Questions on GPS-RTK2 ZED-F9P suitability for application prior to ordering

GPS/GNSS Development Boards
1

I am looking to set up a rover/base pair for tracking the position of athletes during dynamic sports movements. I’ve read through most of the information linked on the product page and in some of these forum posts, and I have some lingering questions about suitability of this system for my application before I drop $$$ on parts. Apologies if these are dumb questions, but sometimes it is hard to evaluate whether a product will work for you when you don’t have it in your hands yet to try…

  1. Antennas on rovers: Sounds like you need a ground plate. Any thoughts on how to put a ground plate on a human doing dynamic sports movements? Are there other options that might work?

  2. Power supply and data storage on rovers: All the examples I’m seeing look like your rover needs to be plugged in via USB to both have power and to transmit data (despite having a backup battery). Can you run the whole thing wirelessly (ie from a battery) and store/log data on the board? Do you need a separate storage chip to do this? If this is possible, how would you communicate with the board to start/stop logging?

  3. Sample rates using real-time RTCM corrections vs applying after-the-fact: all the examples I’ve seen include streaming RTCM in real-time over bluetooth or serial. I don’t need my location data in real time, but I do need the sub-mm accuracy that comes with the RTCM corrections, and at the best sample rate I can achieve up to 100hz. Is it possible to synchronize clocks between the two boards (base and rover), store the data, and then apply the corrections from the base after-the-fact? In doing so, how much would I be able to increase my sample rate?

  4. If you are collecting RTCM data to apply the correction in post-processing (rather than in real-time), do you need a wifi connection at the time of data collection?

2

This may be pretty complicated to setup, especially on a human rather than something like a car or robot but I’ll try to answer your questions.

Any thoughts on how to put a ground plate on a human doing dynamic sports movements? Are there other options that might work?

I’m not sure how you intend on mounting your antenna and ground plate, but the antenna needs to be oriented directly up to the sky. You’re going to need a surface that maintains that orientation to keep connected to the GPS system. Anything could be used for mounting, if in a helmet, glue or Velcro would work. If on a persons body, some sort of strap might work but the ground plate and antenna must have a clear view of the sky.

All the examples I’m seeing look like your rover needs to be plugged in via USB to both have power and to transmit data (despite having a backup battery). Can you run the whole thing wirelessly (ie from a battery) and store/log data on the board? Do you need a separate storage chip to do this? If this is possible, how would you communicate with the board to start/stop logging?

You need power, doesn’t matter if it’s coming through from a computer via a USB cable or from a power supply battery that’s carried by the rover. The battery backup on the board doesn’t supply power to the GPS, it maintains the volatile memory in the GPS so that the GPS can get a faster lock after being powered down. (Without the memory being backed up, the GPS has to calculate where on the planet it is every time it’s powered. If it has somewhat of an idea where it is, getting a lock goes much faster.)

If you wanted to log/store data at the board, a serial logger like the Qwiic OpenLog or serial OpenLog could be used. You would need to program a microcontroller to do the logging though if you’re using the Qwiic OpenLog as it can not log I2C data without some help from a controller. If you wanted to communicate with the board (or send data from the board) you’d need some sort of radio on the rover to do that. You could use Bluetooth, WiFi or some other system as long as you have enough range.

Sample rates using real-time RTCM corrections vs applying after-the-fact: all the examples I’ve seen include streaming RTCM in real-time over bluetooth or serial. I don’t need my location data in real time, but I do need the sub-mm accuracy that comes with the RTCM corrections, and at the best sample rate I can achieve up to 100hz. Is it possible to synchronize clocks between the two boards (base and rover), store the data, and then apply the corrections from the base after-the-fact? In doing so, how much would I be able to increase my sample rate?

This is beyond my knowledge so I can’t say for sure, but it sounds mostly possible? I don’t think you’re going to be able to do 100Hz sampling though as the GPS can’t go any faster than 25Hz. You do need to be aware that you can get sub centimeter accuracy with our product, but not sub millimeter.

If you are collecting RTCM data to apply the correction in post-processing (rather than in real-time), do you need a wifi connection at the time of data collection?

I wouldn’t think so. If you’re doing post processing, you just need the data. Doesn’t matter if you have WiFi or not when you collect the data.

3

Thanks for the response. I have purchased a bunch of components (SD card reader, IMUs, remote receivers, etc) and getting everything set up is going much better than expected! A couple follow-up questions/comments:

the GPS can’t go any faster than 25Hz.

When I try the ‘Example7-OutputRate’ sketch, I can’t get it to go faster than about 3.5Hz. What’s the secret to getting 25hz? (25 is “good enough” for my application)

You do need to be aware that you can get sub centimeter accuracy with our product, but not sub millimeter.

I did mean sub-cm, not sub-mm – that was a typo!

If you’re doing post processing, you just need the data.

I am struggling to find information on how to apply the RTCM corrections in post-processing rather than real-time. What does the RTCM data actually look like? How (mathematically) is it applied to the rover gps data?

4

Are you letting the sketch run for a little while? the displayed speed in the sketch is just an average and will take a bit for enough data to pass to calculate. You should eventually see the value settle close to 10Hz.

I’m afraid I don’t have any information on how to apply the RTCM corrections in post-processing. If anyone has any insight on that, please respond to this thread.