RP6conrad / ESP-GPS-Logger

GPS logger for Speedsurfing, based on ESP32 & Ublox M8N hardware.
MIT License
28 stars 16 forks source link

Incorrect checksums #3

Closed prichterich closed 1 year ago

prichterich commented 2 years ago

The current code writes NAV-PVT sentences with incorrect checksums for points that are filtered (the speed is set to 0 because of fewer than 5 satellites, sAcc > 1m/s, or speed >40 m/s). These sentences are then rejected by readers that check the checksum, which includes u-center and GPS Speedreader.

The cause is that the filtering is done before writing the sentences to log files. The problem could very simply be fixed by first writing the data to the log file, and then applying the filters. The filters make sense for internal speed calculations, but the log files should contain the original information from the GPS chip, not data that have been altered in any way by the firmware.

RP6conrad commented 2 years ago

Yes, this has to be corrected. I’ll change it with the next update.

Greetings Jan.

Verzonden vanuit Mail voor Windows

Van: prichterich Verzonden: zaterdag 14 mei 2022 15:51 Aan: RP6conrad/ESP-GPS-Logger CC: Subscribed Onderwerp: [RP6conrad/ESP-GPS-Logger] Incorrect checksums (Issue #3)

The current code writes NAV-PVT sentences with incorrect checksums for points that are filtered (the speed is set to 0 because of fewer than 5 satellites, sAcc > 1m/s, or speed >40 m/s). These sentences are then rejected by readers that check the checksum, which includes u-center and GPS Speedreader. The cause is that the filtering is done before writing the sentences to log files. The problem could very simply be fixed by first writing the data to the log file, and then applying the filters. The filters make sense for internal speed calculations, but the log files should contain the original information from the GPS chip, not data that have been altered in any way by the firmware. — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you are subscribed to this thread.Message ID: @.***>