Product Compatibility

GPS/GNSS
1

I want to create a base-rover pair for maximizing relative precision. Reading through copious amounts of documentation, I have arrived at the following:

  1. Base station + Antenna
  2. Radios
  3. Rover + Antenna
  4. Position Data

For (1) and (2), I went with a pre-made Express Kit and LoRa serial radios due to previous familiarity, ease of use, and trust that the two are compatible.

For (3), I am leaning towards something with a mosiac-X5. You might be wondering why the mosaic when the chosen base station advertises a lower 0.010 m precision. The thought process is that once I gain experience with what I am doing, I can duplicate the rover and have one serve as a base.

For (4), I would love to cable (ideally USB or USB-C) out of (3) and into a tablet. SW Maps looks to be an offline option for data validation, but this bridge can be crossed later.

A few questions:

  1. What is involved with making (1) talk to (3)?

  2. What might I be looking for to demarcate the large number of options SparkFun advertises at the 6mm precision level given my stated goals?

  3. What do I need to attach to (3) to complete the rover - currently I am set on power protection and a black box logger. I would like the tablet to replace the need for a RaspberryPi or Arduino.

  4. Does the data come out as a .txt if I wanted to write display software for my application to avoid SW Maps?

  5. Am I mistaken in my belief that I can duplicate (3) to act as a rover later on?

I have a working understanding of how to manipulate data and numbers once they are in one spot. One of the biggest goals of this project is to expand skills on getting multiple devices to talk between each other. From everything I have read, I seem to be close. I am more than willing to read more documentation and tinker with a physical setup. I primarily am looking to know that “it’s possible to make these components talk to each other without much trouble” so I can get them in the mail and ready to help me learn.

2

Welcome @ajsobo !

Your #2, the Radios. Or other various forms of communication.

  1. What might I be looking for to demarcate the large number of options SparkFun advertises at the 6mm precision level given my stated goals?

Antenna Choice and the specific combination drives the selection when you start talking in mm’s. But this question can lead to pages of discussion :laughing:

  1. What do I need to attach to (3) to complete the rover - currently I am set on power protection and a black box logger. I would like the tablet to replace the need for a RaspberryPi or Arduino.

Once the link is established (you mentioned radio) between the Base and Rover, the Rover will ingest the RTCM observation data from the Base and calculate the RTK position. The Rover will typically deliver a NMEA stream. You can consume that NMEA Stream in many different ways (Bluetooth, WiFi, Serial, etc) on any device you want.

  1. Does the data come out as a .txt if I wanted to write display software for my application to avoid SW Maps?

Yes, you can parse the NMEA stream for the information. The RMC, GGA, GLL, etc, NMEA messages/sentences are examples.

  1. Am I mistaken in my belief that I can duplicate (3) to act as a rover later on?

I’m not sure I understand your question…but yes you can have many Rovers consuming the Observation Data from (1) nearby Base. FYI: Base and Rover can be the same (or different) Hardware, it’s just a different operating mode (Base vs Rover).

3

Thank you! Nothing is the response sounds conflicting with what I had convinced myself, and the answer to three particularly has me feeling better about my plan.

For the antenna, I have a GNSS Multi-Band L1/L2 Surveying Antenna - TNC (TOP106) and was planning on continuing with it until I find out it is limiting the operation. I have very friendly conditions (open field) in my application.

To clarify (5), which you already answered, I am currently buying a pre-made base and building a rover. Later, I might create a replica of my rover to act as the base. My question boiled down if there would be any foreseeable reason I could not swap the same components in a receive (i.e., rover) mode to a send (i.e., base) mode. My interpretation of your extra response is that switching a rover to a base is fairly straightforward.

If someone had a recommendation or strategy between choosing between the various options of 6 mm precision, I would appreciate the resource. In terms of application, I want to be able to press a button while stationary and collect a sample of positions to average together. I am convinced the simplicity of my goal leaves me with plenty of options. Perhaps the secondary goal is ease of transfer into a tablet. If I can get the numbers into python or excel, preferably via direct cabled link, I am set.

Unfortunately, in a good way, I just found 2-3 more possible products that look like they are capable of what I want.

1 Like
4

SW Maps will likely be the easiest way to get you started. Depending on where you live, see if you can find a CORS (NTRIP) service. You could use your existing Express as a Network Rover and begin collecting RTK positions in SW Maps (with an internet connection) immediately. SW Maps can export an Excel file with all the saved positions.

After you have a good understanding of your workflow, it wont be hard to transition to a Local Base instead of NTRIP. A local base (short baseline) will “tighten-up” the precisions.

That’s a good L1/L2 antenna. If you decide to upgrade in the future, look at L1/L2/L5 for both ends. But the TOP106 isn’t going to be a limiting factor at first.

Fun Fact: Averaging RTK positions in this context (5-6 mm precision goal) doesn’t behave like you would initially expect. Averaging solutions is how we reduce random errors. You will be using a defined workflow to reduce random errors (most are human errors in this context, excluding multi-path). The major contributor of error will be the Atmosphere bias (Ionospheric and Tropospheric delays), which is a “short-term” systematic error. Having a Local Base (with short baseline) helps to cancel out the atmospheric bias almost entirely. With a network Rover (NTRIP), you would need to repeat the work on a different day (with different atmospheric conditions) to cancel out. I think its still worth starting out with NTRIP if it’s available in your area.

As I mentioned, human errors can dominate when you are targeting mm precisions. It’s actually fairly hard to place a Rover antenna within mm’s of the same position from a previous day. Incase you haven’t already, be thinking about how you will physically mark your positions. We typically use a mag-nail, PK Nail, etc, that have a defined center point.

image
image161Ă—331 50.4 KB

Most people will use a 1-piece rod and a tripod for stability when mm precision is required. Getting the rod perfectly plumb is the critical part. The antenna will have a North Reference mark. Make sure to point that reference mark to the North everytime, because the actual electrical phase center is never the physical center of the antenna (could easily kill your 6mm goal). A proper workflow is how we reduce these random errors.

Product Specs : In GNSS receiver specifications, an example of 8mm + 1ppm describes the total error budget for a real-time kinematic (RTK) solution. It is divided into two distinct components: a constant “noise floor” and an additional distance-dependent factor.
Don’t put too much confidence in the first term when shopping/comparing receivers. Your workflow will impact the results far more than the Make/Model of the GNSS receiver, at this level. Operating a Local Base is how you reduce the 2’nd term.

To be fair, your 6mm precision goal is considered Control Surveying. Most of what you will read on the internet in this context will be Static Session, not RTK. Your 6mm goal is optimistic but absolutely achievable. Welcome to the Fun Stuff :smiling_face_with_sunglasses:

1 Like
5

Once you start chasing the last mm or two of precision, the Septentrio Mosaic lineup at SparkFun would be beneficial. But again, your antenna and workflow will impact the results more than the receiver.

1 Like
6

I am back after some preliminary testing. To clarify some earlier goals, I am convinced minimizing random error is my goal. Among several projects, the first boils down to creating an expensive tape measure. I would like to be able to claim that if the receiver does not move, X% of points fall within Y distance. With that uncertainty on both ends and utilizing other equations, I can solve my tape measure goal and accompany it with an uncertainty range. The shared suggestions for workflow have been appreciated, and I feel confident in my validation protocol.

Using NTRIP and initial tests with one device, my statistics indicate a 6mm precision for a pre-selected acceptable confidence interval and a pre-selected number of points to average. Conditions were nearly ideal.

In my intended application (i.e., expensive tape measure), I do not expect to always have NTRIP data, but I also do not need absolute precision. A local base station should be able to provide me the relative precision I desire for my local base-rover pair if I am understanding my source material correctly.

Regarding products with the mosaic chip and my initial testing, I do not seem to be ready to build my own setup from the ground up. That has pushed me down the following path, and I would be grateful if anyone could give confidence in the plan or stop me before I make a mistake.

  1. SparkFun RTK mosaic-X5 - SparkFun Electronics
    1. Comes relatively ready to use
    2. Has specs that appear to be compatible or better than goals
    3. States that it can be a base or rover which is important for other projects down the line
  2. I need to broadcast the RTCM a “short” range (roughly 300 meters)
    1. That led me to a discontinued serial option.
    2. That led me here: SparkFun LoRaSerial Kit - 915MHz (Enclosed) - SparkFun Electronics
  3. The schematics of (1) and (2) lead me to believe they can be connected, but the pin setup is not exactly what I expected. Will the radio (2) be possible to connect to (1)? If so, does the com/connection allow for either listening or talking and/or both at the same time?
  4. I believe the [black] antenna in (1) is for wifi/bluetooth and not serial, so I have been ignoring that for the time being. But this is also useful for future projects.

I suspect the answers are yes and yes.

7

I’ll let the SparkFun folks answer the specific hardware questions about the radios.

I want to offer you words of encouragement for your plans. I think you are on the exact right track. Your application/description is where a Base/Rover will shine. Having the same antenna/receiver at both ends removes any potential bias, for all satellite geometries.

I’m still not convinced that a lot of averaging will help you. The Base/Rover handles the atmospheric errors with double differencing. You’re now dealing with time-correlated errors such as multi-path and DOP. Averaging wont help with multi-path, nor the satellite geometry (in this context). I absolutely could be wrong and forgetting something important. But my gut tells me averaging might actually have the ability to hurt you. I know that sounds nuts.

Personally, with a 300m baseline and tight control of the fieldwork - I’d look at the RTK residuals over a few seconds and pull the trigger on a solution. Goofy things like not orienting both antennas properly (or at least the same) is what can bite you.

You’ve got me all excited and wanting to do some real baseline testing again…it’s been years. I’ll try not to clog-up your thread anymore than I already have, and hope others will chime in.
Thanks! :nerd_face:

8

I need to broadcast the RTCM a “short” range (roughly 300 meters)

LoRaSerial is a fine solution for 300 meters. However, I worry that the mosaic-X5 is capable of receiving so many satellites and constellations that the RTCM data rate may overwhelm the LoRa abilities. The 100mW radios are a little less range but are a plug-and-play solution, and should handle the larger data rate.

Will the radio (2) be possible to connect to (1)?

Yes, the LoRaSerial or other telemetry radio can be connected to the RTK mosaic-X5. Please see the I/O terminal docs, specifically the mosaic-X5 COM1 pins.

I believe the [black] antenna in (1) is for wifi/bluetooth and not serial, so I have been ignoring that for the time being. But this is also useful for future projects.

Yes, the small black antenna is for WiFi/Bluetooth.