GNSS Solutions Question

I don't know much, or anything about the PPK type surveys, however, I noticed that the version of GNSS Solutions you mentioned you are using is version 3.10. That version is pretty outdated. The most current version that I am aware of is 3.80.

Might I suggest checking out Mark Silver's GNSS Solutions videos? He is a guru with GNSS Solutions, and has helped me on more than one occasion. He is a heck of a nice guy, very knowledgeable, and has many videos on various things with the program.

Mark Silver - Igage Training Videos

Good luck! I sure hope this helps.

My procedure for your example:
1 Process location of one base
2 Use that base as a control point and process location of second base
3 Set both base as control points and then process the rover locations

0.02

You HAVE to update to later version. Known issues with clocks in the version that you are using. I can't remember the exact details, but I do remember it won't fly anymore.

Be sure to turn off 'Adaptive Static Detection', then try to remake the job.

If bringing in RINEX files, remember to always click the 'Apply to All' button after setting / checking the HI and antenna for all files (do not skip this step.)

🙂 GNSS Solutions lives on in-spite of all efforts to kill it.

M

hi guys,

I have a v3.8 at the office. tried that with the rinex files from a sokkia 2700 and it won't import. tried all versions of rinex from the sokkia gps and it says that there is something wrong with the gps data file.

Had to backtrack back to v3.1 because it accepted the rinex from the sokkia GPS.

I usually use PPK for bathymetry survey and I select all the points with a dz_conf of less than 0.10. Most of points in the bathymetry survey survey returned a Fix solution. I guess being out in the open sea gets more satellites than on the ground.

I am at a loss at the moment. The Help section of GNSS Solutions does not offer enough explanations on the topic.

Thanks guys.

Is GNSS Solutions really slow or my points are just too many?
I am processing roughly 50,000 points of static-kinematic via PPK.
As mentioned earlier 2 base x 2 rovers each recording @1 sec interval.
Well I finally figured my initial problems regarding the FIXED solution vs
large proc_length_conf/DZ_conf values. As long as they are Fixed then
their positions as ok for my need which is for rough topography.
What I did was to process each rover from the 2 bases independently. Then
I converted the points to CAD & compared the 2 sets. The difference was less
than 0.10 hor & 20 ver. These are only for vectors with Fixed solutions.
Another problem cropped up during the independent processing. Some points
would fix from 1 base but not from the other. It was random so no way to select
them manually. So I had to again reprocess all the kinematic points using
the 2 bases at the same time. This way the software selected the best point
depending on which vector had the least error.
So I started at maybe 9 AM this morning and it is now 3 PM and I have not yet
saved Points file. The vector file was saved about 1 hour ago with 100,000+
vectors.
I am using an i5 64 bit 4 gb laptop. I gues the 64 bit has no effect because
GNSS Solutions was installed automatically into the program file (32) directory
so it is not using 64 bit speed in processing.
Anyone else took this long to process their kinematic topography?

Tip - close your survey window during processing. somehow speed was increased when only the workbook window was open.

FrancisH, post: 333081, member: 10211 wrote: I have a v3.8 at the office. tried that with the rinex files from a sokkia 2700 and it won't import. tried all versions of rinex from the sokkia gps and it says that there is something wrong with the gps data file.

What's the actual error message you get? A newer version of some processing software doesn't accept contemporary data, but an older version does. My spidey sense is tingling.

For v3.8 I have been removing my SD card and using a Card Reader to import the files into GNSS.

I've installed v3.10 on one computer and sent all my previous files there. Have not downloaded anything so far, collecting data today so will be doing so soon.

In my opinion, in any kinematic project RTK or PPK you have to have enough check points to be comfortable the results are what you say they are. In other words re-visit at least a few points with at least 1/2 hour or so between measurements to see if you get different results that are close enough to your needs. The driving principle is to look for is different satellite geometry so the greater the time difference between duplicate positions on the same point the better. In my view residual differences are really all that matter and processing error statistics can be misleading and it can also be very difficult to get the software manufacturer to tell you what they mean in mathematical terms. The error statistics you're showing above "Proc_DX_Conf" etc. are something that spin out of the stochastic model of the least-squares processing of the vector components. They relate to a confidence interval produced from what is called the a posteriori covariance of the solution for each vector. There are different ways to deal with these stochastic results and in particular scale them realistically to the metric of the measuring system. The entire stochastic model of gps positioning is not as realistic as it is for angle and distance measurements because there is not a reliable and consistent way of predicting what one expects GPS errors to be. You can derive a realistic standard deviation of directional pointings of a theodolite by pointing and reading on a target perhaps 40 times. You can't do that with a range to a specific GPS satellite. About all that can be done is to relate the weighing scheme either to elevation angle or signal to noise ratio but again these are all relative to each other rather than the specific distance of a range. In other words they are all meaningful relative to each other but are not really directly tied to the length of a vector or, more importantly, the residuals produced. Frankly, they aren't all that useful, but residuals from one position to another are. Even measuring the same point at the same time from different base stations doesn't help that much. To really be sure of yourself, again in my opinion, you have to measure it more than once at differing times. The greater the time gap the better and the more repeats of the measurement, the better. Good luck with getting your software provider to really tell you what the error statistics mean in mathematical terms.

Latest update for future references,
As I mentioned I use PPK mainly to have the benefit of getting point positions from 2 bases. My DC doesn't allow me to get both
static, kinematic & RTK at the same time. So what I did was to process separately from 2 bases as was suggested here. I then
compared the x,y,z of the same point from the 2 bases. Made a small macro and eliminated the point if the results of the
x,y,z varied more than 0.40m. My reasoning is if I get the position of the rover from 2 bases & their difference is less than
0.40m then my point should be good, right? It is like having a redundunt measurement using a TS so to speak. So from 100,000+ results
of Fixed solutions, I am now down to less than 50,000. That is good too because I am now able to import these into a regular
Microsoft Excel file instead of using Microsoft Access which I am not too familiar operating.
As a final check, I then made some random measurements at regular grid intervals using a TS. The results from the TS & GPS
varies from 0.10m - 0.30m. This I expected since they are from differing technologies but results are ok for my needs for
topography mapping.

The contour map looks good with relatively smooth contours direct from the software. Some lines are crooked mainly due to
the exccess number of points used to process the DTM. Trying to smooth out the lines using the settings in the software.

Hope this helps surveyors who will encounter similar problems in the future. This method is still better compared to
the alternative of using a TS. I got 50,000 fixed points in less than 1 week for open terrain. Using a TS I wwould get maybe
at most 20,000 for the same time in the field.