Getting >20mm accuracy on RTK Torch with local base

Thank You. I have been using this tutorial and a couple others from Spark Fun as my touchstones. I have been carefully going thru them again and noticed a couple of steps I deviated from. Most notably the step where they say:

" Once the RTK product is configured, power it up with microSD inserted, and leave the unit in Rover mode. This will record all the data (NMEA, UBX, and RAWX) from the receiver to a *.ubx file. We do not yet know the location of the antenna so we stay in Rover mode to allow it to compile a large amount of satellite data. Only after we have confirmed its location should you enter Base mode."

I collected the data in ā€œBASE MODEā€. I shall collect new data in ā€œROVER MODEā€ as they suggest.

Also, I will try your suggestion of only using ā€œGPS satellite dataā€ for submission to OPUS. Unfortunately the weather prevents me from doing this today, but Iā€™ll be backā€¦

Run through that next section after you get the rover data and see how that goes :slight_smile:

That is the tutorial I have basing this whole exercise on. It is an excellent reference. As of today NOAA still has not added the FACET L-band to their list of calibrated antennas, so I will still be using SFEFACET as the antenna. Thanks for the response.

@BlackManOps , I appoligize for not answering your question to me in this thread.
The SparkFun tutorial is a great resource. Thatā€™s what Iā€™d recommend using for anyone starting out. And as you already know by now, the PPP I mentioned is indeed Precise Point Positioning (CSRS-PPP)

Sure enough, OPUS only uses GPS data. However, you might not want to limit yourself to OPUS by only logging GPS data with your device on a permanent mark.
The reason is OPUS and CSRS are 2 entirely different animals.

Oversimplified Explanation:

OPUS uses the Relative Positioning Technique, which is based on the CORS Network which are tied to the NGS Network Control. We assume all CORS station positions have been adjusted using this same network. Professional Surveyors and Engineers benefit from the inherent QA/QC, as we have ways to validate our data against the existing network of control points on the ground.
When you submit data to OPUS, your answer is calculated from this network of CORS.
As youā€™ve seen, OPUS has stringent data requirements to ensure their answer is the best available based on the data you submit and time it was submitted.

CSRS-PPP uses Precise Point Positioning Technique which doesnā€™t require/use data from external reference stations. It also supports the GLONASS constellation and a much higher logging rate.

I personally wouldnā€™t say that either choice is right or wrong, depending on the application. Itā€™s up to you, as long as you collect the additional constellation data in the field.

Thanks for the clarification and exposition. Another contributor helped get me over the finish line re. getting OPUS to accept my data file and process it. For everyone here and for posterity here is a brief outline of what worked in the end:

  1. Use Emlid Studio
  2. Use Options (the gear icon) to configure file processing as follows:
  3. Use ā€œLogs Durationā€ setting to trim off the first and last 5 minutes (or so) of data which is the equipment set-up and tear down periods since Sparkfun Facet starts logging data immediately and those beginning/end periods are bad data.
  4. Make sure the ā€œIntervalā€ option is set to ā€œ30 sā€ since thatā€™s what OPUS requires.
  5. Make sure the ā€œTime Roundingā€ option is checked. This aligns timestamps on perfect seconds boundaries. Again, a critical OPUS requirement.
  6. In the Satellites section check the GPS option ONLY. OPUS only processes GPS data.
  7. Donā€™t forget to set your antenna height to where you had it set up it in the field.

Also note that OPUS currently does not support the FACET L-Band antenna (NGS has been promising this 3-generations of product ago!). Iā€™ve been using the SFEFACET antenna type. Doing this will require a subsequent vertical datum adjustment when in Autocad working your COGO points.

Iā€™ve been informed there is a way to accomplish this in RTKLIB, but I havenā€™t worked out how to do the time rounding within RTKLIBā€¦yet

1 Like

I figured it out. In RTKCONV, Click Options. Near the bottom of page where it indicates ā€œReceiver Optionsā€ Enter ā€œ-TADJ=1.0ā€ without the quotes, but in caps (it IS case sensitive). This will put time stamps on exact 1.0 second boundaries and properly tweak whatever needs to be tweaked (i.e. not just re-label a time stamp) I got this from an article on rtklibexplorer here: Using U-blox receiver options in RTKLIB ā€“ rtklibexplorer

1 Like