Slightly non-linear NMEA 2K network

[RWCarr]

I have a 36' sailboat with lots of (mostly) Garmin NMEA 2000 devices from the wind instrument at the mast top to the GPS antenna on the stern rail, with chart plotter, autopilot, VHF AIS, and displays in between.

I have one drop to the chartplotter at the helm that snakes through the boat to a NavPod mounted on the wheel pedestal. I want to add an N2K autopilot control on the pedestal. Do I need to snake another N2K drop from the backbone to the pedestal.

It would be a lot easier to add a T connector to the end of the chartplotter drop cable and a short cable to the autopilot control.

Assuming I have a linear N2K network with two terminators, etc., how will the network degrade if I connect two devices on one drop off the backbone?

[rgleason]

Reconfigure your main backbone using the cable to the station as a backbone extender by removing one of the terminators. Look at nmea2k diagrams. Follow the n2k wiring requirements.

[JBP]

Officially, one device per drop cable. Unofficially, people have done just what you describe and it works fine.

[Knotical]

Two thoughts-

1. Why do you need the AP control head and the CP at the helm? I have CP at the helm and AP control on companionway bulkhead, we spend more time in front of the wheel and having AP there is really nice.

2. If you absolutely have to install both at the helm and don’t have space to fish an additional cable can you replace the drop cable with the backbone cable? That way you can connect both instruments properly.

[RWCarr]

Thank you all.

I may be able to configure the CP drop as part of the backbone. That would be best. There is already a long extension of the backbone to the over-the-companionway speed/depth/wind displays and to the masthead wind.

Otherwise, I'm glad to hear that a small step outside the bounds of N2K networking rules will probably work OK.

BTW, B&G have a great tech note on troubleshooting a N2K network with just a multimeter and a N2K cable. https://www.bandg.com/en-gb/help--support/

[merrimac]

@RWCarr, here is a fixed link: How to Troubleshoot an NMEA 2000 Network using a Multimeter (advanced users)

[wholybee]

I've not seen the official standard or any definitive "you can" or "can not" split a drop cable. So, I don't know if it is "allowed" or not.

I can say this. Electrically, devices are a high impedance, and the network/backbone will not see any electrical difference if you split a drop cable.

Also, Maretron makes a drop cable splitter box. (Multiport box)
https://www.maretron.com/products/cabling.php

[sailingharry]

As I understand it, it will work absolutely fine the way you wish. B&g instruments used to come with the ability to daisy chain, which is electrically exactly what you want.

I think that the reasoning for the n2k standard has to do with reliability and troubleshooting. I think their philosophy is you should be able to remove any device at the t on the backbone and not impact anything except that device.

While I applaud their focus on reliability, it does lead to installation complications.

[smac999]

if you keep the total of the 3 cables (drop to splitter and then both devices to splitter) under 6m I would not be to worried. in fact maritron even sells a spliter block.

[hjohnson]

As with anything, it depends.

An NMEA bus is built the way it is to avoid reflections off the end of the cable. At 250kbps, it's entirely possible for the signal to ring back and forth on the cable. The terminators add loading to the bus, which makes it "infinitely long" electrically, and thus eliminate the echos.

The issue you could run into is if a drop cable is too long, say a drop cable that goes all the way up a 70' mast or some such. (Fortunately, the Garmin stuff you talk about doesn't run CAN up the mast; the cable is carrying the old school Nexus network, which is based on RS485).

Anyhow, having a single drop cable feeding all the stuff in your helm pod is probably not going to be an issue, as long as the entire drop isn't too long.