Support Request: Large Positioning Error from 10-hour u-blox ZED-F9P Recording

GPS/GNSS
1

Dear [Support Team],

I am currently testing the u-blox ZED-F9P module with the SparkFun GPS-RTK-SMA kit.
I recorded raw GNSS data in UBX format for about 10 hours, then converted it to RINEX (*.obs) using RTKLIB (RTKCONV).

After uploading the *.obs file to CSRS-PPP and OPUS for precise point positioning (PPP) post-processing, the returned solution shows a position error of over 1 meter, with relatively large uncertainties (1–3 m at 95%).

This is unexpected, because based on ZED-F9P specifications, I would expect centimeter-level accuracy in PPP (long session, dual-frequency).

Could you please advise:

  1. Is the SparkFun GNSS antenna ANN-MB-00 supported by OPUS/CSRS-PPP calibration databases?

  2. Are there specific RINEX conversion settings (RINEX version, antenna type, marker name, etc.) recommended for ZED-F9P raw data?

  3. Does the ZED-F9P log contain sufficient ephemeris/broadcast navigation data, or should I rely on CSRS/IGS data servers for orbit/clock corrections?

  4. Do you have any recommended workflow for improving accuracy in PPP post-processing with the ZED-F9P?

Your guidance would be greatly appreciated, as I would like to understand whether this is a hardware/antenna limitation, or a configuration/processing issue.

Thank you very much for your support.

Best regards,

2

Hello @Nguy_n_Vo_Ng_c_Th_ch ,

Could you please share the “Phase Ambiguity Status” data from the full CSRS report? It will help identify if your antenna and GNSS signal is causing the issue. You should see mostly green Fixed ambiguities:

image
image769×847 20.8 KB

It looks like your antenna was under trees, close to a high rise building? Perhaps the GNSS signal is the issue?

Best wishes,
Paul

3

Dear Mr. PaulZC!

(attachments)

COM6___57600_250820_145924.pdf (1.22 MB)

4

Dear @Nguy_n_Vo_Ng_c_Th_ch ,

Thank you.

I think it is clear you have bad GNSS signal reception. There are no Fixed Ambiguities in your data:

image
image504×264 25 KB

It looks like you lost signal intermittently for the last 2 hours. The data plots are broken several times:

image
image326×460 28.7 KB

I suggest trying again. I suggest taking your equipment to an area with a clear, unobstructed view of the sky. Collect data for perhaps 1 hour, and submit that to CSRS. If you have a car, you can collect data from your car with the antenna in the center of the roof.

I hope this helps,
Paul

1 Like
5

image.png
image.png1911×1054 562 KB

I am doing a survey like this picture. Please give me advice. Thank you.

Vào Th 5, 21 thg 8, 2025 vào lúc 15:57 PaulZC via SparkFun Community <notifications@sparkfun.discoursemail.com> đã viết:

6

I suggest taking your equipment to an area with a clear, unobstructed view of the sky. Collect data for perhaps 1 hour, and submit that to CSRS.

Perhaps you can access the roof of your building?

7

Please also post a better image of the satellite signal strengths. Your screenshot is not clear:

image
image492×104 55.3 KB

8

OK. I will try again, and let you know later. Thanks

Vào Th 5, 21 thg 8, 2025 vào lúc 17:00 PaulZC via SparkFun Community <notifications@sparkfun.discoursemail.com> đã viết:

9

@Nguy_n_Vo_Ng_c_Th_ch , the data you posted is L1, not L1 + L2.

Your results are “as expected”.

image
image705×419 61.2 KB

Here’s an example of what you’re looking for.

Also note: you will never be able to determine the actual position of a Mark/Point with that type of antenna, unless you can find the Electrical Phase Center for it (doubtful). There are much better choices for PPP, such as this one on the SparkFun site.