No RTK Fix

Q&A forumCategory: QuestionsNo RTK Fix
beatified asked 3 weeks ago

I have my base on a stable 25 foot pole but am never getting any better on my rover than a 3d fix. My rover blinks between 2 and 3 times on the Xbee to GPS for every one blink of GPS Fix. The Base GPS is blinking GPS to Xbee. So I dont know what is going on with it. None the less the rover RTK Led never turns off. My Base Antenna is about 25 feet high or about 5 feet above the roof. There is a hill near by but there is not much I can do about that at the moment. I can raise the fiberglass pole higher if I want but I kept it low to try to make it as stable as possible, It gets flimsy at higher heights but other wise its possible to get it to almost 40 feet. Although this setup is for testing only right now.
Please help me get an RTK fix. I was going to try to get an RTK fix from ntrip client but I cant figure out how to do that. And that was just as a trouble shooting step

Ardusimple Staff replied 3 weeks ago

Hi beatified,

let’s find together where is the problem coming from.
1. Which kit do you have?
2. Did you change any configuration? If yes, you can return to the original files: When this is confirmed, move to the next step.
3. Is the base in TIME mode? You can check this by connecting to the base, enabling the message ubx-nav-pvt and check if you see TIME mode in the FIX Type. If yes, let’s go to the next step.
4. Now connect to the rover. Go and enable ubx-rxm-rtcm. Which messages do you see? And do you see all of them increasing once per second?

beatified replied 3 weeks ago

So I managed to get a Float fix by setting a location manually. Is there a way to improve the fix to dgnss fixed?

clive1 replied 3 weeks ago

It is only going to get better with clean usable data. Noisy measurement data, starting with the base is going to have a systemic impact.

Doesn’t need to be 5ft above roof, simple needs to break the roof line, going to be excluding everything below 5 to 10 degrees anyway. Use the uCenter Deviation Map view over 24 hours to understand the sky view/obstructions locally.
The mount should be solid, you should have a ground plane under the puck type antennas, I’d suggest at least 6-7″ diameter as a minimum, or get a similar diameter pro/survey antenna. If you can make 14-15″ work, do that.
The Survey-In isn’t particularly good, nor repeatable, better to post-process raw measurements against a government reference, CORS. If you can feed the Base with NTRIP data from a local source, use that to get an accurate fix.
The radio antenna should be as high as possible to meet coverage expectations.
Make sure you have a reliable data stream at the rover, use UBX-RXM-RTCM to check.
Have a ground plane on rover antenna, watch for obstructions, and keep away from the human body.

beatified replied 3 weeks ago

Thanks for the extra info clive1. When your referring to clean data are you referring to the xbee radio (I have the LR kit BTW)? Also I’m guessing altitude won’t hurt but at a certain point won’t help either. There are a lot of trees in my area and a small hill that is close by.

As for NTRIP I tried using one who’s details I found and are as follows…
Site Name
Sonora High Scho
CA Zone
California Zone 6
NAD83 Lat
33° 55′ 38.4086″
NAD83 Long
-117° 55′ 43.019031″
NAD83 Ellipsoid HT
Geoid 12B (m)
CA Orthometric HT (m)
NAD83 Velocity N (mm/yr)
NAD83 Velocity E (mm/yr)
NAD83 Velocity U (mm/yr)

I entered the up and port and was able to find SNHS which is close to me but when actually trying to utilize that data it would report an error as soon as I click ok. Am I missing something?

Also how do I check things like USB-RXM-RTCM?

Thanks for helping a newb.

clive1 replied 3 weeks ago

I’m primarily talking measurement noise, the data transmitted in the RTCM3 packets used here are measurements, not “corrections”
You’re trying solve multiple geometry equations and the ability to fit that well depends on the amount of motion (noise) in the data from the Base, it needs to resolve sub-cm for the Fixed RTK to be achieved.

Rigidity of the antenna mounting, and immunity from indirect signals, plays significantly to the measurement noise. The correlators can resolve the carrier signal to a fraction of a mm, so it moving around or oscillating is going to cause grief. I think there are going to be enough satellites in view from all constellations that a few trees will not be a big issue. Having a lot of signals blocked on one side (except North in your location) is unhelpful. Generally people with have issues with antennas placed on window sills as this causes significant masking.

Now if the radio link is failing, that’s going to give another set of issues, mainly in sustaining the Fixed RTK, and recovering signals into the solution is lock is lost on specific satellites.

To get at the UBX-RXM-RTCM reporting go into the Message View in uCenter and drill down to the packet form, and enable it. You should get running totals, and confirmation it is receiving which messages and their integrity over the radio link.