Hi guys
I traversed some stations today around the perimeter of a construction site. I took great care, did 3 rounds and got a 1 in 228,000 accuracy. Great!
Thing is when I loaded it into the survey processing software (LSS) it threw up some excessive difference errors. Please see excerpt from the processing report below.?ÿ There were 7 stations and the total traverse was 371m.
I don't get how I can have such excessive differences between expected and observed angles on loading the survey and then get such good traverse accuracy.
Anyone know what might be going on?
?ÿ
?ÿ
Can you re-observe just that setup??ÿ Might be quicker than sifting through the data.?ÿ I mean, you know you need to re-observe it anyway.?ÿ 0.022m = about 0.066ft, which is too sloppy for your perimeter control between two stations.?ÿ Did you trust a rookie chainman to set up your last prism?
Is your range pole,and total station calibrated??ÿ
Total station was calibrated recently.
I set up all the prisms. They were perfectly centred and leveled. I also checked them again before taking them down. All was good.
Dunno what is going on.
How come the traverse closes so well and has such good accuracy?
The traverse has good precision. The accuracy is terrible...
When was the last time you adjusted your prism tribrachs?
I have seen this happen the prism you are using is not the same as the one in the data collector, some companies use a prism for backsites and a different prism on the range pole, but in the data collector they have it listed as the same prism, or sometimes they have a -30mm offsite when it should be a 0 offset on the prism. Also how are your tripod legs, do they stay tight, do you check the prism level before you tear down the backsite?
I use Trimble gear with Access. When setting over every new station during the traverse after backsighting the previous set on station the deltas of the co-ordinates given by Access were within 3mm (1/8 inch). Surely that means all prisms and station setups were well centred.
My prism carriers are the rotating Leica ones. Once I've centred over a nail I rotate the prism and check level and centering again. I do this every time.
http://www.surveyingequipment.com/product/leica-gzr3-precise-optical-prism-carrier/
Very nice stuff! Yeah, it's hard to say, I'd tend to just re-observe that last setup and see if I can get my problem to disappear.
Is Stn K a sideshot??ÿ It looks like you turned good angles to wherever the prism was at at K.?ÿ It was either not over the point or the tribrach is out of adjustment.
You might also consider using this old trick for blunder detection, shown in Figure 1 at this link and described in the text below the figure.?ÿ Back in school, we were made to prove mathematically why this works, but all I remember is that it often does.?ÿ I've used it for practical purposes only once in nearly 20 years, but man, it was worth it.
https://www.xyht.com/surveying/methods-analyzing-data-part-1/
Running thru the control three times is not the same as turning 3 sets at each setup.
Many things can affect instrument behavior and operator actions.
Implement and add as many cross shots that can be possible.
Any object that blocks any part of the BS or FS can induce error, like having to filter thru any object or the site passing too close to the ground or obstructing objects like fences or buildings or grass and such. Heat waves from weather, exhaust vents and other sources cause difficulty with site and EDM functioning.
A complete analysis of the results will give you more of a conclusion.
?ÿ
Yep. The first time I did this was in school on a hand drawn map.
I got to the bottom of this so thought I'd share what I found out.
The survey processing software I use (when it loads a job file) compares raw angular data per face not the meaned angle between faces. Its comparing an expected angle against observed F1 and then F2 NOT the mean of them which is why it was throwing up the large errors.
On two of the legs I had 30 secs diff between faces but each face only differed from the mean by 1 second. I did 3 rounds of f1 and f2 per set up.
The traverse calculation looks at meaned angles so F1 and F2 differences get cancelled out. This is why I got a traverse accuracy of 1 in 290,000. My misclosure was 1mm of Northing and 1mm Easting.
The only thing that bothers me is the 30 sec diff between faces on two setups. Might do a Hz angle calibration and an autolock check.
This exercise clearly demonstrates the benefit of doing multiple round in both faces.
Anyway I'm happy and will sleep better tonight!
?ÿ
?ÿ
?ÿ
?ÿ
Try using a parasol.
Try using a parasol.
Say what?