Raymarine NMEA 2000 network

[theglo]

I have a system powered by a Raymarine Axiom 7 with 2 x GS165 chartplotters, sonar and a host of NMEA 2000 devices attached.

I'm having an issue where, intermittently, I will lose all NMEA 2000 data.

The NMEA back bone is still powered because my i50 depth stays powered; it just loses depth.

When I open my Network list of devices one of three things happens: 1. I see everything (Network and Nmea) and everything works, 2. I only see Network items and no nmea and no nmea works or 3. I see everything but only the network items work and no nmea works.

I've updated the software.

I've checked every connection I could locate.

Ive attached a video of my i50. At about 7 seconds in you can see it flash. I'm not sure if it momentarily is losing power or if that's normal when it is searching for data.

Thanks for any help.

Heres the link for the video:
https://youtu.be/lrgYVrFsFlc

[goboatingnow]

Are the terminators on the 2k bus

[skipmac]

Any chance something electrical is kicking in to give a temporary voltage drop? Water pump, AP motor, fridge compressor or something like that? Are the batteries in good condition, all the connections clean and tight?

[Dsanduril]

At it's heart NMEA 2000 is a CAN serial network. It is possible for a single errant transmitter to jam the whole network. If you find no other solutions cut the network in half and re-terminate. See if the problem goes away. If it does there's a chance there's a problem transmitter in the disconnected half. If it doesn't go away, cut the active half in half again and repeat. If you can get to a segment that doesn't have any problems then add transmitters back in one at a time and see if you can find one with problems. Depending on the number of nodes and how hard the drops are to access it can be a very time consuming process.

There are some network inspection software tools that would allow you to look at error counts, etc., which might give you a better picture if you have access to them and can get them connected to the network.

[goboatingnow]

Originally Posted by Dsanduril

At it's heart NMEA 2000 is a CAN serial network. It is possible for a single errant transmitter to jam the whole network. If you find no other solutions cut the network in half and re-terminate. See if the problem goes away. If it does there's a chance there's a problem transmitter in the disconnected half. If it doesn't go away, cut the active half in half again and repeat. If you can get to a segment that doesn't have any problems then add transmitters back in one at a time and see if you can find one with problems. Depending on the number of nodes and how hard the drops are to access it can be a very time consuming process.

There are some network inspection software tools that would allow you to look at error counts, etc., which might give you a better picture if you have access to them and can get them connected to the network.

It is not possible for a single can transmitter to jam a can network , both the can controller error mechanism and the anti jabber feature of the transceiver also prevents it. ( not to mention the bitwise arbitration system )

[Dsanduril]

Originally Posted by goboatingnow

It is not possible for a single can transmitter to jam a can network , both the can controller error mechanism and the anti jabber feature of the transceiver also prevents it. ( not to mention the bitwise arbitration system )

That's the intention, but it relies on the controller on each node to behave properly, inspect its own behavior, and take itself off the bus if it finds problems with itself.

If you have single transmitter stuck with a dominant bit on, and that same transmitter ignores turn off commands (which can happen with a bad transmitter) then the bus is jammed. Can also happen if a transmitter sees a different value on the bus than what it transmitted and gets stuck with retries. Both of these are supposed to be dealt with by error catching mechanisms, but they can/do fail. Both are also the result of (usually) hardware failures on the offending transmitter side. When intermittent it can be temperature, or....

Both of these behaviors have been documented in DeviceNet and other CANs.

Not saying it's frequent or common (the whole point of all the error checking in CAN is to make it uncommon) but if you've gone through the obvious and the easy....

[theglo]

Originally Posted by goboatingnow

Are the terminators on the 2k bus

There are 2 white terminals that are missing terminators. I've ordered those earlier today as I've seen that cause this issue before.

[goboatingnow]

The separate can transceiver chips have specific dominant bit anti-jabber timers and will disable the transmit function if triggered. This is separate to all the error passive and bus off systems in the controller itself.

It’s an extremely rare failure mode and while it’s “ possible “ it’s not “ probable “

[theglo]

Originally Posted by skipmac

Any chance something electrical is kicking in to give a temporary voltage drop? Water pump, AP motor, fridge compressor or something like that? Are the batteries in good condition, all the connections clean and tight?

It's none of the above. Batteries are great. It's a 60 foot yacht. I'm working on this. It's not my own sailboat. Thanks for your help though.

[theglo]

Originally Posted by Dsanduril

At it's heart NMEA 2000 is a CAN serial network. It is possible for a single errant transmitter to jam the whole network. If you find no other solutions cut the network in half and re-terminate. See if the problem goes away. If it does there's a chance there's a problem transmitter in the disconnected half. If it doesn't go away, cut the active half in half again and repeat. If you can get to a segment that doesn't have any problems then add transmitters back in one at a time and see if you can find one with problems. Depending on the number of nodes and how hard the drops are to access it can be a very time consuming process.

There are some network inspection software tools that would allow you to look at error counts, etc., which might give you a better picture if you have access to them and can get them connected to the network.

Thanks. These are the steps I've been taking. I'm actually working on this boat at my job. It is not my own boat. I was there for 6.5 hours today tracing the network. I did not find all the connections. It's a 60 foot motor yacht and quite difficult to trace everything. I was hoping for something I was missing that I could look at that would be a bit easier than trying to find the missing pieces of the network I haven't located yet. But, I agree, that may be the process.

THanks

[goboatingnow]

Originally Posted by theglo

It's none of the above. Batteries are great. It's a 60 foot yacht. I'm working on this. It's not my own sailboat. Thanks for your help though.

Check the terminators

After that remove any nmea 2K devices that arnt involved in the suspect data , see if it works then. Slowly add the other nodes back one at a time.

[theglo]

Originally Posted by goboatingnow

Check the terminators

After that remove any nmea 2K devices that arnt involved in the suspect data , see if it works then. Slowly add the other nodes back one at a time.

Yep, will do.

My other current possibilties are:

1. When I ran the software update it seems that none of the NMEA 2000 devices were actively connected. Meaning the software only updated the items on the Marine Network and not on NMEA. I plan to do that again next time it magically has everything connected.

2. The NMEA 2000 backbone is getting power from two sources. One I've found but the other is located somewhere in its run that I have not yet located. I also haven't found the Autopilot Computer (major item I know) and I suspect it may be where its getting its other power source. It seemed strange to me to power the backbone from two sources but after talking to Raymarine that is normal and okay.

[goboatingnow]

Hi.
Just a point the i50 displays “- - -“ when the depth transducer can’t resolve the bottom . It’s a common problem in certain bottom conditions

Have you seen it happen under sail etc.

[theglo]

I'm aware it does that. It's not blinking like when it can't see bottom. It's a data issue. Sometimes it randomly works. Also, there is a seperate sounder on the marine network connected to a different transducer and that is functioning perfectly fine.

It's not just the i50 that goes out but the 20 or so NMEA 2000 devices. The i50 is just a good tell because its on the dash and is powered by the NMEA network as well as getting its data from it.

Thanks!

[smac999]

Originally Posted by theglo

There are 2 white terminals that are missing terminators. I've ordered those earlier today as I've seen that cause this issue before.

sounds like you have a seatalkng network and not a nmea 2000 network. but they are simular.

you can only have 2 total terminators. and they must be at each end. and no white cable can be over 6m.

if you have empty white STNG ports, you can buy black plugs, they are not terminators. terminators are blue.


I would be carefull updating with an itermitant issue. if the nmea 2000 devices drop off while updating. you may brick it.

updating is not going to fix nmea 2000 devices coming and going from the network.