NMEA 2000 bus crash

[Chuck - Raymarine - Moderator]

George,

While I can't dispute that the MFD should not be transmitting the PGN 128267 (Water Depth), the NMEA 2000 system should support multiple sources of depth. In support of submission of a problem report, it is requested that the MFD's diagnostics feature be used to log sufficient data from the SeaTalkng/NMEA 2000 backbone to show to include an instance of PGN 128267 being transmitted by the MFD.


In systems where multiple sources of the same type of data are present, system devices using such data will typically provide a mechanism to specify which source of data will be used by the device. In the case of Raymarine MFDs, have you used the Data Sources feature to specify which of any duplicated sources of data within your system have been used by the MFD?

[george123]

I will be going to the boat later this week to do some additional testing once I received the DST800 back from Airmar. I will reinstall the device and begin troubleshooting. I will also do some data locking to see what is happening on the a75. I will also bring the a75 up to the latest rev.
I will report back

Thanks
George

[Chuck - Raymarine - Moderator]

You're welcome.

[george123]

Well I went to the boat yesterday and reinstalled the DST800 and configured the transducer depth offset, speed calibration, and temp calibration with the Furuno MFD. I also updated the a75 with the 17.45 version. The system would still crash after 20-40 minutes. I began to remove one NMEA device at a time and could not resolve the issue. So I said what devices have a common PGN transmitted. The WSO weather system had a common temperture PGN with the DST800. I changed the device instance number on each of the WSO100 sensors temp, wind, pressure. We let the system run for 20 hours and everything was still working. I do not understand this. The system worked for the last three seasons with no problem.
One other point when I set the data sources on the a75 for the depth and speed after it scanned the network it would come back with device unknown on the display. It had a check mark nexted to it as well. The a75 would still show depth, speed, temp ok. Why do I see device unknown?
I have attached a copy of the nmea2000 network for your review. The only think that changed this season was adding the a75 in May. From May to the end of July the system worked fine. In early Aug is when the buss was crashing.
I am pulling my hair out to understand what is happening.
Also should the a75 be set as master or slave?

Thanks
George

[Chuck - Raymarine - Moderator]

George,

Based upon the system diagram which you have supplied, it does not appear that you have a properly constructed NMEA 2000 backbone. Specifically, the Simnet appears to have been installed as a spur rather than as part of the backbone ... i.e. the RC42, AC28, and AC42 (assuming that these are all NMEA 2000 certified products) should each be installed as an individual spurs to the backbone.

Is its also not clear where the backbone's power insertion point is located. Within a properly installed backbone, the single power insertion point should be located at the approximate midpoint of the backbone's LEN load.

With respect to the system functioning in the past and not functioning now, I can only suggest that it may have been teetering on the edge of the incorrectly constructed backbone's tolerance, waiting for an additional device to push it over the edge.

Finally, with respect to the a75 being configured as a Data Master, this setting only applies to systems featuring an Ethernet network of more than one Raymarine a/c/e/eS/gS-Series MFD. In such systems, only the Data Master MFD will typically be interfaced to the SeaTalkng/NMEA 2000 backbone and it will in turn share data with the other networked Raymarine MFDs. In systems having only one Raymarine a/c/e/eS/gS-Series MFD, the MFD must be configured as the Data Master (the default state of the MFD).

[george123]

I will look at the network design to verify the layout. I am powering the network in the center. In the drawing you can see the power tap just above the usb100 device 12v at 4 amps.
I will make sure the a75 is set as master.

Thanks
George

[Chuck - Raymarine - Moderator]

George,

I do suggest that you review your diagram relative to the Simnet spur.

Additionally, the Load Equivalency Number (LEN) total of the devices to the right of your power insertion point (5) appears to be significantly greater than LEN total of the devices to the left side the power insertion point within your drawing (the LEN load of the RD33 alone is 6). It is suggested that you review the documentation for the products within your backbone to ensure that you have obtained the correct It is additionally recommended that you confirm that no devices are supplying power to the backbone or a ground loop may result.

[george123]

I have rewired the network as seen in the attached file. I also balanced the current distribution across the network. The current is very close to being equal on each side of the center tap voltage supply. I have made sure that each device on the network has its own device instance number. I have used the Maretron N2K Builder to review the design. The Builder passed all review test's.
We are still having random network crashes. Last night we decided to remove the a75 and let the system run. It has been over 24 with no failures. We are letting the system run. Next week when I return to the boat I am planning to capture the NMEA2000 data for your review.
Have you seen a design like the one that I have that has two different manufacturers MFD's sitting on the same NMEA2000/Seatalkng network?

Thanks
George

[Chuck - Raymarine - Moderator]

George,

Raymarine products are quite routinely interfaced with third party products, including those with NMEA 2000 communications interfaces. That said, greatest overall reliability, data sharing, and features will typically result in system which has been constructed of products from a single manufacturer ... as this this the environment in which they are tested. Raymarine currently manufactured products are designed to support NMEA 2000 communications specifications rather than be tested with all third party products having NMEA 0183 communications interfaces ... as doing so would be impractical. NMEA 2000 bus crashes such as you have reported are very rare and are typically the result of insufficient power within the backbone, improper termination, oa damaged / faulty networking component, or product failure. Should you believe that there is an issue with your MFD, then it is recommended that it be sent to Raymarine's Product Repair Center to operated continuously within a large network of SeaTalkng / NMEA 2000 devices over several days.

[george123]

Well we found the problem on the NMEA2000 network. When we checked the network with the power turned off we measured 54 ohms. So we began removing one at a time the devices. When we got to the WSO100 weather station the resistance jumped up to 60 ohms the correct value. When I measured the resistance between the two signal pins on the WSO100 on the bench it measured 500 ohms way to low, it should be in the meg ohms. I have returned the device to Maretron and they have determined that the network transceiver has failed. What was interesting the device was still sending out correct data on the network. So the take away from this you should always test the network resistance with the power turned off as one of the first thinks to do if you begin to have random fails over time.