Difference between PPK coordinates and NTRIP coordinates

[oajioka]

We recently purchased a new Triumph LS with cellular network access for use with the CRTN. I performed a quick test by setting up the Triumph 1M base station and taking some shots on control points we have set near our office, and then taking down the base and connecting to the CRTN and repeating the same shots.

After DPOS the coordinates from the first set of observations match closely with readings we have taken in the past, while the coordinates from the observations made while connected to the CRTN differ by ~2 feet Northing and ~1 foot Easting. My location is in Novato, CA and the CORS station used was "SVIN" about 5-6 miles away. Signal strength to the cellular network was strong throughout.

Attached are the reports from the two projects, where "A" is with the base station and "B" is connected to CRTN.

Wondering if I am missing something in my settings or what. Any insights are appreciated, thanks!

 

[Shawn Billings]

Check the base reference frame for your CRTN profile. It's probably set to WGS84 instead of NAD83.

 

[oajioka]

Hi Shawn, thanks for the reply!

The base reference frame for my CRTN profile was set to NAD83(CSRS). Am I correct to assume I should be using NAD83(2011)?

 

[Jim Frame]

I didn't even know that NAD83(CSRS) was a choice, I've been using NAD83(2011). But I rarely use the derived coordinates for anything beyond search purposes, as I almost always adjust the vectors to known control later.

oajioka: I have a request. For some years I used the CRTN station UCD1 as my RTN base, as it's only a couple of km from my office. However, BARD made some changes that have made it impossible for me to connect to it. Attempts at troubleshooting with CSRC have gotten me nowhere so far, and I'd like to verify that the problem isn't in my setup somewhere. If you find the time, will you attempt to connect to UCD1 and report your results? Thanks!

 

[oajioka]

oajioka: I have a request. For some years I used the CRTN station UCD1 as my RTN base, as it's only a couple of km from my office. However, BARD made some changes that have made it impossible for me to connect to it. Attempts at troubleshooting with CSRC have gotten me nowhere so far, and I'd like to verify that the problem isn't in my setup somewhere. If you find the time, will you attempt to connect to UCD1 and report your results? Thanks!

Jim, I was able to successfully connect to UCD1 via office WiFi this morning. Below is a screencap of my connection settings. When you say you almost always adjust the vectors to a known control later, are you using the "recover base point" feature and then adjusting the recovered base point coordinates to the published station coordinates? Or are you doing a static occupation of a local control point that you also collected with CRTN RTK? Would love to know more about your methods since we are just starting out with this CRTN business.

Thanks, and I appreciate your reply!

 

[Jim Frame]

I was able to successfully connect to UCD1 via office WiFi this morning. Below is a screencap of my connection settings.

I should have been more clear: connecting isn't the issue, it's getting usable correction data. When I connect to P268 (another CRTN station near me) I get good corrections, like so:

(This was taken in my driveway, which has a heavily obstructed sky view, so not great SV reception.) Clicking on the fix/float box shows this:

However, when I change the mountpoint to UCD1, the RTK engine display shows only this:

I believe that UCD1 got misconfigured -- its message stream is different from that of other CRTN stations, as can be seen when you connect via BNC -- but after some initial interest by CRTN management (lots of back-and-forth email with Yehuda Bock, and with Mario Aranha, who manages UCD1 for BARD) I think they believe the problem is on my end. If I can get someone with a different receiver to tell me that they can't get usable data from UCD1 either, I'll be able to go back to CRTN and say that the problem isn't just with my equipment setup.

With regard to my workflow, what I typically do is record the vectors, export them in Star*Net format, and in the Star*Net data file fix the position of the mountpoint. I generally use the mountpoint's values from the CRTN database, though sometimes I use the datasheet position, an OPUS position, or an OPUS Projects position, depending on the needs of the project. On occasion I need to anchor the project to one or more ground stations, in which case I let the mountpoint float and assign appropriate standard errors to the desired ground station positions instead. Star*Net also allows me to translate and/or rotate the adjusted coordinates to a local system or basis of bearings easily, which is useful for boundary jobs. I admit that I'm not a typical RTK user -- and sometimes I'm a pretty dunderheaded one (I still struggle with localizations for search purposes, I just can't seem to get the process clear in my brain) -- but nevertheless I find it a very useful tool in my work.

 

[Sdrake14]

I should have been more clear: connecting isn't the issue, it's getting usable correction data. When I connect to P268 (another CRTN station near me) I get good corrections, like so:

View attachment 9520

(This was taken in my driveway, which has a heavily obstructed sky view, so not great SV reception.) Clicking on the fix/float box shows this:

View attachment 9521

However, when I change the mountpoint to UCD1, the RTK engine display shows only this:

View attachment 9522

I believe that UCD1 got misconfigured -- its message stream is different from that of other CRTN stations, as can be seen when you connect via BNC -- but after some initial interest by CRTN management (lots of back-and-forth email with Yehuda Bock, and with Mario Aranha, who manages UCD1 for BARD) I think they believe the problem is on my end. If I can get someone with a different receiver to tell me that they can't get usable data from UCD1 either, I'll be able to go back to CRTN and say that the problem isn't just with my equipment setup.

With regard to my workflow, what I typically do is record the vectors, export them in Star*Net format, and in the Star*Net data file fix the position of the mountpoint. I generally use the mountpoint's values from the CRTN database, though sometimes I use the datasheet position, an OPUS position, or an OPUS Projects position, depending on the needs of the project. On occasion I need to anchor the project to one or more ground stations, in which case I let the mountpoint float and assign appropriate standard errors to the desired ground station positions instead. Star*Net also allows me to translate and/or rotate the adjusted coordinates to a local system or basis of bearings easily, which is useful for boundary jobs. I admit that I'm not a typical RTK user -- and sometimes I'm a pretty dunderheaded one (I still struggle with localizations for search purposes, I just can't seem to get the process clear in my brain) -- but nevertheless I find it a very useful tool in my work.

I still fumble with localizations myself but good survey QC is always my safety net and I like you find them nice tools when applicable. I use them as searches and sometimes my net is junk but if I am close I use shifts on 1st found land corner to narrow it down on the fly and then generally scrap the system at the end of the day in favor of CCS83 datum for postprocessing then scale and rotate in the end in CAD if I need to be at a record......this seems to keep me out of trouble...

 

[oajioka]

I believe that UCD1 got misconfigured -- its message stream is different from that of other CRTN stations, as can be seen when you connect via BNC -- but after some initial interest by CRTN management (lots of back-and-forth email with Yehuda Bock, and with Mario Aranha, who manages UCD1 for BARD) I think they believe the problem is on my end. If I can get someone with a different receiver to tell me that they can't get usable data from UCD1 either, I'll be able to go back to CRTN and say that the problem isn't just with my equipment setup.

Jim:
Sorry for the delay, I had a chance today to connect to UCD1 and I ran into the same problem: It says I am connected to UCD1, but when I go to the collect screen it shows "STN" and my RTK display shows no data, see screen caps below. SVIN mountpoint still gives me good corrections, which seems to suggest that the problem isn't on your end...

Hope that helps!

 

[Jim Frame]

Thanks for verifying that! I'll resume my conversation with Yehuda. (I'd love to have a Trimble user report their experience as well just to make sure it's not a Javad thing, but so far I haven't been able to find any who use CRTN to try it.)