Adventures in NMEA 2000 Wiring – Part I

22 Responses

  1. Lookout Sailors says:

    This looks like a great idea. I’m intrigued by the Ethernet approach. I’ve wondered if the Seatalk hs network was just a router. I’d like to plug my laptop into the Ethernet backbone instead of a multiplexer.

  2. Chris Ellingsen says:

    It is a good idea to have two separate backbones as you describe. We have done something very similar with seatalk so that the instruments can still be used when the autopilot and other larger loads are turned off.
    I have my doubts, however, that the two chartplotters will relay all the data from one buss to the other, especially through the ethernet. Raymarine is very iffy with this kind of thing and it is never very well documented.
    Your other approach of bridging the two busses with only the data lines is probably better, but you may find that when the underway network is turned off that the powered off devices create a load on the data lines that stop things from working reliably. We found this with seatalk.
    I would suggest making the connection with some kind of relay that only bridges the two busses together when the underway network is powered up, and otherwise terminates the powersave network.

  3. Anonymous says:

    Seems like you could have tested whether or not you’re going to be able to bridge the two networks BEFORE you spent all the money to wire them both into your boat. The way it is now you might have no GPS in your BLUE network.
    From the looks of things the only think you need on both segments is the GPS. If you can afford two chartplotters that are probably max 10′ feet from each other you can afford a couple hundred on a GPS antenna and not have to deal with the fragility of your home made bridge.

  4. Dan Corcoran (b393capt) says:

    I have no illusions will directly bridge the two N2K networks, rather the e-series will internalize the information from one N2K network (e.g. boat speed = 2.00 knots), pass it over the ethernet in another format, and it arrives on the other N2K network (and potentially seatalk and 0183 networks as well) after being retransmitted.
    So … I would not expect any PGN’s the e-series does not recognize to be bridged between the two networks, which would mean any attempts to configure an N2K device or perform a software upgrade, must be done by devices on the same backbone. If a device is configured via PC, that’s easy, I plan to be able to easily move the gateway between networks, otherwise a power off/on sequence add the bow connector temporarily engaged would create just one network end to end.

  5. Dan Corcoran (b393capt) says:

    I have been giving thought to just having the bow connection in place, but Chris has given me another concern to check out first … he wrote that the devices on the unpowered bus can put a load on the data wires. I suspect the impedance / signaling design of CANBUS would decide this without experimentation. Is it just like having six TV’s at home, five of them off, and the remaining TV has no problems? Can anyone weight in on that ?
    Another consideration for a permanent bow connection, what’s the best way to avoid a ground loop, if I am choosing to leave power and ground disconnected at the bow between the two networks, do I …
    (i) Leave the shield disconnected at the bow
    (ii) Leave the shield disconnected at the under way power tap ?

  6. Victor says:

    Interesting approach. In the article you place the weather sensor on the gray-bus, but in the picture it’s on the blue. (Simply a typo?)
    I would prefer to connect the Weather sensor to the gray-bus. As I usually monitor the weather closely when at anchor. Especially the wind angle.
    Regards, Victor

  7. Anonymous says:

    Are you eliminating all Seatalk(1) devices from the boat?
    I have a challenge with the fact that documentation says not to conenct both N2K and Seatalk(1) at the same time to a SeatalkNG network. Conencting the Eseries to N2K via SeatalkNG-N2K adapter means no Seatalk(1) devices.
    That would be easy, except that I have a very nice wireless autopilot control that is ST(1), and the Seatalk-PC-NMEA bridge that is ST(1), and also some ST80Maxiview displays and their remote controller that are ST(1), and so I’m not supposed to be able to use NMEA2000 if any of these are connected.
    Do you simply have no Seatalk(1) items? I don’t see speed or depth transducer pods on the diagram.

  8. Dan Corcoran (b393capt) says:

    Yes on seatalk I have connected depth, speed (actually I am using an ST70 speed pod on NMEA-2000 the last two weeks), and seatalk autopilot that are not in the N2K diagram.
    And not just those, I have numerous seatalk (GPS, autopilot, depth, speed, autopilot remote) and nmea-0183 (Airmar PB200, VHF radio, autopilot fast heading output, autopilot nmea0183 input) devices on Breeze that are connected to each of the two e-series that work together collecting and distributing information across my boat (e.g. boast speed from my tri data display (seatalk) network appears on my N2K and 0183 connected electronics for example).
    Check the documentation your referencing, I can’t imagine what you read to lead you to a conclusion other that these can be easily intermixed.

  9. Dan Corcoran (b393capt) says:

    Victor … the entry is correct, perhaps a little confusing. The weather receiver is what I look at while at anchor, it’s connected to the ethernet and goes into the single chartplotter (nav station) which I would have turned on at anchor. The receiver provides doppler radar images, NOAA forecasts, and more.
    Could easily make a good case for wind direction and speed (barometric pressure, air temp, etc.) to be on the gray network … since the two network backbones are side by side, I could easily make this change on the fly if I desired.

  10. Dan Corcoran (b393capt) says:

    Victor … on second thought, that’s a pretty good improvement, the weather station will be moved to the gray network and be available as well as the weather receiver.

  11. Anonymous says:

    Why not separate the two legs with one power tap to separate breakers. If in fact you are using the Maretron Power tap this can be done. I found a post on Maretron’s Knowledge Base describing the connections of their power tap . This will help avoid a gender issue on the network.
    In reference to the impedance issue the data pair does not power the transceiver the power pair does. If a transceiver is inactive this will not affect the overall condition of your transmission lines.
    Un-hooking the shield decreases the path of stray voltage drain from the cable allowing noise to increase.

  12. Ben Ellison Ben Ellison says:

    When I first saw this post, before it was published, I asked Dan why he didn’t just separately power each side of the same backbone to achieve his goal. The answer is that the underway and non-underway devices don’t line up conveniently.
    But I’m told that this technique works fine, if you can physically arrange the gear on each side of a split power tap. Unpowered CANBus devices purportedly do not cause any drains or problems on the powered part of the network.
    Whereas Dan will have two backbones with split powertaps and may likely connect them, he may end up with an N2K network with four power segments. I believe this perfectly legal as long as they share the same ground, and I think he’ll be able to power any one or two or three segments. I’m hoping he experiments (which is part of design goal anyway).

  13. Dan Corcoran (b393capt) says:

    I would like to say I am experimenting going forward, but that might be a wild claim.
    This isn’t as off the wall as it might look, especially if you take away one chartplotter and install the cord set at the bow of the concept diagram shown (cord set bridges two data wires only, note both yellow power taps share power ground and shield). Let’s call this the “one chartplotter version” of the same concept.
    Now looking back at the once chartplotter version of the network concept diagram, move the two power taps to the bow, make it just a single power tap, with one side of the tap on an underway circuit breaker and the other side of the same tap on a powersave circuit breaker … does the network look at all experimental ? If I had done it that way, does it provide a clear benefit in your mind over running a single backbone ? Does it look like a no brainer that it would work ?

  14. Jeff Shukis says:

    Hi Ben,
    Are there any SeaTalk1 instruments on your network? If so, how have you managed to get their data onto the NMEA 2000 network?

  15. Ben Ellison Ben Ellison says:

    Jeff, I don’t have SeaTalk1 one my N2K network but Dan (who wrote this entry) does on his. Sort of. As he explained in a comment above, he has SeaTalk sensors connected to his E-Series MFDs, which then bridge at least some of the data over to NMEA 2000.
    A number of different brand MFDs do this sort of bridging but it’s usually not well documented and uncertain. I have used a Raymarine pod to turn SeaTalk1 wind sensor output into N2K messages, and it worked OK:

  16. Dan Corcoran (b393capt) says:

    Yes, I have lots of seatalk1 transducers (speed, depth, wind), ST60/ST60+/ST70 displays, autopilot, and autopilot wireless.

  17. Jeff Shukis says:

    Following up on the SeaTalk question: When I added an NMEA2000 network to my existing Raymarine SeaTalk gear (E120 plotter, wind, speed, depth, GPS, Autopilot), I hoped/expected that the E120 would bridge *all* of the SeaTalk data onto the NMEA 2000 network so that my Maretron N2KView software could see it. This isn’t happening on my boat. In fact, I see none of the basic sensor data. I know that the NMEA2000 network is working properly because I *am* seeing engine data on the E120 that is being created by an NMEA2000 box from Maretron.
    Do I have to wait for the Actisense NMEA0183 to NMEA2000 bridge to be released?

  18. Jason says:

    Does anyone know, for certain, why Raymarine indicates that connecting a Seatalk(1)+Seatalk ng to a N2K network is not recommended?
    Or, who has these three network types connected aand sharing information?

  19. Dan Corcoran (b393capt) says:

    In response to a question above from Jeff Shukis : Yes, I say with certainty the seatalk data is bridged from my e-series between the seatalk and NMEA-2000 bus. I was able to observe the data from my seatalk instruments on the NMEA-2000 bus using the Airmar NMEA-2000 to USB gateway and WeatherCaster software. Specifically I saw depth, speed, water temp, heading, and GPS position.

  20. Owen Schneider says:

    Does anyone know if you can generate data in a NMEA 2000 network from the Maretron engine data box and a NMEA 2000 transducer (speed/depth), then put it into a Raymarine Seatalk NG network, add say wind data from an ST-70 TD Pod then put it through an ST-70 to feed ST-60 displays (no data creation just reading) via a Seatalk 1 network? Or is Raymarine being that harsh about creating connections between NMEA 2000 and Seatalk 1. I guess one could just use a NMEA 2000/0183 bridge to do the same along with a NMEA 0183 /Seatalk 1 bridge (E85001)? I would hate to start adding more Seatalk 1 pieces when they are being phased out and if Garmin buys Raymarine.

  21. Dan Corcoran says:

    This article aged pretty well. In retrospect, I would comment that the 2nd backbone was entirely unnecessary. As much as I tried, I never put it to use. On my next boat I just have 1 backbone.

    • Ben Ellison Ben Ellison says:

      Hi Dan, I had the same intention when I stripped all the test systems off Gizmo along with the second N2K network. But after the redo, and a month or so with only one network, I went back to two. The change solved some issues — like the fly bridge autopilot controller showing data but not responding to commands — and I like having the redundancy again. It was a pretty busy single network, however, and included devices from many different manufacturers.

Join the conversation

Your email address will not be published. Required fields are marked *