NMEA 2000 buss crash

[george123]

Four years ago I installed a system that has the following; Furuno NavNet3D display, Simrad AC42/AP28 autopilot, WSO100 weather station, DST800 transducer, two NMEA2000 GPS antennas, and a furuno rs-33 display for nmea data. The system worked great for 3 1/2 years. Six months ago I installed the a75 (software v16.47) display connected to the NMEA2000 backbone. The boat has not been used that much during the past six months. The a75 displayed all the NMEA2000 data that was on the buss. Four weeks ago something strange started to happen and that was randomly the NMEA buss would crash and all the data on the buss would be gone. The system crash would happen 20-40 minutes after the system was up a running. The DST800 would report correct information when everything was running. I started to trouble shoot the system and when I removed the DST800 from the buss the data for the other devices that were on the system would come back. So I concluded that I had a bad DST800. I have sent this back to airmar for evaluation and they have told me that the DST800 is fine. The only new device o the nmea2000 buss is the a75. I have removed other devices on the buss and can not isolate the problem. What I did next was randomly changed the instance numbers for each device from there default 0 number. So now each device has a unique instance number assigned including the weather station which has 4 channels each having it's own instance number. we ran the system for 20 hours with no failures. I do not understand while this seemed to fix the problem, any thoughts.

[tanglewood]

Since the A75 was the last addition, and seems to correspond with when the problem started, it seems the most likely cause or trigger for the problem. The other possibility is that something else failed at the same time, but that seems much less likely.

I would guess that the A75 is transmitting data that is also being transmitted by some other device. And this has caused the issue. Changing all your device instance numbers probably worked around the issue.

Note that Maretron is the only vendor who it super sensitive to device instance numbers. They require unique instances to select between any like devices, e.g. GPSs, depth sounders, etc.. Other vendors select sources based on a combination of the device's function, model, manufacturer name, and the device's unique serial number. So they can select between two GPSs for the same model and same vendor based on the serial number. All devices instances can be zero and everything will work. But this isn't true with Maretron.

Sorry for the side bar, but N2K Instances are greatly misunderstood.

[s/v Jedi]

Any update on this George?

[george123]

Yes we fixed the problem. What we did was measure the resistance on the buss. It should of measure 60 ohms. The reading we got was 53 ohms which was to low. We began removing one device at a time. When we removed the Maretron weather station the buss jumped to 59.6 ohms. We returned the unit to maretron for repair. They determined the data buss driver was bad. Everything is working great now.

[tanglewood]

Good example of diagnosing a problem on N2K.

The down side of a buss structure is that just this sort of problem can take everything out. But the good news is that if you set up your network with accessible connectors, it's pretty easy and quick to locate and isolate a problem.

[s/v Jedi]

Originally Posted by george123

Yes we fixed the problem. What we did was measure the resistance on the buss. It should of measure 60 ohms. The reading we got was 53 ohms which was to low. We began removing one device at a time. When we removed the Maretron weather station the buss jumped to 59.6 ohms. We returned the unit to maretron for repair. They determined the data buss driver was bad. Everything is working great now.

I had the exact same problem with the WSO100. Incl. myself, you are #3 that I hear of this problem.

[smac999]

the A probably has an internal GPS? so you added a 3rd GPS to the system. all with the same instance. may have been the issue. though they are supposed to deal with this automaticly

[lateral]

Originally Posted by george123

Yes we fixed the problem. What we did was measure the resistance on the buss. It should of measure 60 ohms. The reading we got was 53 ohms which was to low. We began removing one device at a time. When we removed the Maretron weather station the buss jumped to 59.6 ohms. We returned the unit to maretron for repair. They determined the data buss driver was bad. Everything is working great now.

Good explanation here.
How do I check the health of my NMEA 2000 cabling before powering my network?

I will be using this to diagnose my PB200 crashing my bus if I can't find the fault in the device/cable. Works fine without the WS & just terminated.