Comparison between Emlid Reach RS+, Leica, Altus and Trimble receivers

Today I had a wonderful opportunity to compare different GNSS receivers with Emlid Reach RS+. We the colleagues from we were measuring 3 points in the city environment with following antennas:

  • Leica GS18
  • Leica GPS900
  • Altus NR3
  • Altus NR2
  • Trimble R4
  • Emlid Reach RS+

Measuring setup was as following: RTK with LatPOS network, 10 sec measurement of each point in FIX, handheld pole (2 m).

The results were really good. Almost all of the receivers showed 2 cm accuracy (standard for cadaster measurement in Latvia). Here is the list of deviations from the measurements of the same points, which were found in the local geodetic network.

Point 9058:

Point is in a small well, located very close to the building. With Emlid it was really hard to get FIX here, but after a small walk initialization was found and everything went well.

Results of the measurements (deviations in meters):

					dx		dy		dz	
Leica GPS900	   -0,033	-0,018	0,006
Emlid Reach RS+		0,016	-0,002	0,006
Altus NR3			-0,014	0,006	-0,006
Leica GS18			0,001	-0,007	0,007
Trimble R4			could not acquire FIX					
Altus NR2			-0,001	-0,013	0,013

Point 5222:

Point is a geodetic sign. Quite good observation of the sky. FIX status was acquired in 1.5-2 minutes. All other receivers acquired FIX in ~15-20 seconds. I think, it is because of L1+L2 receivers.


Leica GPS900        0,003	-0,004	-0,038
Emlid Reach RS+     0,003	0		-0,038
Altus NR3           0,003	-0,001	-0,051
Leica GS18			-0,004	0,004	-0,031
Trimble R4			-0,003	0,001	-0,042
Altus NR2			-0,006	0,005	-0,029

Point 5198:

FIX status was acquired in 20 seconds. I had to experiment with different configurations (Fix-and-hold and Continuous), as well as with different networks in LatPOS system (NETW-iMAX and Virtual RS).

Leica GPS900 	    0,005	-0,001	-0,038
Emlid Reach RS+		0,005	-0,009	-0,038
Altus NR3			0,013	-0,009	-0,044
Leica GS18			0,008	-0,008	-0,051
Trimble R4			-0,003	-0,002	0,015
Altus NR2			0,01    -0,006	-0,046

Overall, I am more than happy with the results. I have to admit, that local NTRIP base station was very near to the receivers, so that’s why everyone have good centimetre level accuracy. I will need to conduct the same experiment, but with the points, that are far from the local NTRIP stations.


You don’t happen to have any Topcon gear do you. I would be interested in seeing that comparison on a network. I am in Texas, US and we only have one network in our area that is not very good for our use-case so I always run a base/rover scenario. I have done my own testing up against the Topcon Hyper V and it’s not even close. Accuracy isn’t bad it just takes 1-2 minutes to get that accuracy with the Reach RS+.


I can find some Topcon gear. We have old Topcon receivers. I don’t have too much experience with them though. If you can help me with the setup, then I can make an experiment here.

1 Like

Nice. What models? Send your questions on! Screenshots of software issues would be great. Thank you!


Would be nice to see a kinetic rover test with PDOP values with the dual- and triple-frequency heads.

What data would you like to see? I have a Topcon Hiper receiver, the Reach RS+ and my phone stress testing right now and am running a value comparison so maybe I already have the data you are looking for.

I can tell already that with all the variables between the 3 devices I am testing, after about 20 minutes their coordinate values are within about a two hundredths of an arc second of each other, or about 1.8ft and 0.55m. This is not bad without corrections. Top disclaimer is that I am in my office at home so it is also a test of each device’s capability to achieve a lock on satellites in adverse conditions. I do not know how the thresholds of each compare, but Topcon is very picky on what it uses. It is using 10 sats whereas the Reach is currently at 20. The phone is at 19, so I assume that it is more comparable to the Reach’s algorithms.

The stress test is that after 15 minutes I have the Topcon and Reach sending corrections at 5hz continuously. I obviously can’t do this on the phone so I have it set to post-process continuously.

Does Reachview have a skyplot?

Nop, it was removed some versions ago.

Thanks @wizprod.

Michael, what I mean is comparing it to a dual- or triple-frequency setup while walking close to south-obstructing walls would be more valuable information than ultimate accuracy and good points since L1 can give you the same mm accuracy in ideal locations, however it takes more work to get a fixed location with obstructions and less-than-ideal location. The PDOP values would be the best indicator of multipath, obscured satellites, refraction, etc.

What would be interesting is to log kinematics while passing by a wall or trees, and see when RS+ starts to lose PDOP and how long the others keep going. It would be a test that reveals the shortcomings of L1-only so we know what to expect. Sort of a test to failure scenario.

Gotcha. Let me see what I can come up with. Next week should be some good field time. I’m already 6 drone projects behind thanks to the unusual rain we have had. That means allot of flying to do. :wink:

I think, that it is obvious, that L1 will lose this challenge so bad. When I am measuring in the city, I almost always have very bad results, in terms of time-to-fix. Sometimes I can’t get fix at all. Pdop values are around 6-11. In my original post, settings were quite comfortable for GPS in the city. Buildings were not so tall and streets were straight. So, some satellites were visible.

1 Like

This topic was automatically closed 100 days after the last reply. New replies are no longer allowed.