Hi, I’m taking over a dGPS set up for a rover test coming up soon. Previous lead on this bit of the project has gone on holiday for a while. I’m attempting to learn by looking through all these posts and manuals - have a few leads but don’t want to spend too long going down a dead-end (only have a week or so until our shakedown).
Problem: RS shows drift (1m within 30s) after being stationary for 30mins. Location, open car park, some trees but only on one side. I’d estimate 135deg visibility where the trees border the car park & basically unobstructed in a perpendicular direction.
We’re not using NTRIP yet (sorting out month subscription today/tomorrow) BUT would / could that result in drift of a static base after 30mins being stationary? Status tab shows 7 good satellites most of the time.
I’d expect NTRIP to correct bias error. But then I’m still learning the ropes here.
If your receiver doesn’t receive base corrections, it’ll calculate only a Single solution, which is a few meters accurate. That’s why the position can drift even if the receiver is placed at the same point for a long time.
However, the rover always receives the same base position that is saved in the Base mode tab on the base. So, if your base continues recalculating its coordinates with a Single solution after you set it up, it’s ok and doesn’t affect the solution calculation on the rover.
That makes sense. We ran a longer test yesterday, and saw exactly what you describe - position estimate varies, drifiting within something like a 3m diameter circle.
We do plan to use dGPS for our rover test. So I’d expect that the relative measurements (i.e. the difference between the gps solution on the base and the rover) is much more stable than the absolute solution for each receiver.
Because we need heading, as well as position for our trials - we’re using a different manufacturer’s gps unit (actually it’s a mini INS, the sbg ellipse 2-N) as this is compatible with 2 antennae on the rover. But I expect that both emild and the gps in the sbg will exhibit the same instantaneous drift in absolute position estimate (down to a few cms anyway) due to variation in propagation of satellite signal to those receivers.
Does that sound right to you? If so, there’s no need for the NTRIP subscription (which I ran out of time to purchase yesterday, luckily).
If you’re going to set the receivers as base and rover, you indeed can work in RTK without an NTRIP subscription. The receivers placed close to each other are in almost the same atmospheric conditions. So, the error caused by the GNSS signal distortion in the atmosphere is fixed in RTK, and you obtain centimeter-level accuracy with the rover. You can read more about it in our How RTK works guide.
One note: to obtain coordinates in a particular coordinate system with the rover, you’ll need to place the base above a known point. If you average the base’s position with a Single solution, you’ll get only relative accuracy. It means that the rover’s coordinates are accurate only to the base.