NMEA 2000 Install Question
Posted by thehorselessjockey@reddit | sailing | View on Reddit | 36 comments
(Picture mostly for interest)
Question for the brain trust: we spent not an in substantial amount of money to have NMEA 2000 sailing instrument package and below deck linear drive autopilot installed in our Catalina 42. The installers appeared to have installed the NMEA backbone and autopilot pump directly to the battery bus (inline fuse but not on switch). This results in an always on state for both. Am I crazy to insist this is not the correct way to install and ask that they both be installed to a spare switch on our panel?
whyrumalwaysgone@reddit
Marine electrician here: Sometimes installers do this because people forget to turn on the network and electronics fail to connect on boot up. Or if its on the same breaker as your nav systems, stuff doesnt boot up in the right order. The power draw is trivial, like milliamps. However, nothing should be left on 100% of the time, its not great. Wouldnt hurt to move it to a breaker, but just know that breaker ALWAYS must be turned on first before the rest of the system.
Installing the network this way is common for charter boats, because charter guests are incapable of following basic instructions and turn on breakers in the correct order. Your installer just has a low opinion of your intelligence, or was trained by someone who does.
Regarding the pilot, you say linear drive then also pump? Hydraulic pump drives are not the same as a linear drive, I'm going to assume you mean the drive ram in this case. The ram itself usually gets power from the pilot cpu directly, not an external source. If you provide the ram/pump with power directly, it will stay on (like actually trying to move) forever. So unless you hear a constant whining and massive power draw, thats not what happening here. You need to understand more whats going on before going nuclear, either get a diagram from your installer or trace the wires themselves.
repOrion@reddit
Is there published information on the order in which elements of a N2K system must be powered on to ensure they are able to sync correctly?
Same_Detective_7433@reddit
That is not how NMEA networks behave typically, your devices SHOULD be able to pick up the network regardless of when it is switched on. Devices that do not are simply not following the NMEA standards. Whether you power the network all the time really is a personal preference, although having a way to depower it when leaving for an extended period of time is probably a good idea.
Nothing sucks more than an unexpected dead battery.
On mine, I have solar, which maintains my batteries, so I leave my network, Starlink and cameras, and Raspberry Pi on all the time, so I can check in on my boat. But I do have a breaker to turn everything off if I want.
repOrion@reddit
Fair enough, but I’m sure you’ll agree that there is often a discrepancy between how things *SHOULD* work and how they do in practice.
Sometimes equipment that is part of an N2K network is able to accept power off the backbone AND has the ability to be powered independently.
If that is the case it’s a relatively straightforward fix to power them up independently BEFORE providing power to the backbone. This way it’s possible to have some control over what finishes its boot up protocol first (cause I can’t imagine ONE power supply waking all devices at once leads to a 100% guaranteed repeatable boot procedure).
Same_Detective_7433@reddit
If your device can be powered by more than one source, then other than reading the manual and finding out why, I would probably never power it from the network, as if they do not isolate properly, you are going to have some weird problems one day. Or you could.
Typically the only devices that I have seen that do this are chartplotters or autopilots, and for an autopilot(Raymarine anyways) they pretty much always specify that you use one or the other. My older one has you remove the wire, my Evolution has a switch to depower the N2K, both my Raymarines were clearly marked and warned you. My Axioms are less obvious, but I do not think they actually power the networks, just power themselves, I am not certain without looking.
I have not run into other devices doing that, but of course there are many many devices so who knows. I would never use both sources.
repOrion@reddit
Now I have to ask … how do you keep the device from being powered by the backbone? Somehow modify a cable so it longer connects power to that device but still carries signal?
This is also common on AIS transmitters.
millijuna@reddit
Per the N2K specifications, the CANBus transceiver must be powered by the bus, even if the rest of the device is powered externally. For example, if I flip my electronics breaker, my VHF and AIS won’t talk to each other, even though they’re both powered independently.
This is a deliberate choice to avoid voltage weirdness on the N2K bus, especially when using devices that operate on 24vdc
repOrion@reddit
Interesting. Our VHF communicates with the AIS on NMEA 0183 not the N2K side.
millijuna@reddit
The M509 that I have pre-dates the mandate for an internal GPS (it's also in a lousy location to see satellites anyway) so it gets its position via N2K from our AIS transponder, which also acts as a ship's GPS. I actually am not quite sure if it pulls in AIS contacts from the transponder, or its own internal AIS receiver.
The only 0183 I have left running on my boat is between my compass (Furuno PG-500) and the Vesper AIS, which translates it to N2K.
repOrion@reddit
You didn’t explicitly or implicitly state as much but I have to ask … have you powered your RaspPi off the N2K backbone?
whyrumalwaysgone@reddit
Published? Not that I'm aware of, this is from personal experience. Its a lousy CANBUS protocol from several decades ago, with multiple electronics manufacturers butchering the implementation a hundred different ways.
Basically you need the backbone power up, each device powered up, then everything starts playing nice. Ideally sensors first, but usually that isnt practical. Im working this week on a big Furuno install that throws a stack of errors on startup until everything comes online. Its pretty common
repOrion@reddit
Yeah I kinda figured it’d be possible with single components being individually powered, but once you start having elements powered from the backbone it becomes impossible to accurately time the startup.
Just thought I’d see if there’s a best practice.
Our boat needs the AIS on by itself first followed by backbone, finally Autopilot CPU.
We’re going to add a bunch more to the N2K this season so 🤞 we can figure out the new startup.
whyrumalwaysgone@reddit
Some parts of the system can be set up with timeouts or use the little "on" wire on your MFD (usually spliced to the red power lead) if you want to automate some of this. Lots of Maretron and Airmar stuff has internal settings you can mess with using an Actisense n2k->USB tool, or sometimes sub menus in plotters or CZone systems
repOrion@reddit
One day I may actually need to get one of these fancy MFDs I hear so much about.
TriXandApple@reddit
Still way better than what we had before.
thehorselessjockey@reddit (OP)
Thanks for the detailed reply.
Apologies for the incorrect terminology (still learning here!). We had an Octopus Drives Hydraulic Linear Actuator installed. They labeled the power “autopilot pwr” so I assumed it was the actuator power but now as you suggest it’s probably coming from the cpu. Would you recommend leaving that directly to battery bus? We were upgrading from an underpowered wheel autopilot (which was on a breaker,hence my confusion).
All our instrumentation (depth/speed, wind, tank monitor) except MFD, radar, and AP are now on the NMEA 2000 back bone. If have them move NMEA to a switch (or do it myself) order of operation would be fire up NMEA then MFD/autopilot/radar?
Powerful_Bluebird347@reddit
I prefer and many do to have the NMEA backbone energized by a switch. On my boat this called “navigation electronics”. The larger items on the backbone need independent power they don’t get their power from the backbone typically. I prefer to have the following switched on independently: AIS, autopilot and radar. There is one caveat to this the autopilot controller does get powered on but the drive unit and computer does not it has independent power from the autopilot switch.
whyrumalwaysgone@reddit
Ah, tank monitor. Thats why they did it that way.
Arguably they could have asked you your preference first, but if you have NMEA2000 tank gages most people want them on all the time. You wouldnt normally expect a customer to want to turn on a breaker and wait a bit every time they want to see fuel, water or black water levels. So backbone needs power 24/7. Generally this gets turned off if battery switch is switched off, otherwise 100% up time is desirable here.
There are quite a few different ways to configure an autopilot install, hard to say from here why theres an always on circuit. My suggestion would be to download the installation manual for your pilot model and see if they require an always on curcuit for memory or something. Unusual but not impossible. Or just ask your installer why they did it this way.
thehorselessjockey@reddit (OP)
Well they didn’t know about the tank monitors as I’ve installed the SCAD style monitors after their install.
I’ll reach out to them and ask the rationale, though I suspect there isn’t a good reason.
mikeboatman@reddit
The backbone should have a power isolation T and on one side tanks, the other electronics. Power tanks all the time, electronics as needed.
NarcisSisyphusRankin@reddit
"However, nothing should be left on 100% of the time, its not great."
If the boat is left in the water the bilge pumps absolutely should be left on 100% of the time."
Ancientways113@reddit
You are not crazy to aak for this.
jacky4566@reddit
Autopilot pump usually does not normally need to be switched power. It only draws power when pumping.
The NEMA backbone, i would want that switeched. Depends what its powering. Ours is on a switch for "Instruments" However some people want it always on. Again, depends what its powering.
millijuna@reddit
I leave my N2K, AIS, and StarLink circuits on 24/7... I'm actually probably going to grab a couple of the toggle guards for them that Blue Sea makes (or 3d Print some guards) so that I don't accidentally bump those circuits off.
For me, big motors (eg Windlass and larger autopilot drives) should be on disconnect switches, either appropriate breakers, or fuse + disconnect, so that they can be locked out when servicing them. To me, it's a safety issue. It's the same reason why my engine has a disconnect in addition to the keyswitch. When I'm working on the engine, I want to be able to positively ensure that it is stone cold dead.
NarcisSisyphusRankin@reddit
That's utterly ridiculous. Every wire on your boat needs to have circuit protection, either a breaker or a fuse.
The only possible exception to that is the starter motor.
I'm an ABYC Certified Marine Electrician.
whyrumalwaysgone@reddit
Its fused
jacky4566@reddit
I wasnt trying to imply that it should be unprotected. Just that it doesn't need to be switch on the panel.
fgorina@reddit
I would like a seitch for the instruments although now I try to have switches at every instrument (I like to let anemomèter and remote access on to
Control wind from home) but I sureny want an independent switch for the autopilot pump.
millijuna@reddit
I just turn my instrument heads off at the instrument, but like you I leave all my sensors running. I check in on my wind sensor remotely, along with bilge levels, battery status, and so forth remotely. I can also run my bilge pump remotely if I so choose.
millijuna@reddit
I would definitely want my N2K network on its own breaker, if only so I can reset it if there is some N2K voodoo going on (it happens occasionally).
Unlike some, I leave my electronics (including N2K bus, AIS Transponder, and StarLink) on 24/7, even when I'm not aboard. It lets me keep an eye on things remotely, including the weather, and I also feed my AIS data into MarineTraffic and VesselFinder.
But each of those items (N2K, AIS, and StarLink) has its own breaker.
diekthx-@reddit
It depends on your AP’s power draw, but I set mine so that the switch just powers a contactor, and the AP is powered through that on a fused home run to the main bus. But mine needs a 30A fuse, and yours may not.
alex1033@reddit
Usually the instruments including the autohelm control screen consume very little power. The autohelm drive takes more, but not in standby mode. So, having them on a switch makes sense only if you really need to save power for something else, e.g., nav lights. But ofc it's nice to have that option.
bill9896@reddit
Certainly the N2K backbone should have a switch. Either now or in the future you might have instruments that are bus powered, and you’ll want the ability to turn them off. The autopilot pump in less of an issue. It is separately switched by the autopilot controller, so has zero parasitic load when the controller is off, and I can not think of any reason you would want power to the pump when the controller was off, or the ability to power down the pump when the controller was running. That said, normal practice would be to have the autopilot drive and controller be powered off of the same switch. That makes maintenance easier since you power off everything and don’t have to worry about hot wires as you are fussing around.
If there is room on an existing panel to add these switches, that is certainly the elegant way to do it. If not, a simple switch and an inline fuse is certainly adequate. It then becomes a matter of esthetics.
Installing the N2K power without a switch doesn’t look good for knowledge or the diligence of the installers. It is not required, but is a very good idea, to have a stabilized power source for the N2K bus. While some things you might connect will be tolerant of voltage fluctuations others will not. This is especially true if the power source you are connecting to has other things connected to it that are intermittent, and high current. Like engine starter, bow thruster, windlass…
visibl3ghost@reddit
Yes, this is wromg. If you have an existing breaker panel then they should have absolutely installed these on appropriately sized breakers. Alternatively. These should have been installed on a spare fuse on your fuse panel.
Hooking these items up directly to a battery bus and using an inline fuse is just lazy and unprofessional, not to mention wrong. An installer should know how not to add an unswitchable, always on parasitic load.
Also, an installer should know to use marine crimp and shrink terminals, and marine (tinned) wire. It's not clear if they did that here.
Also, the crappy "PWR" wire label is hilarious. What is this supposed to mean to someone troubleshooting this system in 5 years. Power to what? If you're going to label your wiring (which an installer should be doing) then they can be more specific. The label could, for example, read (NMEA +12V) or (AP +12V).
While the NMEA network doesn't pull much, the AP may be enough of a constant draw to pull down your battery significantly during a period of anchoring/chilling at a sandbar.
Unfortunately, most installation manuals just tell you what size fuse is needed on their wiring, and they don't mention these other things, so depending on how much of an a-hole/idiot your installer is it may be difficult to get them to do all of the above.
thehorselessjockey@reddit (OP)
To be fair it is labeled N2K PWR. Still feeling this was lazy due to the fact they installed it onto the bus that was behind the breaker panel.
Full-Photo5829@reddit
You are not crazy. The backbone does not need to be powered except when you're actually using it. It should be run from a breaker on the 12V panel.