@lukenz not bad at all. I also use take a piece of paper and tape it on the floor. Give a plumb bob to someone and i would make a mark. And have them get as steady as possible and let them drop it and mark about 10 to 20 times. So now many little dots while everyone watched. Then go through the math and draw the ellipse showing outliers etc. i am such a hands on and visual learner myself And poor at communicating and things especially to those of higher education. Grammatically i use the visual. Now they can see the precision that is going on. Then make them switch hands there off hand do again. Now we have two ellipse the truth is somewhere in the middle lol. I was taught standard deviations by almost same method except it has a ink mark on bottom. I was slow on education it was always one thing to solve some math problem. I was one of those kids that had to see it in action what and how is it applied. Your way works for most because most get it quicker than i did.?ÿ
So double onscreen RMS values for 95% CI and on tight jobs 3x RMS.
We just have our crews change the precision display to 2-sigma / 95% for typical work:
Default in Access is DRMS, which is better than 1-sigma but not by much.
?ÿ
@mightymoe Need to add a hand-level one to the gear list. Not a bad idea. Small compact light weight fits in your pocket. I will make a spirit level one that sets perpendicular to each other now we have solved the problem. Lol. ?ÿ
Tried that APC experiment today. After fourty minutes of letting the network rtk resolve ambiguities, and the estimated solution dropping from 0.04'(H) to a 0.01'(H), I took the two meter snap lock and turned the cardinal directions, with the compression nut snugged, but not locked.
The results were 0.005'+/- difference between the previous turns.
During those fourty minutes, gaining, or losing, satellites had a greater impact on the position.
?ÿ
For a long time I was skeptical about the value of antenna orientation for RTK because I figured the errors inherent in the method swamped those arising from orientation, but one of the guys on the Javad board kept insisting that it mattered, so I looked at the antcal tables for my setup (Triumph-LS and Triumph-2) and found that as much as 0.05' in the vertical can result from mismatched orientations.?ÿ The possible horizontal error is much smaller, as I recall, but since I often use RTK for vertical I now take pains to orient the antennas to north, even though that sometimes makes for some awkward setups.
@jim-frame so in your situation keeping the antenna pointed north means the javad system is indeed applying the ant calibration as part of the elevation and direction of the satellites. ?ÿI imagine they all should be doing this. But i also imagine that since different antennas and manufacturers have different specs some may never see that much error while others could see more depending on the offset value. This also proves the theory of gps gnss is not 2d it is not separated by hz and vertical. Was your horizontal error or difference about 1.5 x??s less than vertical difference. And how does that compare to the anten x file specs of offset values. ?ÿ I still have to get my hands on one of these Javads. I have by the guidance of another member on here will be camping out as soon as hay season is over and spend a few days working with a guy who sells but is a licensed surveyor and user. I am looking forward to it for sure . When i was at the agency i did some work on the team that helped prepare for icd -211 specs so that was passed down to manufacturers for gps. Lots of agencies involved and academia so we were a very small portion. But i was able to read lots on different manufacturers at the time. Javads system did perform very well. Other manufactures did also. They all seemed to have their way about interpreting the specs and niches . ?ÿSo all had pros n cons on different aspects. I wonder if javads system since its a cube has any effect on near noise. Vs other antennas that are round. A radome effects the signals and changes the apc. And different radomes shape size and materials will have different effects. ?ÿWe are living in a time where we can actually see and measure these small changes and differences. ?ÿIt is amazing. ?ÿ20 years ago the effects of different radomes were studied but not seen as an issue because the error was still in outer band if the noise. Fast forward to 5 years ago we had alarms going off and really testing and studying more intensely. Who woulda thunk. ?ÿLol ?ÿ
Interesting thread.?ÿ I have couple of observations/comments.?ÿ First, I find it interesting that we are talking about this kind of precision but are still using a rod for the setup.?ÿ If precision matters that much then use a tripod with multiple setups at discrete times, even then you are looking at around 1 mm of error budget.?ÿ Rods are great for topo shots and monument ties.?ÿ Rods should be checked regularly and transported in a way which minimizes the chance of them vibrating out of adjustment.?ÿ You can do a quick field sanity check using the tailgate of a truck for that matter.?ÿ Rods do get bent from time to time and I would venture to say that every rod is manufactured slightly out of true.?ÿ Second, while RTK precision is much improved over the years, it is still a magnitude behind static GNSS.?ÿ As an example, Leica's latest flagship GNSS receiver, the GS18, has a listed precision of 8 mm hz/ 15 mm vert + 1ppm for single base RTK vs 3 mm hz/ 5 mm vert +0.5 mm ppm for fast static.?ÿ While you may observe better results from time to time, I would never certify anything less than these.
Shawn great topic. ?ÿWow at the interaction that it has created. Ya most definitely have people thinking. ?ÿWe have 2 things going on here. One the error of a bubble and how best to reduce that. 2nd we have APC offset going on. ?ÿYou have drawn a lot of great thoughts for sure. This is the type of things that make me love being back in surveying. Problem solving different theories and methods to achieve goals.
?ÿ
It seems that the more you learn, then then the more we can see our need to learn.
N
If you use matched, name brand antennae, the combined horizontal offset will be 1.4 mm or less. The value of a (nearly) independent observation far exceeds this trivial error.
This discussion has not clearly distingushed accuracy in finding your location in the world vs accuracy in measuring the bearings and distances within a project. Some techniques improve one at the expense of the other.
@bill93 so true. ?ÿRelative accuracy vs absolute accuracy. I have been playing with network rtk to be on datum around project but the. Use base and rover to have better relative ties on the project. So in this case the network rtk is closer to nad83 per ngs cors. And i can use my base to establish relative control throughout the site and i do move my base around so i get some redundancy from one day to the next. Yesterday i tied to some traverse points and noticed i was systematically off in same direction throughout. Today I found my issue on fixed height tripod. The bubble was bad. Note enough to sweat but enough i reversed the bubble today and re shot every thing as far as control. I will see how it comes out when I perform least squares. Rtk and traverse data. Should work well from field numbers so far. ?ÿGot to get the bubble fixed though. That stuff drives me crazy. Lol.?ÿ
