Like most surveyors I spent the weekend setting up an NTRIP Caster at my office. I already had a Leica 1200 Pro transmitting corrections on an TCIP address so was not that much of a task just had to drill some holes in my firewall for the new port.
The problem is when I connected to the caster (the rover was about 150 feet from the base) it locked right away, my initial precision was great but as I watched it slowly degenerated to a point where I lost the fixed solution. I then disconnected from the caster and reconnected and got another good lock but then it would gradually get worse. As long as I was willing to disconnect and reconnect I would have a good solution but the precision always gradually decreased until I would lose the solution. I setup a second unit and connected to Leica Smartnet and locked in and the precision would stay good. I setup a third unit and locked in the to TCIP connection to my base transmitted from "REALTIME 2", same gps base different port address, and locked in with good and steady fixed solution.
The first unit I connected to my caster was an old Leica 1200 and as a final check I connected with one of our new Leica GS15's with the same gradual loss of lock as the 1200.
My base is a Leica 1200Pro with GPS only no Glosnass.
Has anyone else seen this using a Caster or any idea what the problem might be. I am connected to the base with a Windows Server 2003 using Com1 to Port1 9600,n,8,1 with RTCM 3.0 at 1 second intervals.
gschrock, post: 366653, member: 556 wrote: This is puzzler as there is nothing distance dependent in the nature of a caster; it is simply an "exchange" for data streams... Casters like the Unavco and JPL ones are accessed by science folks in other parts of the world (though not for RTK corrections) and only have to worry about latency. Smartnet delivers corrections to you via a caster in another state.
Without connecting and doing some tests I can't see a lot of variables; could try a higher baud rate like 19200 on the serial connection (though that should not matter).
Thanks for replying, I was thinking about the problem and did a decode on the message string and it seemed to have an overlap so I used the option in the Leica 1200 to add a "Carriage Return' - CR" at the end of the message and that solved the problem right away. The gps unit locked on at a 0.05 and stayed that way.
Next problem is I connected a second unit to the caster and fixed and the precision on both units dropped to over 2.5. I'm going to redo this later in the day to make sure not a problem with the sv configuration. Should the quality of the data stream go down with the number of users connected?
Is it possible your base position is not correct? for every 6 meters or so of error at the base, that causes a 1 ppm error at the rover. If you are far enough off it may never initialize, or initialize incorrectly and then lose the init.
John Hamilton, post: 366680, member: 640 wrote: Is it possible your base position is not correct? for every 6 meters or so of error at the base, that causes a 1 ppm error at the rover. If you are far enough off it may never initialize, or initialize incorrectly and then lose the init.
Hi John, The base was a part of the Leica network until a couple of years ago so the position should be good. I send a 12 hour rinex file to OPUS every couple of months to be sure it hasn't moved plus when I had a good solution with the Caster I did a check point at a control point at the office and the elevation was within 0.03 feet.
Just a thought...sometimes software expects decimal degrees, other times DD.MMSSsssss, etc. without it being labeled, and watch the sign of the longitude! Most software expects west to be negative.
