Expected RTK accuracy; NTRIP; 40km baseline

I have the same question. If the distance from the NTRIP base station is 40km, is it possible to still get centimetre accuracy ?

In reference to:

mod edit: split into new topic

Hi @mirkozhou920605,

If you want to survey with Reach RS+ in RTK, the distance to NTRIP station shouldn’t exceed 10 km. Otherwise, you may experience accuracy reducing. Note, that accuracy is decreased by 1mm each km of baseline.

For PPK workflow, the distance in 40 km should be fine.

1 Like

Thank you for your help.
The project I will be doing doesn’t need super accurate. If it can achieve half meter accuracy, it would be acceptable. So is it possible to exceed 40km baseline by sacrificing some accuracy but still achieve half meter accuracy ?

@mirkozhou920605 You didn’t mention which device you wanted to use. @tatiana.andreeva gave an example using Reach RS+, so I will assume that is the device you plan to use.

If we look at the published spec ( https://emlid.com/reachrs/#nav-specs ), then the rule for horizontal accuracy of a stationary rover is:
5 mm + 1 ppm

So at 40km:
= 5 mm + 40 km * 1000000 mm/km / 1000000 parts
= 5 mm + 40 mm
= 45 mm

Remember that you will be operating outside of the recommended limits if you do this. See FAQ “What is the maximum distance between base and rover units?” at: https://emlid.com/reachrs/#nav-faq

If you want to be operating inside the recommended limits, then Reach RS2 is a better device for RTK at a distance of 40km from the NTRIP station. At that distance it has the same expected RTK accuracy that Reach RS+ would have with the PPK workflow mentioned above.


Thank you very much. What about Reach M+ ?
I intend to use the Reach M+ as the moving rover, receiving the RTK corrections from NTRIP base station around 40km away.

The RS+ and M+ share the same internal hardware. It is only the antenna that is different. They still share the same spec though ( https://emlid.com/reach/#nav-specs ), so the answer is the same.

IF you want real-time (RTK) positioning with an NTRIP base station at 40km, then you should probably get the Reach RS2.

You could get away with the cheaper M+, but you would be operating outside of the recommended limits and the RTK experience will not be as good as it would with Reach RS2. As @tatiana.andreeva already mentioned, if you used the PPK workflow instead of RTK, then you could confidently stay with the Reach M+.

1 Like

I suppose in conclusion one could say that you could try Reach M+ for the use-case you described and it may be acceptable, but we can’t recommend it. It would be your risk.

Some questions to ask yourself are: Does it matter if the fix drops to a coarser float solution? Does it matter if it takes a long time to achieve the fixed solution or to regain it if lost?

If you have tolerance for this kind of thing, then it might be worth the risk to try out the Reach M+ for your use-case.

The reason why I use Reach M+ is because it can be mounted on the Pixhawk vehicle. Is there any GPS receiver available that has similar size with Reach M+ but use the same hardware of Reach RS2, which has the multi-band GPS chip ?

:grin: You describe the hardware that a lot of people here want, but it is not available here (yet?). The RS2 was just released, so if there is an “M2” to be produced, I guess it would take some time before it would be announced, let alone be available to purchase.

We don’t want to send you away to look for competing products though, so maybe you will have the patience to wait (or are willing to take a risk and try the M+) :slight_smile:

1 Like

If you are doing standard drone work, then do you really need RTK workflow? Most people don’t and are fine with PPK workflow (and the Reach M+).

The vehicle I’m trying to build needs to be able to take a series of measurements at a grid of points. So it needs to arrive every measurement location with a fairly accuracy (within 0.5m error) :sweat_smile:

RE my question:

Is the accurate measurement location needed for real-time processing (RTK)? Does the job depend on it, or can the accurate measurement location be calculated later (PPK)?

Yeah, it can be calculated later . But then does it mean the vehicle needs to run on the field twice ? First time is for collecting the raw data and then PPK ; second time for doing measurement on the calculated locations?

(I deleted my previous post after re-reading your replies.)

So what you have is:

  1. An autonomous vehicle with a Pixhawk controller.
  2. Access to an NTRIP station that is 40km away.
  3. Several coordinates that the vehicle needs to travel to where it will take measurements.
  4. A need for the vehicle to arrive within +/- 0.5m of those coordinates.

What you need is clarification on whether Reach M+ is a good choice for this use-case.

Additionally: It is OK for you to learn about the precise arrival coordinates afterward by PPK workflow, but that doesn’t change the fact that the vehicle needs to arrive within +/-0.5m in real-time.

Please correct me if there are any errors.

(I’ll wait for @tatiana.andreeva’s opinion)

Hey there,

With the baseline in 40 km, you may experience difficulties with achieving the fix solution. So the unit may just remain float with meter-level accuracy. As @bide said, it’s hard to say certainly how it will behave with such a baseline.

In your case, I’d recommend using the second Reach device as a local base. You can establish its position with PPK and then broadcast corrections to your Reach rover over LoRa. Note, that you need an external LoRa module for Reach M+ for this.


Thank you for your help. Then can I use another Reach M+ as the local base instead of using the expensive Reach RS ? Also I’m wondering is Reach M2 gonna be available soon ?

1 Like

Yes you can . To set it up with Lora your only a few hundred dollars different.

Is it possible for the local base receiver to transmit the RTK correction to the Mission planner through physical connection, which then transmit the signal to the Pixhawk and GPS receiver through telecommunication ? Because right now I have a pair of telecommunication device already, so I’m wondering do I have to get another one ?

Hi @mirkozhou920605,

You can connect the base to Mission Planner over the USB cable/external radio.

Choose Serial option in ReachView position output and COM port in MP while configuring base to inject corrections to MP.