A couple of years ago I installed a rudimentary NMEA backbone on Morgan’s Cloud to support the new B&G Zeus³ plotter we bought to replace our venerable, but sadly departed to the electronics graveyard, Northstar GPS, and to back up our primary navigation system.
This year I have added to that network to support a replacement autopilot brain.
When I’m doing this kind of thing, my scepticism about the quality of most marine electronics tends to kick in, and never more than when I looked at the chintzy NMEA 2000 cabling included with the plotter, autopilot and related gear.
And I was particularly concerned given the potential for NMEA 2000 networks to completely shut down and take every piece of connected electronics with them. This is the great weakness of daisy-chain networks of any type, and doubly scary on a boat where one drop of water in the wrong place, a chafed cable, or a corroded contact can bring the whole thing down. See Further Reading for more on this.
Thanks, John. I’ve just purchased a 9 inch B&G Vulcan for my pilothouse, “liberating” the previous 7 inch Vulcan for use at the outside helm, and I was looking for a more robust way to connect it all together. This is helpful and, I agree, well worth the fairly trivial extra cost.
Hi Marc,
You will love the Maretron cables.
I like the use of that multiport box, too. I’m already in the market for one of those Actisense 0183-2000 converter gadgets, so I can feed compass, GPS and AIS data to my Furuno radar display in a more “tidy” fashion. More jobs over the winter!
Lest anyone be tempted to buy shielded cable by the 100m spool and terminate it themselves with field-assembled connectors…. I assure you, no matter how good you think you are at cabling, you will NOT match the reliability of the Maretron factory-made cable assemblies for anywhere near as cheap as the Maretron price.
Totally agree. Like John I started life as an electronics techie, so I used to be very familiar with soldering, and to be blunt most people who aren’t doing it all the time are pretty bad at it.
Hi Philip,
So true, and trying to troubleshoot a bad solder joint on a daisy chain network like NMEA 2000 is a special kind of hell!
Hi Matt,
So true! Even though I have professional electronics technician training and a lot of experence I avoid making up my own cables whenever there’s an alternative.
Agreed, although pre-assembled isn’t always an option if the cable needs to run through a tight space the connector won’t fit through.
I just upgraded my Autopilot to a new Raymarine Unit and was at my wits end trying to connect their Seatalk NG backbone connectors together. It was straight forward enough, align the two arrows, push together and twist the collar, but could I get the collar to engage with the pin by twisting! I actually split one of the plastic collars thinking it must be a tight O ring by firmly pushing the parts together and twisting the collar. I will be expanding the network when I replace the plotter, so a timely article, again! And the price of extension kits ….
Hi John and all,
Yes, I agree. Maretron is a quality outfit.
When I re-instrumented back in 2012 I ripped out all the old data wiring and started anew with Maretron and I do not believe I have had a hiccup since. I found their N2K designer really helpful but posed (for me) a steep learning curve (it is likely more user friendly now). I also shied away from the more complicated, but very tempting, Maretron offerings (such as a weather station at the top of the mast), but friends have given good reports of this piece of kit. Mine was a Furuno-based instrument display/brains system, which, in hindsight, used some of the same senders/sensors that Maretron uses.
And, what Matt says is entirely accurate: buy the pre-made cables. That said, I used a handful of the field-use terminals (when cables had to be snaked through small passages) and they have stood the test of time. Having a couple of right angle terminals really helped a couple of times.
Oh, and as I remember, I ordered straight from Maretron. And agree: they were very helpful.
My best, Dick Stevenson, s/v Alchemy
Hi Alastair,
Glad it was useful. You will be amazed by how much more easily the Maretron connectors go together. I was.
Just for the sake of curiosity, the Maretron N2K media looks very much the marine equivalent of the the original industrial DeviceNet media specification.
https://www.rockwellautomation.com/en-au/products/hardware/allen-bradley/connection-devices/network-media/devicenet/1485-round-media.html#
Maretron look like they’re sourcing a very similar product, just with different colours. Maybe even from the same factory I’d hazard a guess.
DeviceNet is an open protocol network stack, targeted at industrial applications, that originally came from within Rockwell Automation in the early 90’s. The physical layer, like N2K, is based on the original Bosch CAN chip now used in tens of millions of vehicles worldwide. The underlying tech is actually exceedingly reliable; the problems arise when in industrial or marine settings the end user puts together networks without reading the manual thoroughly.
In my experience 99% or more of problems that customers encountered were due to cable problems, or failing to observe to design rules on terminators, dropline lengths and power supplies voltage drops.
Hi Philip,
That’s interesting. And yes, I agree that most network problems are self inflicted (same with Ethernet). That was why I covered Maretron’s design software and their willingness to answer questions.
That wouldn’t surprise me. NMEA 2000 is essentially just some marine-specific messages added on to the widespread automotive CAN standard SAE J1939. Its cabling and connector standard *is* the DeviceNet spec, copied word for word.
Also, I still have not forgiven Raymarine & Simrad for their utterly moronic, brain-dead marketing decisions to slap their own proprietary connectors on what is still supposed to be NMEA2000 with a DeviceNet physical layer underneath.
John
In short, I agree. I’ve used and been a fan of Maretron almost since inception, I specify it on virtually every new build and refit project. It is relatively easy to design and install, and program and its ability to accept generic non-Maretron devices is also very attractive (I can select my own networked smoke/CO detectors, for instance, one that includes the features I wanted). I’ve used it on vessels from 26 to 100 feet. Yes, it really is that good.
Maretron is owned by Airpax (they were a 50% owner for years and last year completed the merger), manufacturer of circuit breakers and other electrical equipment. When this merger occurred many expressed concern; While I am often concerned about such acquisitions in the marine industry, as they frequently spell lower quality and poorer service, in this case I wasn’t worried as I’ve known Airpax to be a great product and believed this would only enhance the Maretron brand.
Hi Steve,
Great to have your confirmation on that, particularly given that you have infinitely more experience with this stuff than I do, thanks.
Also good to hear that the ownership change won’t be an issue. I can remember being concerned when Simrad was bought by a larger company and everything I hear says I was right to be.
And I can give you half a dozen other examples of marine industry acquisitions that resulted in lower quality products and support, it’s usually bad news, this one seems to be an exception. I wrote an editorial about it a few years ago https://stevedmarineconsulting.com/photo-essay-december-2015/?upm_export=print
If you only have a few units to link this is an equally good quality option – truly plug and play (for up to 8 connections) in my experience: https://actisense.com/products/a2k-sbn-2-nmea-2000-connector/
They also do a 4 connection unit.
Hi Paul,
That does look good for a small setup although it is worth keeping in mind that if the network expands at a later date one will need to do complete network redesign and rebuild to convert to the more normal backbone and drop configuration. It’s also worth considering that using that box will result in longer cable runs than backbone and drop. Of course none of this is a problem if we are absolutely sure the network will not expand in the future.
John, you mention the unique tees at which supplemental power can be added. I have a NMEA 2000 system with a fair number (15+) of devices on it. Sometimes when I lean on a power-hogging piece of equipment like an electric winch or bowthruster, I get a momentary “low CAN BUS voltage” alarm, even though my battery bank SOC may be very high at the time. I wonder if adding in some power at a second location in the backbone might ameliorate this? (I note that, since putting into practice your excellent tips on backing boats that don’t steer in reverse, I’m leaning on the bowthruster less and less. Ha.)
Hi Craig,
No, that won’t fix it. This probably means that the cabling from your battery bank to the main buss bars is too light or there is a poor connection somewhere. I’m also guessing you have a 12 volt system? If so the best fix is to go up to 24 volts on any boat with that level of high current loads, but that’s a horror and expensive too.
The other possibility is that your battery bank is not big enough, or the batteries themselves are ageing so that the internal resistance is rising, due to sulfation, to the point that the voltage drops a bunch when the current increases. Ohms law: https://www.morganscloud.com/2016/05/24/one-simple-law-that-makes-electrical-systems-easy-to-understand/
The first thing to check is the cable sizes. Blueseas has a great app to calculate the correct size: http://circuitwizard.bluesea.com
If the cables meet spec for 3% drop, next try cleaning every single connection with fine emery cloth.
In addition to all of John’s suggestions above (which I’d try first as a matter of good practise), I would also run my N2K network power from an Isolated DC/DC converter supply.
Victron Orion-Tr 12/12-18 (220W) would be a good choice. This way it really doesn’t matter what happens on the battery side, the output side to the network power will be tightly regulated.
Hi Philip,
That’s a very good idea that I had not thought of and would probably be good practice for any larger and/or mission critical NMEA 2000 system.
That said, and as you say, it’s important to be clear that this does not solve the base problem that Craig is experiencing, particularly since if left unfixed these voltage drops could have other effects like maybe the autopilot kicking out at a bad moment.
Thanks, John and Philip. I will check cable sizes and connections. The builder (Valiant) seems reputable and doesn’t appear to have skimped on too much, so with luck it is connections or total bank size. I have only 4 Lifeline batteries (brand new). That just might not be enough. (And I like the converter idea, though I do understand the risk of treating a symptom rather than The disease.).
Hi Craig,
I have a Valiant (V-42-128) and I suspect the wire gauge sizes are more than adequate and I suspect your batteries are group 31s which, in new AGM Lifeline, should be way more than adequate to power a complex instrument package (~~~400 Ah), even when a powered winch demands current. I also have a fairly complex N2K Maretron based set-up and have for 12 years or more, but do not have the big load items you mention.
If I remember correctly, Maretron has a diagnostics capability that they sell which might determine any glitches in the N2K system that would contribute to the warning you receive.
Those big current items are an amp-sink, but the bow thruster, I would guess, is only used when the engine is on.
I would lean strongly against the DC to DC converter until you find the problem. There is something going on which can only cause problems in the future.
When trouble-shooting, I always ask myself: “What just happened?” In your case you got new batteries and I would start there and look at any connection that was touched. Was the N2K package working well before the new batteries?
It could be as simple as there being a stainless-steel washer used on one of the battery terminals that is strangling the current when big loads occur.
I would start at the batteries and take apart, inspect, clean, goop with copper-kote or the like and reassemble every connection from battery to where the N2K picks up its power. Don’t forget the negative side. Document every terminal wire as you go along. And, yes, the positive distribution posts are a pain to get to (at least on my boat) but need to be worked on also.
Let us know what you find.
My best, Dick Stevenson, s/v Alchemy
Hi Dick,
Yes, I would agree to really dig in and look for a bad connection.
That said, I would not recommend getting involved in diagnosing the network itself, since I think that’s a red herring . As I understand it this only happens when high loads hit, so I think it’s pretty clearly a cabling or bad contact problem.
Hi Craig,
I agree Valiant are a good builder, but I would still check the wire sizes against the app I linked to. The issue may be that back when your boat was built these kinds of very high loads were comparatively rare so the builder may easily have used their standard wire sizes without realizing that the game had changed.
I’m planning an article on this very issue, inspired by the huge DC loads people are taking on by changing to electric cooking though an inverter.
Thanks John, Philip and Dick. I will go into detective mode. One further interesting fact that occurs to me after thinking through the helpful advice: My bow thruster is not wired through the DC panel. It is 24v and its switch activates a solenoid that puts the two house banks in series. That circuit has its own 400a fuses and big Guest switches. if the low voltage is occurring with two machines (bow thruster and winch) that do not share any cabling on the positive side, must any high-resistance connection causing unacceptable voltage drop be on the negative side? Just a thought. I’ll investigate and report back. Thanks again.
Hi Craig,
That sounds logical.
Hi John,
Is there any reliable solution for NMEA2000 cabling when one has no choice but to cut a terminator? I had to do this for the cable from my wind instruments after spending 3 hours on a bosun’s chair trying unsuccessfully to pull the cable and its fat terminator through the mast. At the moment I’ve just connected the 5 wires of the wind instrument cable to the boat’s NMEA2000 network via an open bus bar. I intend to replace it ASAP – but with what? I see Matt’s post about unreliable on-site splicing so what else to do?
Timely article! Thanks.
Hi Robert,
Maretron makes field-attachable connectors/ terminals for just that situation. They come male and female and in right angle which is very hand in some situations. I have used them in protected areas for 10+ years without a problem.
You might ask Maretron about a top of the mast connector and how to weather proof it. Ham radio people put me on to coax seal a while back that is marvelous for this function.
My best, Dick Stevenson, s/v Alchemy
Thanks Dick. I’m looking at the Maretron field connectable connectors now :>)
Rob
Another vote for Coax-Seal, which has also been recommeded to me by amateur radio pros for our upcoming SSB installation. (http://coaxseal.com/).
Hi Robert,
If it were me I would come at this differently and not take NMEA 2000 to the top of the mast particularly with field connections. Way too many chances for intermittent problems thought the whole network due to the length of the drop to the top of the mast and the chances of chafe etc. The way I look at this is mast head instruments are a luxury, but installing them as NMEA 2000 increases the risks to mission critical stuff like radar, GPS and autopilot on the rest of the network.
I believe you can get self powered wind instrument packages that then have a NMEA 2000 converter box at the bottom of the mast, which would be a better option. If there is no way to get away from running NMEA 2000 to the top in your configuration, then I would strongly advise you make that a separate backbone with it’s own power T and connect that to the rest of the network with a bridge: https://www.maretron.com/products/nbe100.php
That way, if you have issues in the mast, all you will lose is wind data without potentially bringing down the entire network:
https://www.morganscloud.com/2013/05/26/nmea-2000missing-the-obvious/
Hi Robert,
Here’s an example wind package which does not rely on sending NMEA 2000 to the top of the mast: https://www.bandg.com/en-gb/bg/type/instrument-sensors-and-transducers/wind-sensors/ws310-wind-pack20m-cblinterface/#prl_key_features
What they are doing, I’m pretty sure, is using NMEA 0183 to go to the masthead (much more reliable over long distances) and then converting it to NMEA 2000 at the bottom.
I second everything that’s been said here, but I want to emphasize 2 points most people miss. Buss bar/barrier strips (and ring terminals) outside in the salt and spray are notoriously unreliable. Pretty much any watertight connector is a vast improvement. If you’re in a tight spot (e.g. the masthead) I actually like heat-shrink crimp-on butt splices filled with dielectric grease. Just be sure you get the right size so the heat-shrink will seal on the cable. leave plenty of service loop and if you have to replace something, just cut it off.
The second and more insidious problem is that N2000 is a CAN-bus and as such is not designed for a star topology. In general that means that without an active repeater you shouldn’t run a line off a backbone t-piece for much distance at all, let alone up a mast. In general you should keep T-offs less than 1m from the backbone, though in the real world you’ll get away with longer. The suggestion of adding a separate bus up the mast (and the associated terminators on each end) does solve that problem. Otherwise, you really should be running the backbone up and then back down, to not have a very long spur off it.
CAN busses with lots of t-offs that go long ways will work, but will also fail randomly and intermittently, as this effectively ruins the termination. Different silicon in different devices will all behave slightly differently and with so many things talkign on the bus normally it is very difficult problem to diagnose without an oscilloscope.
Hi James,
That all makes sense to me.
One small quibble is that while I agree that keeping the drops as short as possible is a good idea, I think I’m right in saying that NMEA 2000 spec is for a maximum drop length of 6 meters, so aiming for 1 meter could force one into making the backbone unnecessarily long and convoluted as well as routing it in undesirable places.
To me a good compromise max drop length would be 4 meters, which on most boats allows a nice clean backbone along the side of the boat well out of the bilge with drops able to reach all equipment.
I also have the Maretron multiport boxes installed on the vessel and can attest that this is a really good piece of kit that works and that simplifies cabling. Prior to installing that, the original installer of the gear had just daisy chained some components. Apparently not good, nor recommendable practice.
It is possible to find these components at a favourable price on ebay.
Iris is in principle a 24v vessel. To feed the 12v electronics I have cabled in an isolated 24-12v DC/DC converter from Mastervolt. In principle is might also be an advantage to have a 24-24v DC converter to feed sensitive 24v. I nearly fried my 24v electronics on a longer passage when the regulator of the wind generator called it a day – and the generator started to produce voltages above 30V.
https://www.mastervolt.com/products/dc-master/
Greetings.
My boats have many high-quality Maretron (and Actisense) cables and connectors, as well as lots of industrial DeviceNet gear I’ve bought on EBay. But you have to be careful because the NMEA 2000 cable and connector standard is a subset of DeviceNet. I tried to explain here:
https://panbo.com/quality-nmea-2000-cabling-the-ebay-way/
Hi Ben,
Thanks for the warning. Scanning your article said to me that the rest of us should just stick to Maretron and be done with it!
“A clever power T that can be connected to two power sources one primary and the other backup”
Not really. What the Maretron Micro/Mid PowerTap does is supply separate 12v feeds to each leg of the network, thus reducing possible voltage drops at the far ends. (And Maretron Mid size cables can also reduce voltage drops.)
It’s also female on both backbone sides so that exposed male pins are never hot even if you break into a network while it’s powered up.
https://www.maretron.com/products/pdf/Micro%20Cords%20Tee%20Pwr%20Datasheets.pdf
Hi Ben,
Good to hear from you. And thanks for the correction. Just looked at the manual for the T and clearly I got that wrong. I will fix it.
Thanks, Ben. I’m about to order a whack of Maretron cabling and connectors and an Actisense device to allow my radar to display AIS and bearing data from the NMEA 2000 backbone. It’s the quality of the commenters as much as John and Phylis’s excellent writing that keeps me coming back. Oh, hey…I can do italics now! Nice.
I just read an article on Seabits about a new terminator from Actisense. Steve has mostly used Maretron the last 10 years, and still does, but this terminator seems like a smart piece. Firstly it’s universal, meaning it has a female plug in one end and a male connector at the other end. The unused one is covered by an included plug. The terminator also has an LED that lights up green when there is more than 9 Volts present and the network is terminated correctly. Pretty useful.
The article: https://seabits.com/actisense-smart-terminators/
Hi Stein,
That does seem like a good piece of kit, particularly the lights since I gather that low voltage is a big contributor the the flaky network stories we hear.
Stein and/or John, I hope this question isn’t off-topic, but I intend to order the Actisense smart terminator along with a range of these Maretron NMEA 2000 cables and T-fittings. I am slightly confused (and a call to the local chandlery didn’t clarify matters) with the best way to hook my Furuno 1815 standalone radar into my NMEA 2000 network with a NMEA 0183-NMEA 2000 “converter”, such as the Actisense NGW-1-ISO device (https://actisense.com/products/nmea-2000-gateway-ngw-1/).
My goal is to feed data from my heading sensor and AIS source already “seen” on my plotter to my radar to enhance its functions and, in essence, to use it as a second plotter, minus the chart overlay. I feel because I don’t need to take any 0183 data from the radar, but simply to feed existing NMEA 2000 strings to the radar, the “standard” NGW-1 is all I need, not the NGW-1-ISO-AIS device, which converts 0183 output to 2000 data. While I could use some sort of terminal block setup to transfer between my base VHF AIS target data fully in 0183, this solution seems tidier and more comprehensive. I realize I will have to fiddle with the baud rates on the radar, but that’s not an issue for me. Thank you. This is a very helpful thread about the pitfalls of too much ambition in the networking department.
Hi Marc,
I really don’t want to get into this kind of personalized electronics and networking support. That way lies madness for me. That said, I have just done exactly what you are contemplating and will be writing about it, and the challenges of hybrid networks generally, which will help.
That said, I can give you a hint: you need to take your information from a careful read of the Actisense web site, not calling the Binnacle. For example said careful read will yield that the only difference between the standard NMEA 0183-NMEA 2000 converter and the AIS one is the configuration that they are preloaded with at the factory, the hardware is identical and you can reconfigure it to be either in the field.
All the manuals are available at Actisense so you can figure it all out before you spend anything. I would also recommend drawing yourself a diagram of how things will work, to clarify your thinking, prior to getting started or even buying anything.
Understood. Thank you, John.
I agree regarding the Maretron quality and like using myself. However, I find the field connectors to be horribly tedious. Also, depending on where you are assembling them, (often in a tight dark space with is why your are using it in the first place) the connection with the 5mm lengths of wire and tiny screws can be suspect. They are better suited to assembly on a lab bench.
I have really like the Phoenix solution. Phoenix is an industry leader with quality stuff. Here is an example but there are other options depending on needs: https://www.mouser.com/ProductDetail/Phoenix-Contact/1424649?qs=%2Fha2pyFaduifh0fNZC5%2FGMiM20GJwMUnVH9oeO7uiioc63wQ4PiY1A%3D%3D
Illustrating the ease of assembly: https://www.youtube.com/watch?v=XbxvOmVIodM
Hi George,
That’s an interesting alternative. That said, I’m always a bit nervous about using generic stuff that was not designed for boats. Of course this may be fine, but others thinking of using it should check out how these connectors will do in the marine environment. The good thing about Maretron is that marine is most of what they do.
As to field connecting, my general recommendation for us amateurs, even ones like me who were once professionals, is don’t. Just too much potential for problems.
Hi all,
I believe that all that has been said about the benefit of manufactured connectors and terminals is accurate, but I am also clear that there are many of us, when re-instrumenting their boat, who will be compelled to use a few field connectors where snaking the terminals/connectors through tight spaces proves unworkable.
And, yes, doing the terminals is fiddly work, but it is straightforward, basically un-skilled work, where going slowly and carefully and inspecting makes up for having no great skill and experience. At least, that was my experience, when I re-instrumented Alchemy 10+ years ago and none of the ½ dozen or so field installed terminals have given me any trouble.
A couple of suggestions:
Practice where comfortable and there is good light.
Leave a sizable service loop so you are not working right against the instrument you are connecting to: its length will allow much easier work and will allow for doing the work in a more comfortable position.
Ensure the wire is well secured near the connection so there is no chance of vibration.
Document/label every field connector on the schematic you are planning to make of the whole re-instrument array.
My best, Dick Stevenson, s/v Alchemy
Hi Dick,
That all makes sense. Good to have the fill as I have not yet had to do a field connection on NMEA 2000. I did reread the the Maretron instructions and it certainly looks doable to me, albeit fiddly. Anyway, as you say, care and attention makes all the difference with this kind of thing.
And good point on the importance of documenting what we did.