We’ve been developing a product for a customer with your RTK boards. We’ve purchased 8 of them via different retail channels and as we are approaching to the end we couldn’t achieve accuracy level documented on your site or u-blox documents.
There are diagnostic files and deviation maps of two different RTK-SMA boards attached (ubx files mailed to support).
As you can see, not mm or 1 cm accuracy, location travels 5 to 10 cm.
We committed 2cm to our customer.
Our question is, there is a problem with cable (5&10m RG-58) or antenna (TOPGNSS-106) or NTRIP service or this is the actual capacity of the chip in real life and below cm accuracy is, -in lab conditions etc.- marketing gimmick.
If we can solve the problem, we’ll start the production phase.
While you are using the bare RTK-SMA board, do consider your reference frame and accuracy of the monument you are basing off of. This tutorial may be helpful: https://docs.sparkfun.com/SparkFun_RTK_ … ification/
Above is plots captured without (left) and with (right) RTCM corrections. Note how the points very while your points look like they are on a grid. Yes the scales are different, but it looks like your data is being truncated.