Inconsistent correction results

I have done a ppk survey with inconsistent Y,Z results.
Initial: Setout and survey GCPs
Check: Check GCPs.
The initial and check survey was done during the same base setup.
Base: IT 175.021822 -40.89502974 13.948 (NZGD2000)

Attached are: (Link below)

  • Both Base and Rover Observation files
  • Survey observation file
  • Corrected file (CSV)
  • Corrected file (excel)

For some reason the X-coordinates compares well (Setout/check) BUT
Y and Z has some real big discrepancies. (Spread).

When does the lines turn orange (processing)?
Is the data still usable when I fix a point?
How do I make sure that when I fix a point, that the data is sufficient?
Is it possible to have a warning on the screen if the Rover has not sufficient data to position accurately?
I have checked the satellites and pdop continuously but it means I have to go to receivers. It is a roundabout way and then I still missed a point although I surveyed it.

Can you advise please.

Hi @Wouter,

I see there is quite a lot of data. I’ll examine it all thoroughly and get back to you.

Thank you, Svetlana.
I really appreciate your efforts.

Hi @Wouter,

I think I finally found the reason for this. Initially, I received the same results, but then noticed that QZSS signals are weirdly affected by something in your area. Also, Beidou satellites are noisier than others. Here’s the screenshot of the rover raw data:

QZSS looks the same on the base. Since you’re working with single-band devices, it’s better to exclude all such factors. So, I’ve calculated the solution based on GPS, GLONASS, and Galileo satellites only. And obtained averaged spreads for all coordinates between 5 and 8 mm.

Also, I changed the Filter type to Combined. Otherwise, the z107x2 point was in Float.

Can you check if it works the same way for you, please? I’m very curious to sort it out.

Hi Svetlana,
Thank you for all your effort.
I have re-calculated with the settings you have advised.
My results are now similar to yours.

But z105 had major discrepancies due to 1 observation wasn’t fixed.
I re-calculated with "fixed"only and adopted the lower result.

I have done the survey before with the base on the BM.
I will re-calculate that too for a comparison.

I will leave Beidou out of my operation completely from now on.

One more question: Can the new RX be linked to the RS+ ?

All the best and thank you, again.

Otaihanga Results 221103

Nope, the RX expects a multi-frequency correction-signal.
And as the RX is a dedicated network-rover, it cannot function as a correction source for the RS+ (the RS2-series can though).

Hi @Wouter,


Have you tried Combined Filter type? Maybe it will help obtain a Fix for z105?

Actually, Beidou can be better on another site. So, I’d not give up on it forever. If you need some help with raw data quality analysis, feel free to ask me!

Christian speaks the truth. Reach RS2+ can be a base for Reach RX, Reach RS+ can’t. But if you don’t have one, you can research NTRIP services in your area. They all usually provide multi-band corrections.

This topic was automatically closed 100 days after the last reply. New replies are no longer allowed.