QK-A031 Seatalk connection

Quark-elec AIS receiver and NMEA MUX products questions

QK-A031 Seatalk connection

Postby d.holburn » Sun Aug 20, 2017 8:31 pm

Dear Quark-Elec team,

Following sea trials of my newly purchased QK-A031 with Raymarine C70 and ST60 series instruments, I need to ask the nature of the electronic interface to the Seatalk data line on the QK-A031.

Whenever I connect the link to the Seatalk data line (yellow as per Raymarine convention) to the D terminal of the QK-A031, the SOG display and the NM display of my existing instrumentation become corrupted. SOG freezes at a fixed value unrelated to the actual speed, and NM is set to a low value. If I remove the yellow data wire (no other changes), correct readings are restored - but of course there is no Seatalk data feed to the QK-A031. As you can guess, the amount of diagnostic hardware I can transport on board during sea trials is limited, but I am confident about these observations and have repeated them a number of times.

As you can see from other submissions to the forum, I have succeeded in seeing the WiFi output from multiplexed Seatalk and NMEA signals, in Adhoc and in Station mode. I have made no connection to 'NMEA out' on the QK-A031, and none of my WiFi equipment is connected back to the instruments. Therefore, I am forced to the conclusion that the problem arises from the connection of the Seatalk data line to the D terminal of the QK-A031, since removing that connection instantly restores my instruments to correct operation.

I suspect the QK-A031 is either
(I) overloading the source (I doubt this), or possibly
(ii) it is putting additional data packets on the Seatalk data line.

The documentation text gives the impression that the QK-A031 is only a 'listener' for the purposes of Seatalk, but the block diagram in the Application Note:


gives the opposite view - as it shows (on both p2 and p4) the 'Seatalk In' signal buffered and made available as 'Seatalk Out' - as we know, Seatalk is a multiple-talker, multiple-listener broadcast 1 wire serial bus with devices connected in daisy-chain fashion. It's just not clear whether the QK product is a talker as well as a listener. I can see no separate Seatalk Out terminals.

If the QK-A031 Seatalk connection is genuinely listen-only, then I would have expected your reply to Paul Clark a few weeks ago:
Re: Can A031 convert NMEA message to Seatalk port
by admin » Thu Jul 20, 2017 2:18 pm
to have said that 'converting NMEA message to Seatalk port' would be impossible. However, that was not your reply!

I would be grateful for some confirmation as to the true situation. If the QK-A031 does talk to the Seatalk bus, is there is a software/GUI option somehow to turn off talking to the bus? Alternatively, can you advise on an electronic solution that would allow me to make the connection truly uni-directional - I believe a simple buffer should do this.

I should mention that I am a professional electronics engineer, so am ready to experiment. I have read Thomas Knauf's web pages regarding the reverse engineering of Seatalk with interest. Life is just too short for me to develop interesting pieces of kit like the QK-A031, so I am grateful for the opportunity to work with it.

Your advice, either on the forum, or by PM, gratefully received.

All best,

David H
Posts: 4
Joined: Sat Aug 19, 2017 9:51 pm

Re: QK-A031 Seatalk connection

Postby d.holburn » Thu Aug 31, 2017 2:03 pm

I've PM'd admin with an update on this, but of course it could be of more general interest. Here is a summary of the email, with some additional results.

This is a further report on my sea trials of QK-A031 with the following inputs:
NMEA 38400 b from Raymarine C70 chart plotter NMEA out, all NMEA sentences enabled for the moment
Seatalk 4800 b from local seatalk bus with wind, depth and speed instruments.

I am taking no wired outputs from the QK-A031, purely WiFi, in Station mode. None of my WiFi readers (iPhone, iPad) is connected back to the NMEA or Seatalk buses.

I reported some observations in the forum: viewtopic.php?f=3&t=9939
With all three Seatalk lines connected, I noted erroneous readings in my existing instruments. With the yellow data line disconnected at the QK-A031, these erroneous readings disappeared (but of course there was no Seatalk input to the QK-A031). I conjectured that:
1) QK-A031 input might be imposing too much load on the Seatalk bus - seems unlikely - my system is quite small (i.e. few Seatalk devices connected) and I would imagine QK-A031 has been designed not to impose significant loads on this bus;
2) QK-A031 might be writing messages to the bus. The Application note and Forum commentary are not totally clear on this point and may be contradictory.

I decided to try a suitably designed buffer in the Seatalk D line at the point of connection to the QK-A031. Circuit is attached. With this design, even if the QK-A031 is trying to write to the Seatalk bus, there is no possibility that the data can get back to the existing instruments. My observations are that fitting this cured the erroneous readings on the existing instruments while allowing the QK-A031 to operate.

Seatalk non-inverting unidirectional buffer
QK-A031-seatalk-buffer.jpg (102.32 KiB) Viewed 46 times

I would be grateful to know whether or not the QK-A031 has the capability to place data on the wired Seatalk bus. My observations seem to suggest it may be doing this and corrupting my instrumental displays, but is being prevented by insertion of the buffer. However, if I accept the statement that the QK-A031 only listens to the Seatalk bus, no buffer should be required. The App Note on the QK-A031 shows that the QK-A031 Seatalk input is buffered and appears as Seatalk Out - but there are no separate marked terminals on the unit with that designation. I wonder where Seatalk Out is connected? If it turns out that the QK is able to write to the Seatalk bus, can this be controlled in software?

Here is some data extracted from the WiFi stream, which appears to contain NMEA as well as translated Seatalk.

1 - Does the QK-A031 place a standard prefix before the translated sentences from Seatalk input? If so, which of these is it?
2 - Please can you explain what are the $STALK sentences? They do not seem to map to any meaningful Seatalk transmissions based on the table in your App note.

Listening to:""
Port no: "2000" (TCP)


Many thanks

David Holburn
Posts: 4
Joined: Sat Aug 19, 2017 9:51 pm

Return to Marine Electronics(AIS/NMEA MUX)

Who is online

Users browsing this forum: No registered users and 1 guest