GPS USB stick NEO-6T: correct expectations

So I’ve been playing around with the two NEO-6T dongles and RTKLIB to see how things work out in practice and what kind of precision I can get in the field. My antennas at the moment are pretty poor:

I’m going to buy some stuff at digikey for a different project and I’m considering buying $75 antennas like these:

There’s another available at half the price, minus magnet and a different antenna feed.

What I tested in the field is how long it takes to establish a proper solution for a base station location. After 30 mins, the variance was still at around 3m.

question 1. What should I expect in practice? A rover fix with an established base station is said to take about 10 minutes max. How long should I expect to get a usable fix for the base station itself?

question 2. Which of the above antennas would you recommend? Is it going to make the performance a lot better or would it not make a big difference, given the module only uses L1 and not L1+L2?


We’ve been able to achieve great results in the field with U-blox receiver, using them both as a base and a rover.
It requires a proper satellite skyview for fixed solution to appear quickly. If the satellites are visible and high then it will take just a few minutes. If the skyview is bad you may not receive a fixed solution at all.

We didn’t test that square antenna from Tallysman, but have been working a lot with the round one and it is really great - SNRs are ~50 and higher for satellites in zenith.

Ok. So just to confirm, you have been able to get a good base station fix in little time too, or did you need to run it for >3 hours?

There’s little documentation on how to establish that and in one article I read they left the in “SINGLE” mode until the solution converges. As I said, when I did that with these antennas in a not so optimal place, I waited 30-40 minutes and read 3m from the greyed out numbers just below the lat/lon/alt solution on screen (RTKNAVI). From other statements online, they claim you need to leave the module in place for more than 6-12 hours before it converges. I checked on the plot and the position indeed fluctuated n/s and e/w by 2-3 meters over time.

Is there a good tutorial / article online that describes how to achieve a usable base station fix, explaining the specific settings you need to make in the “options”?

Have you seen ?

There are several methods of determining base station position:

  1. If you have another base with with known position you can setup your base as a rover with option “static” and wait for fixed solution. After fix you can use this coordinates as position of your base.

  2. Log raw data for a long time (several hours) and then postprocess it in RTKLIB.

Precise base coordinates are required only if you need precise absolute coordinates for rover, if relative position is sufficient you can simply use coordinates from single solution.