OK, I now have a Bug, that affects Reverse shift, and dpos

Nate The Surveyor

Well-Known Member
So, I go to a job, that is already started. I have started this job several years ago, with the Topcon gear. I have made a sideshot, with the Topcon Legacy E, using TDS. I set a RR spike, for an extra base. Point Number 960. It has an elevation of 148.17'. It's a BOGUS elev. but for the most part, this job is for HORIZ control.



Now would like to occupy this point 960, but it's in a cow field. So, I move over some 150', and set a NEW base. It is called BASE3. I make it an autonomous coord value.

I want to use M-Local to do this.

Then, I go and take a shot on number 160 above. This is NEW point 1474. It shows up as being about 2 feet off horizontally, and a huge amount vertical. This is because the old Topcon based coord has a bogus Elev.

So, I realize this. And, I make a NEW point 960, and call it 960El. thats 960 with the Elevation of 1474 included with it.
Now, I localize. Then, I shoot a number of other points, and EVERYTHING is good.

I come home. And, DPOS it. NOW, I get a shift of 1.99' FROM the AUTONOMOUS position. BUT, it is only about 0.05' from the coord created by M-local.
NOW the trouble starts.
All the other points, that should be fitting by hundredths, are missing by 0.8'. And it is ALSO showing up as a difference between RTK and RTPK.

I would like it if somebody would look at this file, either via RAMS, or I can send the file. It has some 1500 coords.

I am trying VERY hard, and it has a BUG in it.

Number 1474 SHOULD be right on top of 960El. And the separation between RTPK and RTK is very close on this point.

But, the rest of them are about 0.80'.

I'm in over my head now.

Frustrated.

Nate
 

Michael Stazhkov

Developer
JAVAD GNSS
So, I go to a job, that is already started. I have started this job several years ago, with the Topcon gear. I have made a sideshot, with the Topcon Legacy E, using TDS. I set a RR spike, for an extra base. Point Number 960. It has an elevation of 148.17'. It's a BOGUS elev. but for the most part, this job is for HORIZ control.



Now would like to occupy this point 960, but it's in a cow field. So, I move over some 150', and set a NEW base. It is called BASE3. I make it an autonomous coord value.

I want to use M-Local to do this.

Then, I go and take a shot on number 160 above. This is NEW point 1474. It shows up as being about 2 feet off horizontally, and a huge amount vertical. This is because the old Topcon based coord has a bogus Elev.

So, I realize this. And, I make a NEW point 960, and call it 960El. thats 960 with the Elevation of 1474 included with it.
Now, I localize. Then, I shoot a number of other points, and EVERYTHING is good.

I come home. And, DPOS it. NOW, I get a shift of 1.99' FROM the AUTONOMOUS position. BUT, it is only about 0.05' from the coord created by M-local.
NOW the trouble starts.
All the other points, that should be fitting by hundredths, are missing by 0.8'. And it is ALSO showing up as a difference between RTK and RTPK.

I would like it if somebody would look at this file, either via RAMS, or I can send the file. It has some 1500 coords.

I am trying VERY hard, and it has a BUG in it.

Number 1474 SHOULD be right on top of 960El. And the separation between RTPK and RTK is very close on this point.

But, the rest of them are about 0.80'.

I'm in over my head now.

Frustrated.

Nate
Nate, please send the project to support.
 

Nate The Surveyor

Well-Known Member
OK, its named Bonnerdale231
and, my LS Number is 00435
and, it is sending now. It's at 5% sent.
Nate
PS it may be there in about an hour
 
Last edited:

Nate The Surveyor

Well-Known Member
OK, a few questions.
1.) did you receive the file?
2.) have you found the problem?
3.) is it fixable, or do I need to do something?
REVERSE SHIFT is an incredibly useful program. But, not if it introduces errors.
Thank you,
Nate
 

Nate The Surveyor

Well-Known Member
Another Javad User just called me, and said he'd had an anomaly with reverse shift too. He thought it had somehow "Lost" the Reverse Shift, somewhere during the day. He said he had set 5-6 corners, and had to go back the next day and reset them. He said he had "Fixed" in in his office software, (Carlson) and uploaded it all to DESIGN points, rather than to understand and solve what went wrong.
I do think there is something wrong. But, I don't have it solved yet.
Thank you,
Nate
 

Michael Stazhkov

Developer
JAVAD GNSS
Ok, any answers yet?
Justin team said that Justin doesn't see any discrepancy between RTK position recorded in the JPS file and the post-processed RTPK position. I am trying to reproduce that result in the J-Field and then compare the result and what is recorded in database. It can help to understand the issue.
 

Nate The Surveyor

Well-Known Member
20220305-03.18.22_00435_Processed_Point_Info.png

20220305-03.19.56_00435_Processed_Point_Info.png
20220305-03.20.05_00435_Processed_Point_Info.png

Maybe this can help.
Nate
 

Michael Stazhkov

Developer
JAVAD GNSS

Michael Stazhkov

Developer
JAVAD GNSS
So, I go to a job, that is already started. I have started this job several years ago, with the Topcon gear. I have made a sideshot, with the Topcon Legacy E, using TDS. I set a RR spike, for an extra base. Point Number 960. It has an elevation of 148.17'. It's a BOGUS elev. but for the most part, this job is for HORIZ control.



Now would like to occupy this point 960, but it's in a cow field. So, I move over some 150', and set a NEW base. It is called BASE3. I make it an autonomous coord value.

I want to use M-Local to do this.

Then, I go and take a shot on number 160 above. This is NEW point 1474. It shows up as being about 2 feet off horizontally, and a huge amount vertical. This is because the old Topcon based coord has a bogus Elev.

So, I realize this. And, I make a NEW point 960, and call it 960El. thats 960 with the Elevation of 1474 included with it.
Now, I localize. Then, I shoot a number of other points, and EVERYTHING is good.

I come home. And, DPOS it. NOW, I get a shift of 1.99' FROM the AUTONOMOUS position. BUT, it is only about 0.05' from the coord created by M-local.
NOW the trouble starts.
All the other points, that should be fitting by hundredths, are missing by 0.8'. And it is ALSO showing up as a difference between RTK and RTPK.

I would like it if somebody would look at this file, either via RAMS, or I can send the file. It has some 1500 coords.

I am trying VERY hard, and it has a BUG in it.

Number 1474 SHOULD be right on top of 960El. And the separation between RTPK and RTK is very close on this point.

But, the rest of them are about 0.80'.

I'm in over my head now.

Frustrated.

Nate
Hi Nate. It seems you are right, original RTK coordinates were changed somehow. DPOS / RTPK coordinates are correct. I have compared original RTK coordinates, recorded into the JPS file and, as Justin team said, it is almost the same compare to post-processed.
At the moment we don't reproduce the issue on a test project yet. We are trying.
I am going to add some protection into the J-Field database to detect such attempts to rewrite the original RTK coordinates.
Thank you for pointing out the issue.
 

Nate The Surveyor

Well-Known Member
@Michael Stazhkov , thank you for looking into this. Reverse shift was "broken" for a while. I assumed it was probably fixed now. But, I have had a couple of malfunctions with it in the past. I could not figure out what went wrong... But this time I tried to.

We, the end users are in a difficult situation here. Kind of a blind trust situation. We don't do the math, on GPS, and we use it all the time. This forces us to find anomalies, and point them out, but we don't fully understand them.
So, thank you for understanding, and giving your attention to this.
Do what you think is best. I'm guessing you saw my concerns, ie, possibly some coords were off by 0.80'
Nate
 
Top