"Meindert Sprang" wrote in news:47ce60c3$0
:

"Manlio" wrote in message
...
Presented by Raymarine in 2007, I am unable to find around a white
paper or some clear and complete presentation of this protocol.
Someone may give me an hint ?

As Seatalk has always a proprietary and thus closed protocol, I don't see
them publish any info on Seatalk NG....

Meindert

If you do a google search on this news group, you'll see that I provided a
link to a document that describes the protocol for normal SeaTalk. The
author was very responsive in some questions that I had.

While I can't tell you if there are differences between normal SeaTalk and
SeaTalk NG, I would bet that they are the same and that main differences
are in the physical and datalink layers. IMHO, RayMarine has too much
invested in lots of software to be making fundamental changes in the
protocol.

-- Geoff
www.GeoffSchultz.org

"Geoff Schultz" wrote in message
.. .
If you do a google search on this news group, you'll see that I provided a
link to a document that describes the protocol for normal SeaTalk. The
author was very responsive in some questions that I had.

Yep. http://www.thomasknauf.de/seatalk.htm

While I can't tell you if there are differences between normal SeaTalk and
SeaTalk NG, I would bet that they are the same and that main differences
are in the physical and datalink layers.

It is indeed very different, NG is based on NMEA2000.

IMHO, RayMarine has too much invested in lots of software to be making
fundamental changes in the protocol.

Well, since NMEA2000 is based on CAN, it is relatively easy to run NMEA2000
alongside some proprietary protocol on the same CAN bus and call is Seatalk
NG.

Meindert

"Meindert Sprang" wrote in
l.nl:

"Geoff Schultz" wrote in message
.. .
If you do a google search on this news group, you'll see that I
provided a link to a document that describes the protocol for normal
SeaTalk. The author was very responsive in some questions that I
had.

Yep. http://www.thomasknauf.de/seatalk.htm

While I can't tell you if there are differences between normal
SeaTalk and SeaTalk NG, I would bet that they are the same and that
main differences are in the physical and datalink layers.

It is indeed very different, NG is based on NMEA2000.
IMHO, RayMarine has too much invested in lots of software to be
making fundamental changes in the protocol.

Well, since NMEA2000 is based on CAN, it is relatively easy to run
NMEA2000 alongside some proprietary protocol on the same CAN bus and
call is Seatalk NG.

Meindert

Quoting from a RayMarine web site, "SeaTalk NG is an NMEA 2000 compatible
system, which can be interconnected to NMEA 2000 networks with an adapter
cable. It can also be interconnected to SeaTalk and SeaTalk2 networks for
backwards compatability with existing Raymarine installations."

Since this is backwards compatible with SeaTalk and SeaTalk2 networks, I'd
bet that the underlying protocol (application layer) is the same. While I
don't claim to know anything about the NMEA 2000 protocol, I would be
amazed if there wasn't some way to encapsulate a propritary message within
it without having to translate it to native NMEA 2000. This is done all of
the time in other networks.

So I'll go back to my guess that the application layer is the same but the
underlying layers (transport to physical) have been altered to NMEA 2000.

-- Geoff
www.GeoffSchultz.org

"Geoff Schultz" wrote in message
.. .
Quoting from a RayMarine web site, "SeaTalk NG is an NMEA 2000 compatible
system, which can be interconnected to NMEA 2000 networks with an adapter
cable. It can also be interconnected to SeaTalk and SeaTalk2 networks for
backwards compatability with existing Raymarine installations."

They only tell you half the truth... Seatalk and Seatalk NG/aka NMEA2000 can
be interconnected, but there is a small interface box/cable involved which
does the translation between both physical layers AND low level protocol.
The actual content of the datagrams might still be the same, which makes
sense.

NMEA2000 is invented by NMEA and is an application layer on top of CAN,
which is competely different from SeaTalk. Seatalk NG could be the same
type of datagrams used in Seatalk, but stuffed into CAN frames and sent on a
CAN network. There are a few similarities between Seatalk and CAN, like the
short datagrams, collision detection and the connectionless protocol, it is
merely publishing data on a network.

Meindert

So I'll go back to my guess that the application layer is the same but the
underlying layers (transport to physical) have been altered to NMEA 2000.

Incorrect. The bridge cable is passive. The NG cabling, it appears, just
has extra wiring to accomodate the older SeaTalk network. I'm guessing they
did this to allow bringing older SeaTalk gear into the network without
having to run a lot of additional cabling. That is, bring a line off an
older SeaTalk-based transducer and tee it into a new NG network. Leaves
them an opening to attaching existing gear (sensors, gauges and the like) to
the new network. So if you pulled a new NG 'backbone' network you would be
able to connect any older gear onto the cable without keeping the old
network.

It would not have made much economic sense to go with the NMEA2k protocol
stack and not take advantage of the economies of scale offered by
standards-based chipsets.

-Bill Kearney

"Bill Kearney" wrote in
:

So I'll go back to my guess that the application layer is the same
but the underlying layers (transport to physical) have been altered
to NMEA 2000.

Incorrect. The bridge cable is passive. The NG cabling, it appears,
just has extra wiring to accomodate the older SeaTalk network. I'm
guessing they did this to allow bringing older SeaTalk gear into the
network without having to run a lot of additional cabling. That is,
bring a line off an older SeaTalk-based transducer and tee it into a
new NG network. Leaves them an opening to attaching existing gear
(sensors, gauges and the like) to the new network. So if you pulled a
new NG 'backbone' network you would be able to connect any older gear
onto the cable without keeping the old network.

It would not have made much economic sense to go with the NMEA2k
protocol stack and not take advantage of the economies of scale
offered by standards-based chipsets.

-Bill Kearney

I'm lost. You say "incorrect" and then state that the cable is passive.
If it's passive, then the my statement is correct. Also, here's the link
to the document describing the SeaTalk protocol:

http://www.thomas-knauf.de/rap/seatalk2.htm

If you look at the datagram definitions, you'll see lots of parameters
which are highly specific to RayMarine equipment and have no mapping to any
NMEA sentances. I've never seen a document describing NMEA 2000, so I
can't comment on that. However, I can't imagine mapping the all of the
SeaTalk datagrams to NMEA 2000 and back. With tons of existing equipment
out there, you need to provide a method of backwards compatibility.

I just looked at the RayMarine ST70 documentation and you can bridge a
SeaTalk network onto a SeaTalkNG network through an ST70 instrument. There
is no way to directly connect a SeaTalk network to a SeaTalkNG network.

Also, you are NOT supposed to bridge a SeaTalk network to a SeaTalkNG
network with NMEA 2000 devices on it. I'm not sure what this implies.

-- Geoff
www.GeoffSchultz.org