RTPK issue

[Mark Wheeler]

This morning (with LS+ on state RTN) I shot a USGS bench and RTPK worked fine and was in agreement. Late morning, in another job I shot a point 3 times (3 minutes, 3 minutes and 10 minutes). In a fairly open environment (but close to high tension lines) the RTK collected well, but the RTPK (set for every 30 seconds) would come up in disagreement by 4-10' hor, 3-10' vertical for a round or two. But by the end of the shot bowed out of an RTPK answer. Since there was an issue I upgraded to Jfield 3.0.11.824 and am still running GNSS 4.2.0211.011. Late afternoon I tried collecting a point in the Job that worked in the morning. Any RTPK results were many feet off in both directions and the point did not store an RTPK value.

I then updated an LS (different unit) to Jfield 3.0.11.824. It is running on GNSS 3.7.10.210408. Went to the same spot outside and it collected RTK fine but could not come up with an RTPK value (set for every 30 seconds). All shots were in the fairly open sky.

Could this just be an anomoly that happened durning the day? What local or environmental factors might cause RTPK to be incorrect? Again, the LS+ RTPK worked earlier in the morning. I will check this out again tomorrow to see if the issue resolved itself.
Mark

 

[Alexey Razumovsky]

Mark, Send a project to Support would help to investigate the issue

 

[Mark Wheeler]

Projects sent: 1. LS 479 Test2msm4: Points 121 & 123 2. LS+ 803 Test base rover: points 236-239 11/20 am (RTPK worked) , points 240-242 11/20 pm (RTPK Bad), points 243 RTPK Bad, 244 (Rtpk good for 1 round). Under these conditions RTPK will verify very quickly.

 

[Mark Wheeler]

Alexey. Typically the best time of day in our area is early to mid morning. This morning I tried the LS+ for a few points and RTPK worked normally. Later this afternoon I tried it again with RTK normal, failing RTPK. Before supper did another 6-8 points (off RTN in one spot). RTK worked perfect. RTPK was poor (either 8-10' hor & vert or no solution). When I turned Glonas off the RTPK worked normally again. Flipping Glonas on and it was either failed or Float quality solutions. In my situation it looks like Glonas is messing up RTPK.

 

[Alexey Razumovsky]

Mark, thank you for the data. I appreciate alot projects like this.
I looked into the base data and saw that RTN provides only L1 Beidou signal. It may mislead RTPK behaviour in some cases. You might switch off BDS instead of Glonass.
In the meantime it's not a critical problem for RTPK v.4. I have got good solutions for all points mentioned above. Intervals 30, 60 yields solutions. Interval 120 (default value) doesn't. An interference appears at the end of observation and 2 solutions (120 sec interval splits file in 2 parts) doesn't agree. So in open shorter intervals are better.
I would recommend also to reRTPK vector with updated interval in a case like this.

 

[Mark Wheeler]

Alexey,
For consistency I shot the roof of my truck in an open parking lot at three different intervals today (11/23/2021 EST). Points 255-256 (7:27-7:30am), points 258-269 (11:24am-12:02 pm) , points 270-272 (5:04-5:10 pm)/ The point description notes if a Constellation is shut off for that point collection. Today re-sent updated File: Test baserov. Unit 803 to support. The RTN generally provides corrections for Bedoi B1, B2, B3.
Thanks,
Mark

 

[Mark Wheeler]

Alexey,
The last file sent to support (Test baserov) on Tuesday was from the LS+ where the RTN gives corrections for B1,B2 & B3 if available.
Mark

 

[Alexey Razumovsky]

Mark, I saw B3 on one satellite only.
I found an issue with local RTN. RTPK dismiss B2 signal. Just fixed.

 

[Mark Wheeler]

Alexey, when you said you just fixed this issue, are you referring to the next release?
Mark

 

[Alexey Razumovsky]

Alexey, when you said you just fixed this issue, are you referring to the next release?
Mark

Yes

 

[Mark Wheeler]

Alexy, A few more pieces of Information. The RTPK issues were with the LS standard (as rover) and LS+ (as rover) using the State RTN. Earlier in the year RTPK worked great with the State RTN. Now it works correctly about 25% of the day, probably depending on sat. coverage. This morning I tried LS+ (rover) with LS (base) and it worked perfectly in the open and in thick woods. In the open it only took seconds to get RTKP agreement. When I switched by to the State RTN, no good RTPK results.

 

[Alexey Razumovsky]

I wanted the the files that RTPK struggles to fix. Its a good opportunity to check updated engine.

 

[Mark Wheeler]

Alexey,
11/26/2021 sent file Test baserov to support. location fairly open to sky, light rain. 1. Points 277-278 10:55-10:57 am EST. This is still good rtk window. 2. Points 279-282 11:49-11:59 am. This is where things typically deteriorate locally for RTK. Not terrible today.

 

[Alexey Razumovsky]

Alexey,
11/26/2021 sent file Test baserov to support. location fairly open to sky, light rain. 1. Points 277-278 10:55-10:57 am EST. This is still good rtk window. 2. Points 279-282 11:49-11:59 am. This is where things typically deteriorate locally for RTK. Not terrible today.

Thanks

 

[Alexey Razumovsky]

Mark, be aware aren't enough Beidou B2 signals in MA state RTN for post processing in RTPK mode

 

[Mark Wheeler]

Alexey,
I think RTPK is great. I have been running it on my standard LS for a year and on the new LS+ since early summer. RTPK has worked great on the MA RTN for quite awhile. It wasn't until very recently that I have had RTPK solution issues on both units. I updated a few times to the newest pre-release in order to try to minimize the down time during the middle part of the day. Something has changed. I think it is more than B2. I tried a quick test tonight and the RTK was real slow and RTPK was wrong or no solution. I went out again a little later and alternated turning glonas off.
1. Shooting points 283 (bad RTPK) & 284 (no RTPK) . 2. Turning glonas off 285 (Good RTPK), 286 (good RTPK) 3. Turned Glonas ON (Bad RTPK), 286 (Bad RTPK), 4. Turned glonas off again 286 (good RTPK). The point descriptions are "TABLE" for all constellations and "Table no glonas" for GPS, Gal & Bed only. I resent the updated project Test baserov at 7:10 pm EST Unit 803.

 

[Alexey Razumovsky]

Mark, thank you for the files. Below Justin processing results.

It seems that ~1 min long sessions aren't enough for stable RTPK solution in range up to 17 km in medium canopy. New RTPK failed on pp.283,284.