Multiple GPS inputs onto Seatalk - better or worse?

R

Ric

Well-Known Member
Joined
8 Dec 2003
Messages
1,723
Visit site
Currently I have an ancient Raymarine RC425 chartplotter with external GPS antenna which inputs GPS NMEA into a Seatalk network via an E85001 NMEA-Seatalk converter.

I have just bought a Raymarine E7D chartplotter with inbuilt GPS antenna which I will connect to my Seatalk network. I intend to leave the RC425 wired into Seatalk as a backup GPS, with an independent switch so I can select it if the E7D goes pop.

I have also bought a Raymarine AIS650 transceiver which comes with its own GPS antenna and Seatalk connection.

Most of the time the AIS650 and E7D will be on - therefore both putting GPS data into the Seatalk network. Will this cause a conflict?

What happens if I switch on the RC425 too?
 
It will cause problems having two separate units inputting GPS location sentences into the network. The plotters should have an option to switch off the internal GPS units and listen for external inputs. you can switch on the internal one later if there is a fault, alternatively use the internal one and output that onto the network instead.

Chris
 
Firstly you might not have to worry about the AIS650. I have the earlier 500 and it won't put out GPS data. It's only for consumption by the AIS unit. IMHO that's a rather wrong interpretation of the specs but there you go. The 650 manual makes no mention of putting out GPS data so I assume the same applies.

Also...sorry if I'm stating the blindingly obvious but I believe the E7D and AIS650 have no seatalk 1 connectivity (SeatalkNG yes, Seatalk 1, no)

One solution might be to buy a multiplexer which will do failover of data sources and seatalk conversion. Plug the nmea-0183 outputs from both plotters and the seatalk bus into the multiplexer and program it to fail to the RC425 if no GPS data is seen from the E7D. I've no personal experience with these devices but the ones below allegedly do that stuff:

http://www.shipmodul.com/en/index.html
[EDIT: Posted that too soon. Just removed one of the two multiplexer references as it looks like that's seatalk to nmea only, not the other way round. Could be wrong...Don't take my word for it!]

They're pricey though and you've already got the E85001, so the simple solution might be a physical switch to switch between the nmea-0183 tx lines from each plotter to select what goes into the E85001 and thence to the seatalk bus
 
Last edited:
Why/

The data comes in a stream of bits each identified by its position in the stream usually starting with a timing pulse. While the baud rate will mean that the stream comes at a similar frequency the individual timing will be quite random. So a pulse meaning S could be confused with a pulse from another device being part of a stream giving time or speed. Indeed the pulses could all be combined to give a continuous "1". So killing all data. So just one source at a time It is how digital data works olewill
 
William_H said:
The data comes in a stream of bits each identified by its position in the stream usually starting with a timing pulse. While the baud rate will mean that the stream comes at a similar frequency the individual timing will be quite random. So a pulse meaning S could be confused with a pulse from another device being part of a stream giving time or speed. Indeed the pulses could all be combined to give a continuous "1". So killing all data. So just one source at a time It is how digital data works olewill
Click to expand...

Seatalk allows multiple talkers on the same line and has collision detection and recovery. Collisions happen all the time and it copes. One source at a time does not mean that only one talker can be connected to the data line in a collision detection/recovery system like Seatalk, as long as each talker does not talk all the time. Or Ethernet.

If two sources are writing the same data on the same line I can't see a problem other than a bit of wasted bandwidth, but there's plenty available on the average Seatalk system anyway.
 
AngusMcDoon said:
Seatalk allows multiple talkers on the same line and has collision detection and recovery. Collisions happen all the time and it copes.
Click to expand...
I don't think seatalk will be happy with two GPS data streams in the system.

I have a NMEA 0183 system with 2 GPS units (one for back up) if I turn both on at once the system, as I would expect, starts getting confused losses data etc. By turning on only the minimum NMEA sentences I can reduce the corruption, but not eliminate it.
Seatalk is similar in many ways to NMEA and I would not expect to see it managing much better, but I have never tried it so its not impossible it will work.

I don't think the gps in the AIS will cause any problems as this is normally not broadcast over the network
 
noelex said:
I don't think seatalk will be happy with two GPS data streams in the system.

I have a NMEA 0183 system with 2 GPS units (one for back up) if I turn both on at once the system, as I would expect, starts getting confused losses data etc. By turning on only the minimum NMEA sentences I can reduce the corruption, but not eliminate it.
Seatalk is similar in many ways to NMEA and I would not expect to see it managing much better, but I have never tried it so its not impossible it will work.

I don't think the gps in the AIS will cause any problems as this is normally not broadcast over the network
Click to expand...

Seatalk is similar to NMEA in that it is serial and 4.8kbaud. That's where the similarity stops. NMEA is single talker as it has no collision detection. Seatalk is multi-talker because it does have collision detection. In a Seatalk system all the transducers writing data on the bus are connected to the same single data line. I can't see why there will be any more problem if two of the talkers are writing the same thing than the normal situation of multiple talkers writing different things.
 
AngusMcDoon said:
Seatalk is similar to NMEA in that it is serial and 4.8kbaud. That's where the similarity stops. NMEA is single talker as it has no collision detection. Seatalk is multi-talker because it does have collision detection. In a Seatalk system all the transducers writing data on the bus are connected to the same single data line. I can't see why there will be any more problem if two of the talkers are writing the same thing than the normal situation of multiple talkers writing different things.
Click to expand...

Even if it accepts the data without crashing I can see problems handling the slightly different positions from the two GPS units.
There could be an almost instantaneous change in position as the input switches from one gps to next. Even if this movement is only 10m or so all the derived information from changes in position will see see see some very weird results. This could be removed with filtering and averaging, but the net result will be inferior to single GPS.

The other problem is 2 GPS units give you twice the chance that one has failed. The failure mode could involve an incorrect position rather than no position. Unless there is some means of disconnecting this GPS unit the overall position displayed could be wrong even if one unit is functioning normally.

Seatalk information can be diverted and turned off and on with simple electrical switches although it less versatile than NMEA in that the the input and output cannot be separately switched. Switching the GPS units in this way (if possible) is preferable IMHO.
 
I've just been reading the documentation with the E7 and it seems that there is a deconfliction algorithm which can be set to use only one GPS input. Thus if I switch on my second GPS,the signal will be ignored unless the E7 fails.

I have also found that the AIS650 must use its own dedicated GPS antenna in order to transmit GPS position, even if connected to Seatalk. Strange that Raymarine didn't just use the GPS data from the Seatalk input.
 
Ric said:
I have also found that the AIS650 must use its own dedicated GPS antenna in order to transmit GPS position, even if connected to Seatalk. Strange that Raymarine didn't just use the GPS data from the Seatalk input.
Click to expand...

It's a requirement of the AIS specs.

Pete
 
William_H said:
The data comes in a stream of bits each identified by its position in the stream usually starting with a timing pulse. While the baud rate will mean that the stream comes at a similar frequency the individual timing will be quite random. So a pulse meaning S could be confused with a pulse from another device being part of a stream giving time or speed. Indeed the pulses could all be combined to give a continuous "1". So killing all data. So just one source at a time It is how digital data works olewill
Click to expand...

This is not true for SeatalgNG /NMEA 2000 networks
 
noelex said:
With the Raymarine RC425 quoted in the OP we are talking about Seatalk. Seatalk NG, despite a similar name, is very different.
Click to expand...

The giveaway is actually the E85001: Not sure the RC425 even has seatalk: The OP says it's connected to the seatalk bus via the E85001.

To repeat what AngusMcDoon says in posts #6 and #8, seatalk 1 does have collision avoidance: Unlike NMEA-0183 where the signals from two simultaneous talkers would likely produce garbage, seatalk talkers wait for the bus to be free for a bit, then start talking, but they also read back what they've written. If what they read isn't what the wrote (ie it's garbage because another talker started talking at the same time) they stop and wait for the bus to become free for a bit again.

I do agree with what you say in post #10: The ability to have two simultaneously active GPS units doesn't mean it's a good plan: "Failure" could involve inaccurate information or mucking up the bus rather than just going quiet.

The OP does seem to be using "seatalk" for both seatalk 1 and seatalk 2 so I wonder if Ric plans to bridge the two with an ST70 or a seatalk-seatalkNG converter not mentioned in post #1
 
You are correct - RC425 has no seatalk - it outputs NMEA into Seatalk via E85001.

I do indeed also have a ST-STNG converter.

My plan is that most of the time the RC425 will be switched off. If the E7 goes pop, then I will just switch on the RC425.
 
AngusMcDoon said:
Why?
Click to expand...
How will the network decide which position to use. I had two fixed GPS plotters fairly consistently producing positions 0.001 mile apart - because the aeriels were 6 feet apart.

Even if they were side by side, lag in acquiring new positions as you move will inevitably introduce differences.
 
You must log in or register to reply here.

Share this page

Top