Sunday, January 13, 2013

More on Chart Plotters

Editor's Note:  Since writing this post, Simrad has released support for Jeppesen C-Map charts in addition to the previously supported Navionics charts.  With this addition, Simrad chart availability matches Furuno, and allows for use of the same charts with Coastal Explorer.

I've been researching chart plotters as part of trying to finalize what I want to use.   As discussed earlier, the various vendors offer different chart sources, so I decided to look into it from that angle.  I made a list of the more out-of-the-way places where we envision going, then checked to see what was available from each vendor.  You can see the chart below, but here's my take-away:
  1. Garmin is out.  Lots of missing coverage, including no coverage of Japan.  What's with that?
  2. Furuno has the best and most diverse coverage because they offer both Navionics and C-Map.  The only thing they do not have (nobody else does either) is the antarctic peninsula, and it would probably be unwise for us to go there in our boat anyway.
  3. Simrad (via Navionics) also has excellent coverage.  The only place they fall short relative to Furuno is the stretch of Russia between the Aleutian Island chain and Japan.
Given this, I'm comfortable with the Simrad coverage, and can fill in any gaps using C-map (or other) charts on Coastal Explorer or MaxSea running on a PC.


Region
Simrad
std Navionics
Furuno proprietary
version of
Navionic, C-Map, HO-57
Garmin proprietary
Coastal Explorer
std navionics, c-map, HO57, HO-63
Aleutian chain
Y
Y
Y
Y
Pac rim
Y
Y
Y
Y
Greenland
Y
Y
Y
Y
NW Passage
Y
Y
N
Y
Svalbard
Y
Y
Y
Y
Falklands
Y
Y
Y
Y
S. Georgia
Y
Y
N
Y
Antarctic peninsula
N
N
N
?
Russia, Aleutians
to Japan
N
Y
N
Y
Japan
Y
Y
N
Y



I've also spent some more time looking at Furuno's most recent announcements to try and get a sense of where they are going.  Here's my reading of the tea leaves:
  • The NavNet3D products are not receiving any further development.  This includes the MFD8, MFD12, and Black Box.  These products are about 5 years old now, and the last software update was 2 years ago.  An update has been "very close" for at least 6 months, but still nothing. Other than bug fixes, I don't expect to see much new.
  • In the mean time, the TZ line has been released with touch screens.  They also just announced a Black Box version.  All new features appear to be coming out in the TZ line which is part of why I think they have stopped evolving the NN3D products.
  • It strikes me that they have focused on chasing Raymarine and Garmin with all this 3D and touch screen stuff, neither of which are of any interest to me, and in doing so, left their core historic following hanging out in the breeze.  The NN3D is too flaky, and I'm seeing even more complaints about the TZ.
It's sad, but I think they really screwed the pooch.  It will be interesting to see what comes out over the next few months and maybe there will be a surprise in there somewhere.  But at this point I have reshuffled the stack and Simrad is the plan of record with Furuno next, and Garmin pushed below the cut line.

So a few days ago I came across a really good deal on a Simrad NSO black box with a 17" monitor and bought it to try out.   I plan to bench test it for a while, see how it links with MaxSea and Coastal Explorer, try the charts, and generally see how it works.  Then I'll take it over to my boat (which in on the hard) and plug it in to see how it works with everything else.  The new N60 won't be set up the same as the Grand Banks, but it will be pretty close so I should be able to get a pretty good feel for things.

NMEA 2000, aka N2K

What the heck is that?  NMEA stands for the National Marine Electronics Association, and NMEA 2000, or N2K for short, is a standard way to interconnect marine electronics devices.  It's the replacement for the NMEA 0183 standard which is nothing more than computer serial ports.

The big problem with NMEA 0183 is that computer serial connections are meant to be point to point.  It's two things, and only two things, talking to each other.  If you have 5 devices, all of which need to talk to more than one other device, you get an explosion of connections.  This has been mitigated in 0183 by recognizing that the real restriction is that only one person can be talking, but multiple people can be listening.  The result is that boat electronics build using 0183 area a rats nest of Talkers and Listeners.  When the GPS talks, the VHF, Chart Plotter, and Auto Pilot need to Listen.  When the Chart Plotter talks, the Auto Pilot needs to listen, but not on the same port as the GPS (remember, only one talker is allowed on each connection).  When the AIS talks, the Chart Plotter needs to listen, but again on a different port, and that port needs to run at a different speed than everything else.  When the rate compass talks the Chart Plotter, Radar, and Auto Pilot need to listen.  It gets complicated very fast, and for someone climbing onto a boat to diagnose a problem, it can be a real head scratcher.  It was fine when boats only had a few instruments, but with a lot of instruments is gets ugly very fast.

NMEA 2000 fixes this by creating a communications "bus" which is nothing more than a single set of wires running through the boat that everything taps into.  Smarts in the attached devices allows them to ensure only one person speaks at a time,  just like a computer network.  Devices can be added and removed by connecting them to the bus with T connectors.  Any device can talk on the bus, and any device that cares can listen.  The rats nest of wires and ports disappears.  Another cool side effect is that failing over from one device to another is trivial, and in most cases happens automatically.  For example, let's say you have a satellite compass which gives position plus a highly accurate heading.  And you also have a second GPS, plus a rate compass.  You can set up the auto pilot to prefer the most accurate device, in this case the satellite compass, but if it fails, to use the GPS and rate compass instead.  This fail over happens automatically.  If you were using 0183, you would have to rewire things, or set things up with selector switches to pick which device to use and manually switch in the event of a failure.

But N2K is not without it's challenges.  Like any new standard, it takes a while for different manufacturers to support it, there can be different interpretations of the standard which cause problems, and bugs in products.  There also is a provision of vendor-specific messages which on one hand is essential for things like device configuration and firmware updates, but also allows vendors to venture off and create their own thing rather than work with others.  That said, I don't see any way in which it's worse than 0183, and lots of ways that it's better.  Other complaints argue that NMEA should have just used Ethernet, or that it should have been a switched topology rather than a bus.  Although interesting bar room banter for nerds (like me) those arguments have no practical relevance for someone outfitting a boat.  N2K is what it is, and you can use it, or use something else.  I just argue that of the choices available to us, it's the best one.

Another argument against N2K has to do with reliability, the ability of a single failure of just the right type to lock out the whole bus, and compatibility problems between devices.  There is truth to all of these arguments, but every year that goes by they diminish.  All the same arguments were made about computer networks early on, and we seem to have gotten by it just fine.  I think the trick to success with N2K is to not push its limits.  Don't put too many devices on, test things carefully, and consider multiple, separate buses for larger configurations.  For example, put your navigation equipment on one bus, and put all the house lighting controls on a separate bus.

For a little background, N2K is based on the CAN bus which was create for use in cars.  Cars have a whole pile of computers on board, and one or more CAN buses are used to connect them together.  Another standard at work is J1939, also based on the CAN bus, and used for diesel engine control in current model diesel engines.  NMEA borrowed these existing and well proven technologies, and added in their own marine-specific messages to make it work in boats.  It's akin to coming up with your own written language, but using standard envelopes, addresses, and the post office for all the delivery.

To me, the benefits of N2K clearly outweigh the negatives.  Others may arrive at different conclusions, and that's fine too.  We implemented N2K on our current boat, and attached to it are:

- Chart Plotter
- GPS
- Weather station
- All the autopilot components including 2 control panels, the course computer, rate compass, rudder feedback, and remote control.
- 0183 converter to connect the VHFs
- Water tank level sensors
- Trim tab position sensors
- 2 Mareteron displays that can display a myriad of different information

On the new boat we plan to do the same, plus add a small number of additional devices.  We will be doubling up the auto pilot devices for redundancy, adding a couple more tank sensors for fuel levels, and adding temperature sensors in the equipment spaces.  It's not a lot, and I'm confident that we will have a reliable system.

Sunday, January 6, 2013

Fuel Management

You'd think this would be a simple topic, right?  Fill your tanks and off you go, keeping an eye on the gauges until it's time to fill up again, right?  Unfortunately it gets a little more complicated on the N60 for a variety of reasons.

First, there are 4 tanks to keep track of.  Having multiple tanks is not unusual in a boat, but I think 4 is on the higher side.  Our current boat, for comparison, has 2 tanks.  The 4 tanks on the N60 are arranged as two 1000 gal main tanks, and two 170 auxiliary tanks for a total of 2340 gal.

Oh, but there is a 5th tank called a day-tank, or perhaps more accurately a run-tank.  The engine and generator draw their fuel from this run-tank, and the run-tank gets filled from any combination of the 4 storage tanks.  Why, you might ask?  Reliability is the answer.  Contaminated fuel is reported to be the most common cause of engine failure, and the idea behind the run-tank is to create a known-good supply of fuel, even if the storage tanks have questionable fuel in them.

When everything is working properly, the run-tank is kept full via gravity feed from any one of the storage tanks.  But if your stored fuel is in questions, you can instead route all fuel through a filter on its way to the run-tank, thereby filtering out any crud  before it gets in the run-tank.  Ideally you would run all fuel through the filter before it goes in the run-tank, thereby ensuring good fuel at all times, but running through the filter is a manual process.  You have to operate a pump to move fuel through the filter.  That means that every few hours you need to go down in the engine room and run the pump long enough to top up the run-tank.  And if you forget, you can easily run out of fuel and find yourself dead in the water.  It doesn't seem like a very good solution overall, but I haven't figured out how to best use it, or how to improve it.  Maybe in time I'll come up with something.

Going back to the 4 storage tanks, there is another complication.  As equipped from PAE, the smaller tanks have no gauge at all, and the big tanks have a site gauge that only reads the bottom 600 gal in each tank.  This strikes me as one of the most brain dead things I've encountered on this boat.  I don't understand how anyone can set out on an ocean crossing that's expected to consume 90% of the boats full fuel capacity, and not have tank gauges.  Yes, fuel burn is pretty predictable, and yes, you can calculate how much you will use.  But I'd want to be constantly cross checking the expected fuel level against the actual level.  A crossing calculated to consume 90% of the boats fuel means you have a 10% margin of error before you find yourself in very deep trouble.  Rough seas, opposing winds, opposing currents, and route variations to avoid weather will all impact fuel consumption and can pretty easily consume that 10% of margin.  The way the boat is equipped, you have no confirmation of how much fuel you have burned until you have used about half of it.  Only then do you get your first confirmation of how much fuel you actually have used, and how much you actually have left.  And only then can you compensate if external factors are negatively impacting your consumption.  I just don't get how a boat designed for ocean crossings can be equipped like this.  It's like crossing an ocean with your GPS turned off until your reach the half-way point, relying solely on dead reckoning until then.  At the midway point, you get to turn on your GPS for the first time to see where you actually are vs where you calculated you would be.  And the solution is so simple - just put gauges on the tanks like every other boat on the planet.  That's what we did, but why on earth it's not standard equipment is beyond me.  For all that PAE does right - and that's most everything - they have these odd blind spots here and there that leave me scratching my head.