan error had been 0.050-0.60mm.
Help!I have checked your forums extensively and have a couple of things I will check:
1/ Check the prisms along a baseline to work out Prism Constant (seems to be the most recommended thing to do)
2/ Check the temperature and pressure settings. Would incorrect temp and pressure have introduced such unsatisfactory errors? The day was probably low 20 degrees Celsius, though there was a noticeable heat-shimmer shooting between the two SPMs which were about 130m apart.
3/ Do I need to check the Instrument Constant if the machine was only serviced a week ago?Am I missing something obvious? Do you have any other suggestions to help me narrow down where the error might be coming from?
Anyway, thanks for taking the time to read this. Any further guidance you can provide would be much appreciated!
weakwilled
You may be comparing a ground measurement to a state grid measurement which are usually different.
Sorry RFC, stupid mistake. It is all metric and I should not have put the mm in there. 0.030 is 30mm. 0.050 is 50mm. 0.020-0.30 should read 0.20-0.030 (20-30mm). 0.050-60mm should read 0.050-0.060. Archaeologists and measurements huh!
Agreed
Nikon instruments take "30" as the reflector constant, not -30.
There are many settings on the TS and Software to account for.
You may be in prism mode, target mode or prismless mode.
Then there is the offset in the TS and Software to the focal point of your prism, target or reflective surface.
Then your control monument locations can be published in Grid or Ground.
If you are running with elevations, there is the HI of the TS and the HI of the BS and FS that at random have different values.
Now that was the easy part......;-)
I think you were on the right track with prism constants. You may have the instrument applying a constant, and your data collector (the "TheoLT") applying another. 30mm is the typical non-zero prism constant, and that's the error you are seeing without your dc.
Get your self a 25' (7.6m?) tape measure and a flat piece of sidewalk. Extend the tape along the sidewalk. Set your instrument up over some convenient mark at one end of the tape and measure to another at the other end with the data collector. Measure the distance with you instrument and compare it to the taped distance. Is it right? If not by 0.030m , reverse the prism in its holder (without changing the instrument setting) and try again. If that doesn't fix, check the instrument for prism offset settings. Once you have the instrument measuring right without the dc you can add that to the mix.
Pressure settings would have to pretty wildly off to cause the magnitude of error that you are describing . But it may be possible. 1000mb would be a safe standard setting near sea level. 900mb at the top of a mountain, perhaps.
Grid factor errors would not be that great if you were set to use an appropriate zone. Are you in Australia with your dc set to Oregon State Plane, North Zone or some equally remote zone? that could foul you up. But the instrument without the dc attached would not likely be applying grid factors. So you could have more than one thing going on here. Prism offset on the gun, bad zone in the dc.
Anyway, there are some ideas. Good luck. Let us know how it turns out.
You should set reasonable temperature and pressure (absolute, not weather report pressure) but failing to update during the day won't make the kind of errors you mention.
The temperature sensitivity is on the gross order of 1 mm per kilometer per degree C. So a 10 degree change during the day on 130 meters might change the reading by a millimeter or so. Updating pressure during the day usually isn't critical unless you are going for super precision over large distances.
I have to agree with Norman, prism offset: probably a setting in the instrument. The only thing I would add is to do the settings and checks somewhere you can sit down next to the gun with the instructions in front of you.
A long sturdy table works great. Then you can sit down and go through it all. You can measure directly between the instrument and prism. Work through the temp/pressure input and all the prism offsets and figure out what's wrong and right.
Set up the inst, and another tripod, and tribrach, and prism, 10 Feet apart.
Shoot the dist, manually, and then shoot it with your software.
Then, measure it with the box tape.
Quick and dirty way to check prism offset, and TS offsets.
Nate
I don't understand, what does "0.050-60mm and 0.020-0.30mm" mean?
Grid coordinates will not inverse to ground distance, you need to account for the scale factor.
More information like where you are and the coordinates on your control plus state plane zone and your raw data would be helpful to have.
Sometimes the software also does it's thing, to compensate. And, ditto on the SPC thing. It may need scaled, to fit SPC.
N
Dave Karoly, post: 347412, member: 94 wrote: I don't understand, what does "0.050-60mm and 0.020-0.30mm" mean?
I think he means "within a range of 0.050 meters to 0.060 meters (50 mm to 60 mm) etc. The mixed units are quite confusing but nothing else makes any sense at all. He tried to explain above, but still blundered the decimal.
There are some good suggestions above, but there is more to the problem than a prism offset or a grid scale factor if repeated measurements between the same points give numbers over a 10 mm range. Either his centering is not repeatable enough or there is an instrument problem.
I agree, whatever distance it is spitting out should be consistent.
Where are you located? Maybe one of us can swing by? I have always wanted to help do some surveying on an archaeological site.
Wow! What a run of replies. What a bunch of champs. You've all given me much to think on, but I will try to answer what I can now.
Brad: I am based in Tasmania, Australia...so you're very welcome to come and help...but it might be expensive. Tasmania is the nicer part of Australia, as the animals aren't so angry. No jokes about the Tassie devil either please...
Ekillo and Adam: Grid vs ground measurements eh? These are all ground points, as far as I am aware. To better explain (apparently I have not done so very well so far) -
I am working on a site which has a series of permanent survey marks already established. These have their co-ords provided and were last checked by the (proper!) surveyors responsible last year. I am setting up over one SPM and sighting to another. In the program I use to orient the machine (TheoLT) I input the instrument station's co-ords, plus the instrument height and then the co-ords of the backsight and its height. I then take the backsight, with the program providing me a report of displacement (expressed as a 3D misclosure). Normally - and I have done this hundreds of times on other archaeological sites and with other machines (Leica's generally) - the three misclosure readings for Ds, dD and dH are below 0.020. In this instance Ds and dD were between the range of 0.050-0.060 and dH 0.020-0.030. This was no matter what prism system (fixed or pole) that I used. I also got this displacement when I oriented the machine without TheoLT. Even though the displacement was less on the second set of SPMs I tried, I was still getting an error. In my first post I referred to Ds, dD and dH as x,y,z - which probably confused the issue. Sorry.
Errr, have I explained that better? Or worse?
Dave Lindell: Are you saying I should put in "30" as the prism constant in the machine, instead of "0' which it currently is - even though the prisms have "0" written on them as the constant?
A Harris: Thanks. Am in "prism" mode (there is only "Prism" and "N-Prism" [reflectorless] on the Nikon). The Prism offset is one of the variables I am trying to eradicate. I made sure the interface software has no offset. Control points are all ground. All heights of machine, prism and pole (where used) where triple-checked
Norman Oklahoma: Thanks for the options. Checking the baseline is something I will try first thing Monday if I can't work out what is going on. This machine has always lived at this site and therefore the area coding and pressure settings SHOULD be fine. I will double-check.
Bill93: Thanks as well. I didn't think temp or pressure could cause THAT much of an error! And yes, I did mix up my measurement units like a doofus.
MightyMoe: Nice dog! Yup, Monday will be fun, frustrating day of sitting with the machine, drinking lots of coffee and alternating between fits of madness and melancholia
NateTheSurveyor: Thanks for the instructions re. the baseline. I saw you had posted this elsewhere, so, sorry you had to repeat it!
Dave Karoly: Does the info I provided above make a bit more sense? Basically the program I use spits out a displacement report, which, when it is wrong, makes me sad and then a bit angry when I can't work out what is causing it. I ditched the program to see what happened when I oriented 'manually' (using the TS inbuilt program). It was still orienting with the unacceptable displacement (as in, providing a different x,y,z measurement for the backsight than what it should be - by the error margin outlined above). I can provide data if it is helpful (machine is locked in office over weekend)
Phew, that was an essay. Thanks to all. You've been a great help so far. I am still frustrated (and definitely an amateur), but have a bit more hope now I can sort it.
Cheers
Weakwilled
weakwilled15, post: 347439, member: 10822 wrote: Grid vs ground measurements eh? These are all ground points, as far as I am aware.
I don't know what system is used there, but in the US we often use "State Plane Coordinate" systems where the easting and northing values are on a flat projection of the earth and therefore the map grid distances are distorted with respect to on-the-ground distances. Information is available to tell you what the local scale factor is, of the form 0.999xxx or 1.000xxx so that you can convert between the mapping grid coordinates and ground distances.
So if your State Permanent Survey marks are given coordinates in a similar flat-map projection system, you need to learn how to convert them to ground distances.
Are you able to photograph and post up your prism setup for us? It may help more than words ever could.
One showing the whole setup (not too close) and another or two showing the markings.
Also, you mentioned that these are previously established points you are working with. You should find out how were they established. It's worth considering that you could be chasing someone elses error problem. Also like others stated, could be a scale factor thing going on.
Hello Bill93 and kjpls. Your suggestions have raised the possibility of local scale error. I am in UTM (s) 55 and, according to a websitey-thing for Geoscience Australia I found, I should have a local scale of 0.99986851 for the point I am shooting from. The point I am shooting to has one of 0.99965882. I know the machine is set to a scale of 1. Is that difference really going to account for the displacement error?
Kjpls, I can photo the setup if it will help...
