GPGGA location decoding failure due to "invalid uncompressed location"

41 views
Skip to first unread message

Houzuo Guo (Howard Guo)

unread,
Nov 21, 2022, 1:32:42 AM11/21/22
to aprs.fi
Hey there.

There's a discrepancy between the GPGGA sentences transmitted over d-star by Kenwood TH D74 (2016) and ICOM IC-52 (2021) - the former transmits leading 0s in the latitude field whereas the latter does not use the leading 0s. Consequently, aprs.fi labels the latter as invalid due to "invalid compressed location". See attached screenshot for an example.

Are the leading 0s mandatory in GPGGA sentences? Or might it be a corner case aprs.fi overlooked?

Thanks!

Kind regards,
Howard EI9LC
nmea-gpgga-problem.png

Heikki Hannikainen

unread,
Nov 21, 2022, 2:09:59 AM11/21/22
to aprs.fi
Hi,

The packets at this point are not GPGGA formatted, they are APRS
formatted. Probably a DPRS gateway did the conversion to APRS format
somewhere over there (EI2PMD-C, I think).

The packets in the screen shot are APRS uncompressed packets, and to be
precise, the error shown in the screen shot is "invalid uncompressed
location".

In the APRS uncompressed format, the leading 0 is mandatory - the APRS
packet format has fixed field widths. See APRS101.PDF, "Latitude Format"
and "Longitude format", "Position coordinates", pages 23-24 (page
numbering), page 33-34 (actual PDF pages). The protocol is not great, but
this is how it is and it's difficult to change now.

http://www.aprs.org/doc/APRS101.PDF

If I made the aprs.fi decoder accept invalid position formats, you'd still
have all the *other* APRS decoder implementations rejecting these packets,
including a whole lot of radios and other embedded devices rarely getting
firmware upgrades. aprs.fi is just one of the tens of actively used
decoders. I'd guess the TH-D74 itself will not decode that APRS packet if
it receives it on an APRS frequency.

The DPRS gateway needs a fix for this. Or the TH-D74.

- Hessu

Houzuo Guo

unread,
Nov 21, 2022, 4:48:32 PM11/21/22
to apr...@googlegroups.com
Kiitos paljon Heikki!

--
You received this message because you are subscribed to a topic in the Google Groups "aprs.fi" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/aprsfi/1FQ-it5QkFs/unsubscribe.
To unsubscribe from this group and all its topics, send an email to aprsfi+un...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/aprsfi/alpine.DEB.2.21.2211210859450.4348%40jazz2.he.fi.

Houzuo Guo (Howard Guo)

unread,
Dec 14, 2022, 4:49:25 PM12/14/22
to aprs.fi
Just a brief follow up to this discussion - so at the start I noticed several quirks in the APRS transmissions (relayed by D-Star) coming from ICOM ID-52:

- When ID52 transmits location in the legacy NMEA mode it is missing the leading 0s in the encoded longitude, hence causing "invalid uncompressed location" error.
- When it transmits in the new D-PRS mode, the APRS network does not receive any position reports.

The digital gateway used during my transmissions is a pi-star (version 4.1.6), therefore I followed up with the maintainer of ircDDBGateway Jonathan G4KLX (The gateway is not sending D-PRS message to the APRS gateway software) and explained both of the quirks.

In my follow up comment there are 7 example transmissions with raw dumps of the embedded data frame, in case they might come in handy for software developers.

Wish y'all an early merry Christmas and a brilliant start to 2023!

Kind regards,
Howard, EI9LC
Reply all
Reply to author
Forward
0 new messages