Very long range test of the RS2 LoRa RTK link


(Luke Wijnberg) #1

The new RS2 GPS has a built in LoRa radio for RTK corrections. The corrections are sent in RTCM format at 0.1W and has a claimed range of up to 8km. A group of us beta testers came up with the idea of a competition between us to test the range of these units. Running the latest firmware ( v2.20.2), we planned a few locations, designed some RF propagation plots, filled up the tank with diesel (the car, not the RS2) and headed out.

The first rough test was done locally. I wanted to see what the penetration though features like buildings, mountains, etc was like. The base was erected on a standard tripod and tribrach. Settings used where as follows:


I decided to reduce the amount of data being transmitted to the absolute minimum so the radio wouldn’t missing any stray messages, compromising a fix.

What we initiall found was phenomenal data transfer at 6-8km away. We had no issues with communications behind vegetation, buildings, etc. What blew my mind was a few fixes we had BEHIND a mountain!



Note that the mountain is over 1km of solid granite (no tunnels!) but the signal must have bounced off the valley walls behind me. At this point I must admit that I checked to make sure that I didn’t by mistake have a NTRIP connection running!

Our next test was to check the range. As mentioned, Emlid states that the radio is only good for 8km. But we were sure we could extend that and make full use of the dual frequency RTK baseline. We chose a spot in the KwaZulu-Natal midlands in South Africa at a restaurant on top of a fairly predominate hill. The plan was to test the range at 15km, 25km, then a long push to 35km.


RF propagation plot

Base location. Note the stock equipment setup

The first 2 fixes where fairly uneventful at 15km/25km and wasn’t even recorded as the base was clearly visible.

However, as we began climbing up the pass into Highmoor Reserve, things got interesting. With the RS2 mounted on a pole and out of the car window, we found data was streaming straight away from 35km. We found a pretty place to stop and recorded this:



Since the radio signal was so strong, we decided to push on up the pass to the reserve office. I was not planning on doing this as the RF plot didn’t show much signal up there. What a surprise when we arrived at the top and signal was just as strong as ever! 39.8km away!





A fix took about 6 minutes to achieve which is normal considering that only GPS satellites where being used. We could have continued further but we sadly ran out of road access and time. This is an amazing output for such a small device and shows good design, radio (and antenna) tuning.

Despite my initial hesitation about the LoRa communication design in the RS2, this unit is a vast improvement over the previous generation in the RS. Using the LoRa’s low power design means that a full days RTK work (and well beyond) is feasible. But don’t be fooled by that. What we have seen over the past few weeks of real life testing is that it is a perfect sound and powerful enough for any job site.

Future testing will be to include other constellations and to find a test area that exceeds 40km to find the units true line of site communication limit.


(Timd1971) #2

@Luke_Wijnberg

Wow. That is just awesome. Good work and thanks Luke! ; )


#3

That’s awesome! I’m hoping my RS2 units ship this week and I will be doing similar testing here in the eastern counties of the UK.

With that sort of range, I’d also really urge Emlid to consider releasing a compatible “standalone” LoRA radio with serial output.

Such a radio would be a killer app to integrate with existing non-Reach rovers, to be able to stream in RTCM corrections, for the Precision Ag (autonomous steering) and survey markets.

It would make things like this ad last week from a big Trimble reseller look like a complete overpriced anachronism…


(Timd1971) #4

Quite amazing what shams people and businesses fall for.


(Barry Bolton) #5

Good work fellow SAFFAS :slight_smile:


(David Hofer) #6

Good work Emlid and Luke. I thought you might get that LORA signal that far because your up that high and just about line of sight. It would be interesting how far you can get at the same elevation as the base ?


pinned globally #7

#9

Not sure @Luke_Wijnberg going to like this… :yum:





That is 52km baseline and fix close to a cell tower… Nuts!!
And no, i didnt use ntrip and hacked the cell tower :wink:


(Luke Wijnberg) #10

Well done, Tore!


(Luis Galina) #11

Hi, Tore! Have you checked with known points to verify that those fixes are correct? What constellations of satellites do I use?


#12

No, this is pure LoRa tests.
Base forwarded gps and glonass at 1hz/9Kb sec and 20dB.
Rover used all at 5hz


(Luis Galina) #13

what you could do is put a stake and then go back to stake out
I would like to do tests but I still do not receive the RS2


(Timd1971) #14

@TB_RTK Yes, maybe speed-walk, skateboard or roller-skate this time? HAHAH! :rofl:


#15

As always, very nice work @Luke_Wijnberg. I have questions from above figures. I like your rover tripod mount, could you share some info on that? That radio propagation plot is cool, what used to model?


#16

First 40km, and now 50km @ 9kb/s bandwidth? @Luke_Wijnberg and @TB_RTK you guys are really pushing the (previous) boundaries! Great Job, both of you :+1:


(l3technologycambodia) #17

Great tests and results. It would be interesting to test RS2 in normal use with Base and Rover are at similar elevation and survey in relative sub-urban area where there are buildings and trees. My RS+ could not go over 1km-2km in surveying this environment.


(Luke Wijnberg) #18


I am running some data collection for a company that does post processing software. Today I collected some points in my neighborhood. The area is quite dense with trees, buildings. The profile above shows the big hill in the way and a bunch of radio antennas on the top of it. Most of our cellphone masts are in the 900mHz range so fairly close to the 868mHZ I was running. Also there are plenty local WiFi routers on the way. I had no issues with comms in my entire neighborhood, which is fairly amazing. I am starting to wonder if that is not a typo in the RS2 specification page and that its a 1W radio, not 0.1W!


(Luke Wijnberg) #19

The rover is simply a carbon fibre survey pole with a tripod clamp. The legs are telescopic and the clamp a rubberised spring loaded one.
The base is a standard survey wooden tripod and tribrach. The tribrach has an adapter on it to 5/5th male. A 30cm fibreglass extension pole lifts it up so the antenna for the LoRa is free.

My aircraft engineer luckily has all the tools to make this stuff for me in his workshop!


(Luke Wijnberg) #20

Ja, @TB_RTK, why don’t you go back and repeat this? You didn’t do a very good job of it the first time around!! Im sure a full day of driving back and forth is exactly what you feel like doing!


(Luis Galina) #21

@Luke_Wijnberg Hello Luke! because you only use Gps and Glonass is a condition for double frequency?