What am I doing wrong? (Can't get a reliable RTK Fix)

Q&A forumCategory: QuestionsWhat am I doing wrong? (Can't get a reliable RTK Fix)
beatified asked 8 months ago

I have the simpleRTK2B – Starter Kit LR and a 24 inch ground plane.
I need help trying to get a reliable RTK fix with NTRIP data so that I can setup my base station.
I spent probably 3 hours today trying different things to see how things were working. I tried to document my lack of success in the hopes that someone might be willing to school me.
Here are some pictures of my environment and the way I have the antenna setup. This is temporary but I do intend on installing things roughly this way.
I have both the Mag mount antenna and the survey style non calibrated antenna
https://imgur.com/a/BKivCYx
And here are some screenshots of my results
https://imgur.com/a/8PtgnKr
Check the descriptions with the pictures for details.
This is the location of the SNHS NTRIP Data 3.51 Miles 5.64 km from my location
And the location for the CCCS NTRIP Data 3.33 Miles 5.36 km from my location
And my location.
I never got a RTK Fix for more than a few seconds at a time. And even that didn’t happen very often.
Can you please advise?
Would a better antenna help?
Am I fighting an up hill battle with this location? If so I have other options actually any roof since the apartment complex is owned by my family.
Is there something configured that I messed up in U-Center even though I uploaded the config file as directed in the Configuration Files page?
Thanks for your help.

Ardusimple Staff replied 8 months ago

Hi beatified,
A few comments/ideas:
1. With the survey antennas there’s no need of additional ground plane.
2. The installation location seems to have not full visibility of the sky. It should be enough to get centimeter accuracy but it’s not ideal conditions.
3. We see superlong antenna cable, more than 5 meters is not recommended.
4. Electronics are not boxed. Airflow/direct sunlight affects stability of the zed-f9p, it’s important to use it inside a box (even carton box is enough for tests).
5. 5.5km should be ok to reach centimeter level accuracy.
6. Make sure your base station is not in survey mode at the same time that is receiving corrections from the base station.

beatified replied 8 months ago

Thanks so much for the pointers.

Should I be concerned about the temperature of the SimpleRTK2B? Can it handle being is shade like for instance below the Ground Plane?

Does stable temperature help the performance?

My cables are the cables that came with the antennas so they shouldn’t be too long. But I would like to find shorter cable for the Survey Antenna. I see the connector for the survey antenna is TNC-K is this ever referred to differently? I really don’t see much labeled as TNC-K. I would Ideally like much shorter maybe 500mm. And what impedance is the wire? 50 Ohm?

Thanks again

2 Answers
clive1 answered 8 months ago

It doesn’t look like a terrible location, although the type of ground plane you have here is primarily suitable for flush mounting antennas.
What does the deviation plot look like for the antenna with the receiver in unaided 3D mode, over a 24 hr period? You’re looking to see a tight clustering, with at least 99 of the fixes within a 2.5 to 3.0 m radius. If not you need to look for a better location. Figure you want a clear view south, as most satellites aren’t arcing into the north +/-30 degrees. uCenter’s sky view shows visibility.
Personally I would not rely on random NTRIP providers.
Find a NOAA CORS close to you, and use that to establish a survey grade location for your base antenna. Google imagery tends to have distortions, but should be adequate to about a metre or so. Your picture with the pin on another wing of the apt complex is definitely not good.
https://alt.ngs.noaa.gov/CORS_Map/
AZU1, CNPP and TRAK are about 30 KM out, probably a bit further than I’d like but going to have better provenance, and give a feel for the quality of data you’re outputting, and if the post-processing can resolve the ambiguities, and get some tight convergence on a solution.
 

beatified replied 8 months ago

Clive, Thanks for the Tips.
Hopefully I’ll be able to do some more testing in the next day or two when I get a little time.
I haven’t done a deviation plot as of now at least not at that location but I will give that a try. And its good to know what to try to shoot for as far as quality of fix without RTK.
I have pretty much absolute freedom as to where I place this thing as I have both internet and power on all roofs in this complex. Here is a map that might be a little better quality in some ways.
How Can I tell good NTRIP data sites apart from bad?
Thanks again.

clive1 replied 8 months ago

>How Can I tell good NTRIP data sites apart from bad?
It gets to be a bit subjective.
The NOAA-CORS roster has a lot of quality checks, and auditing.
Infrastructure run by local municipalities, counties and states, often DoT or GIS Dept run, are usually safe as they are managed by surveyors and they are reliant on these things to be correct.
Randoms running NTRIP via RTK2GO, I have less confidence in.
If you have a good survey grade location for your antenna, you’re going to know if a given NTRIP is good to a few cm, or metres off, or has some vertical offset.
One method to quantify/audit an unknown provider would be to take a recording of 30 mins of RTCM3 data, feed it through an RTCM3 to RINEX converter, and then post-process that against a NOAA-CORS in proximity. If the convergence numbers, or ability to fix ambiguities is poor, or the solution doesn’t match the reported antenna position (top of RINEX, or within 1005 msg) you might conclude it is not a usable data source.
Setting up your own base station, and infrastructure, allows you to control the quality and repeatability within your system.

beatified answered 8 months ago

Thanks again. I think I’ve got some promising results. Here is a screenshot
This isn’t a rigid mount it is currently at the top of a 35 foot Fiber glass antenna mast so that i can get it up in the air without needing access to the roof. Rigid Mount will come soon, after I create my enclosures.
Do I need to be worried about the NTRIP Errors or do they just happen from time to time?
And should I have 3D/DGNSS/Fixed as well or just Fixed as the other screenshot says?

Ardusimple Staff replied 8 months ago

(1) NTRIP is as reliable as your internet/wifi/ethernet connection is.
(2) just FIXED is ok.

Ardusimple Staff replied 8 months ago

(2) i mean both DGNSS/FIXED and FIXED is ok.

beatified replied 8 months ago

Great Thank you.