uBlox f9 multi-frequency gnss

Can Emlid integrate this into a Reach RS?

The u-blox F9 platform will underpin the next wave of u‑blox positioning modules targeting mass market industrial and automotive applications. It uses GNSS signals in multiple frequency bands (L1/L2/L5) to correct positioning errors caused by the ionosphere and deliver fast time to first fix (Fast TTFF).

Its ability to receive signals from all GNSS constellations (GPS, GLONASS, Galileo, Beidou) further improves performance by increasing the number of satellites visible at any given time. Stand-alone u‑blox F9 solutions robustly achieve meter-level accuracy.

To achieve centimeter-level accuracy, u‑blox F9 offers optional on-chip real-time knematic (RTK) technology. In addition to offering an open interface to legacy GNSS correction service providers, it supports the main GNSS correction services, bringing RTK high-precision positioning to the mass market.


Wow this would be amazing!

Does this mean it exposes the raw timings the same way our current Reach devices do? And in that way, we could continue saving our raw logs and using the open-source RTKLIB for real-time and post-processing?

Or does this mean the chip handles the processing internally in a proprietary manner? And therefore loosing the ability to save raw logs and work with RTKLIB?

I hope it is the former.

This is a very early release of the chip, there is no info yet on the samples, availability or technical documentation so it will be a while until we see them in real life.

Nice and fair. Thanks for sharing.

This looks super amazing, but be careful RTKLIB is not really meant for dual frequency receivers yet. (https://rtklibexplorer.wordpress.com/2018/02/12/improved-results-with-the-comnav-k708-receiver/)

1 Like

@vbnhu I did a quick read of that link. What I took from that is the one particular dual frequency receiver had a superior internally processed result as compared to the RTKLIB result. Has the same variance in results been found with other dual frequency receivers?

Maybe you know more but it looks to me like RTKLIB does process dual frequency from what I saw there. It may not be perfect, but I guess there’s always room for improvement!

Besides, as @igor.vereninov mentioned, the uBlox f9 is in it’s infancy, so I think RTKLIB has time to catch up if it needs to. :slight_smile:

Dear Bide,

RTKLIB is claimed to be able to have “Ionosphere-free linear combination with dual frequency measurements [which] is used for ionospheric correction”. But on the other hand it is not able to do wide lane combination.
See the post: https://rtklibexplorer.wordpress.com/2017/06/20/a-first-look-at-rtklib-with-dual-frequency-receivers/
And some background:


With the introduction of these new, more affordable receivers I hope RTKLIB and other open-source software (like http://www.gogps-project.org/) will have better support for L2 and L5 measurements.

1 Like

@vbnhu Thanks for that! I’ll give those links a read and get up to speed.

I’ve got very good results with RTKPOST while post processing data from dual frequency receivers.
Nevertheless, PPP processing was inaccurate, and I didn’t try RTK with a dual frequency rover.

I happen to have some Commnav K706 dual frequency flight test data, alongside with a Reach:
We have a GPS GLONASS VRS NTRIP Base with 1Hz and a physical base with 20Hz GPS GLONASS data.

Of course the Reach and the Commnav antennas are different but mounted on the same aircraft. We are not sure if the poor performance in post processing comes from antenna, receiver or PPK software, but we are definitely far from where we want to be in terms of fix ratio and precision.

This topic was automatically closed 182 days after the last reply. New replies are no longer allowed.