Friday 22 October 2021

Electronic/Navigational Upgrade Project

NMEA – National Marine Electronics Association

“NMEA 0183 is a combined electrical and data specification for communication between marine electronic devices” (Wikipedia)
The issue:

Wireless sports a complete B&G H1000 electronics/navigation suite, including wind, speed, depth, auto-pilot and chartplotter (CND).  This system was one of the best available 15 years ago, but has since been replaced and out-done.  
Now while my system is working without any issues, I have decided to make some improvement to my onboard electronics, and it has been an uphill battle sorting through the world of marine electronics on my own.
This post (as it continues to develop), should serve to assist anyone with an outdated chartplotter/autopilot controller who wants to utilise good existing instruments with a state of the art software and hardware package.
It is the story of 2 generations of upgrade.  Initially I simply wanted a GPS signal routed to my Icom VHF radio.  Then as my knowledge of what was possible grew, I decided to replace the CND unit and develop a DIY navigation system.

PROJECT #1 - GPS TO VHF


1.       Get a signal from my GPS antenna to the NMEA input at the back of my ICOM M602 VHF, in order to display a back-up GPS position, but also to enable my DSC (digital selective calling) and location-enabled distress signal.  The ICOM M60f2 allows you to broadcast a per-determined nature of distress and a GPS location to any vessels in the area with a DSC unit (the Coast Guard in NZ do not monitor DSC on VHF).  



The B&G H1000 system is comprised of a masthead unit which produces data on wind speed and direction (recently refurbished), a depth sounder, a knot meter, a compass sensor unit, and a GPS antenna.  These units collect data which is fed to the CND unit, which is also the chartplotter and ‘brain’ of the system.  The information is processed, and distributed to the analog wind display, the multi-function LCD display, the autopilot display, and the autopilot itself.  A tremendous amount of data is available to me while underway, from SOG to the location of the nearest fuel depot.
The B&G units all talk to each other over a B&G-only system called Fastnet2, but the language used is NMEA 0183.  My system incorporates a special unit called a “Universal Interface”, which is like the “back-door” to the system.  This interface allows non B&G components to send and receive data (0183) to and from the H1000 system.

The Universal Interface



I can achieve my goal by using connection 1 (NMEA tx+)  behind the black cover.  This is the NMEA 0183 transmit connection from the overall system.  You can see the inside of the box here:


B&G Universal Interface


The wires you see in this photo are connected to the alarm, letting me know when I’m about to aground, when I’ve dragged anchor etc.

This option would limit my ability to eventually extract data from the system into a multiplexer, so to simply extract a GPS signal, there was a better way...and that is the CND 'black box' inside the electrical cabinet.

The CND (Chartplotter) PWR Box


CND PWR box internals


Solution

The best place to grab an GPS nmea 0183 signal is from the CND PWR box, which is where the chartplotter and GPS antenna come together to be energised, but also where they connect to each other.

Connect the GPS to the VHF at the CND PWR box.


The method for connecting the GPS to the VHF from the CND PWR box is to connect the blue wire from the CND/Chartplotter (NMEA OUT +) to the NMEAIN (+) pin on the back of the VHF.  Then you connect the black wire from the CND (NMEA OUT -) to the NMEAIN (-) pin on the VHF.  You can see both wires in the photo above.


NOTE!!!!!  This view is of the radio input, the cable is mirrored 





 


I have to take a minute to recognise a real guru on this subject, who really pointed me in the right direction.  Here is the secret:

John,

You can get GPS information from 2 places in your system.
  1. From the Blue (Data +) and Black (Data -) on the data / pwr cable that comes from the H1000 CND. (See page 46 of the attached manual). This is in the Junction box that you have identified and you can see the blue wire in the photo.
  2. From the H1000 Universal Interface terminal 1 (Data +) and Terminal 2 (Data -)
You should be able to use 1 of these outputs to feed the VHF radio and the other one to feed the MUX. I would use the UAI output to feed the multiplexer as the CND NMEA output does not supply wind information.
I hope this helps.
Best Regards
Craig McMillan
Advance Trident


UPDATE 31/7/13 -  I have completed goal #1, specifically the connection of the GPS antenna to the VHF radio and the input of my MMSI ID. I can now use DSC to alert other boaters should I have an emergency. 
Note the positon information now displayed on the VHF.

Thank you to the guys at Bay Marine Electronics in Tauranga, NZ.  They had the tools to find the data stream from the GPS and get it into the VHF.



Also worthy of note is the wiring color codes of the GPS antenna itself.





UPDATE 06/09/21


After painstaking research over a couple years...


Project 2 ~ New DIY Navigation System


This is my chosen path to replace the B&G CND unit, and fully-integrating my nmea 0183 data (from instruments) with a modern navigation package, AIS, and data from my charging system, engine, and onboard tanks.





NOTE: The B&G instruments receive power from the 4 way HUB and therefore do not rely on the CND unit for the small amount of volts they consume.  This comes from my DC breaker panel.  This allows me to delete the CND without compromising the instruments.


There are several components to research and source, so I will be detailing each step along the way.  Suffice to say I will be running Openplotter (opensource/FREE chartplotter software) on a Raspberry Pi 4, monitored on dual screens, one in the cabin and one at the helm.  

This solution will cost approximately $400USD in total, approximately 1/10 of the cost of a new B&G navigation system!!


Here's the heart of the system, a Raspberry Pi 4 Model B 2019 Quad Core 64 Bit WiFi Bluetooth (4GB):


And this is the cooling case for it:




The Raspberry Pi will display on a Eyoyo 2K Portable Monitor (mounted on the bulkhead in place of the current chartplotter).



To get data from the B&G instruments, GPS, and AIS to the RPi (computer), and then to send data back to the autopilot, I will need a multiplexer (MUX).  This device combines data coming in from different locations and at different speeds (baud rates) into one output (USB) that can be sent to the computer.

I selected a simple and affordable ($89 USD) unit manufactured in Australia by YakBitz


All of these parts are in transit....watch this space!!




17/9/21 Update!

Here is the Raspberry Pi4b hot from the courier, I am installing it in a cooled aluminum case.

These things are TINY! I imagine all of this will be easy to install discreetly. 

Update 22/9/21

Here's the recommended wiring between a B&G H1000 Universal Interface and the Yakmux unit, courtesy of Steve at YakBitz.








23/9/21 - Refined drawing showing power and data connections to existing equipment.







Update 25/9/21





The removal of the CND unit will leave a hole, needed a plate to cover and act as a base for the new monitor or equipment (placement tbc).  Still had some scrap teak.  Bit of cutting and shaping and oiling later...










Update 4/10/21

I've created a mounting board from marine plywood, and stuck it to the inside of the bulkhead (in the small locker starboard of the companionway) using 3M 5200.  This will allow me to screw down the required components.  To hold the panel in place for curing I used what I had at hand...a teaspoon.
Here is the Yakmux unit wired at one end.
I'm not sure yet how to open the jaws on the B&G UI end of things.





Update 6/10/21

I think the most challenging part of the project thus far has been running the wiring from the starboard instrument locker to the port side DC panel and navigation station.  A convoluted series of small openings and transition over the engine meant moving back and forth between the cabin and engine room.  I accomplished 3 objectives:

  1. Pull the GPS antenna cable back to the instrument locker from the black junction box (see older post above) in the DC panel cabinet.  The GPS cable will be wired directly to the Yaxmux.
  2. Pull a 7.5a cable from the instrument locker to the DC panel
  3. Pull a long USB cable from the Yakmux to the RPi at the nav table.

I then installed a terminal post in the instrument locker and confirmed 12v.  I wired the Yakmux to the B&G Universal Interface and the GPS antenna to the Yakmux.  Power and ground was applied to the Yakmux and antenna (the UI is powered along with the entire B&G system from the DC cabinet 4-way box).

The very small wires connecting the electronics were soldered or crimped as required, I'm not a huge fan of dealing with tiny wires tbh.  The Yakmux and GPS antenna draw milliamps.

Turned the system on and proceeded to look for a signal in Openplotter and Signal K (more on this later) and can confirm that data is being received!

Grainy photo...tiny wires

It's ALIVE!!!





7/10/21   SETBACK!!


My B&G 4 Way Hub has shorted out.  The 5amp circuit breaker tripped and I traced the problem to a corroded and broken pin on the B&G 4 way hub power connection.

The 4 way hub connects the instruments at the forward end of the boat with the instruments and equipment at the aft end of the boat.  It also introduces 12v DC power and grounding to the system through a cable that looks identical to the Fastnet2 cables, but has only 2 wires (Red + and Blue -).

It was at the power connection that the pin burned out and damaged both the hub and the power cable.

This takes down all of the instruments until I can find a replacement.  The system is discontinued so this might be a challenge. 

4 Way Hub




Fastnet2 connector


Update 8/10/21

I have located a new/old stock B&G H1000 4 Way Hub in Florida USA.  It is being couriered to NZ and should arrive in a week or 2.

As I have had no luck in finding the power cable (which is known as H1000HCP in the B&G literature), I am forced to create one.

Because the B&G chartplotter is now redundant, I have 2 useable Fastnet2 connectors at either end of the chartplotter's cable.  This cable, unlike the HCP power cable, has 4 wires with a shield rather than 2 wires.

I need to understand how the original power cable was wired if I am to replicate it with the spare cable.  I used my multimeter (ohms) to test continuity and found the following arrangement:



On the H1000HCP factory cable (PLEASE REFER TO THE DIAGRAM ON THE RIGHT, WHICH IS THE CABLE), the red wire (+) went to hole #1, and the blue wire (-) went to hole #5.

On the cable I have available to replace it, I have discovered that the red wire also goes to hole #1, and the metal shield goes to hole #5.  So this confirms that I need to apply 12v to the red wire, and connect the shield to my negative bus bar.



Image on left is the hub with male pins.  Image on the right is mirrored, and represent the female cable end.  The confirmation of the color of the 3 data cables come from Oppedijk.com


If you are struggling to understand the B&G Fastnet2 system, please spend some time on the following website:

Oppedijk.com


Update 15/10/21:

This arrived today all the way from Florida USA!  Might be the last one on the planet but it will put me back into action next week.

Black instead of yellow, but this is the real deal.


Update 21/10/21

I have finally got the system figured out, with all of my B&G H1000 instruments and the GPS antenna sending good data to the RPi4 and Openplotter.

This was not straightforward at all, and at first it didn't work.  I got some great help from Steve at Yakbitz.  He looked at the data that was coming in from the B&G network and immediately suspected a voltage drop imbalance.  Briefly, this is when the voltage drop on the Tx+ or Rx+ lines differ substantially from the drop on the Tx- or Rx- lines.  Voltage drop occurs in any wire over distance, so the distances must be the same.

Because I was connecting the Tx+ and Rx+ leads from the Yakmux to the B&G Universal Interface over a distance of about 100mm, and the Tx- and Rx- lines were being grounded all the way back to the DC panel (3meters), the differential voltage drop was enough to corrupt the data stream.  I grounded the negative data lines to the screen inside the UI and this instantly fixed the problem.

Data + lines (green and yellow) are the same length at the Data - lines (black)



I am tempted to start some sort of B&G H1000 help forum.  I really think this gear is top quality, but it is older and has been abandoned by B&G.  I have learned so much about how it works now I'm sure there are others faced with the prospect of replacing good gear when they want to upgrade a single component.

Next I will be refining and developing the system to include dashboards, a 2nd monitor at the helm, integrating the system into my autopilot, and creating a panel to display my battery data, tank, and engine data.  I will also re-connect my VHF to receive GPS data using an RS485 connector.

USB to NMEA cable

ICOM 602 wiring





Tuesday 19 October 2021

In Hot Water! ~ Adventures in Plumbing!

I've owned Wireless for nearly 10 years and I've never had hot water aboard.  The original Paloma califont was removed a long time ago.
I've begun the process of installing a new califont,  which will include installing the deck flue within a dorade box built years ago ( https://kiwi-nonsuch.blogspot.com/2013/08/dorade-box-construction-project.html); installing the califont, and plumbing water and gas.




25mm teak board to be re-sawn into 2 dorade tops.  A local mill have offered to do this with a very large bandsaw.  10-12mm isn't ideal, but the top is under no stress and this stuff is EXPENSIVE!

Update 13/7/21:

So the scale of this project is setting in, and it's expanding.  My original freshwater pump is beginning to play up, and I think it may be ready for replacement. I figure it's time to refresh the system entirely as some new buteline tubing is required for hot water.

I've purchased this tidy unit, it's a pre mounted pump and accumulator.   It even has wiring internal to the panel.   It comes with a strainer as well.  It's a Seaflo 42, which has a bypass valve to further reduce pulsing.









The old pump - possibly original

Space for new pump and accumulator


Done! Very tricky working under the head sink, little to no room to crimp fittings and wires.  Improvised with some flexible hose.




To Do:

1. Fabricate flue pipe to pass from cabinet above galley through the deck into dorade box. (DONE - I bought a 4" to 3" exhaust adaptor from an automotive supplier and applied couple coats of epoxy paint)
2. Remove patch, pass flue through hole to measure and cut.
3. Wrap flue with heat wrap, then secure in place with high temp epoxy putty.

Update 2/8/21:

So the outlet of the califont when mounted in place is offset 12mm to port of the hole in the deck.  This is a problem, because it is enough of a misalignment to prevent any solid flue from connecting the 2 points.  

I happened across a product that is used in the trucking industry.  It is a stainless steel flexible pipe that comes in a range of diameters.


This will allow me to position the califont roughly 300mm from the ceiling, and pass the flexi tubing through the deck, bending it as required to land in the reducer I have created for the outlet.  I'll position the dorade box, mark and cut the excess tubing off with a hacksaw.  I want as much protruding as possible, despite having a baffle in the dorade...obviously seawater getting in would be bad news.

I think standard exhaust header wrap or equivalent insulation will reduce the liklihood of anything getting too hot in the cabinet above the califont.  The temperature of the gases leaving the califont, and the corresponding heat in the flue is a matter of conjecture.  Clearly LPG is being burned, but the heating is sporadic and very short lived in an on-demand heater.  Even 10 seconds of running the hot water is unlikely to heat the flue to the point where it could cause a burn or a melted cup.  Still, wrapping the entire thing in insulation is cheap insurance.


4. Fabricate top panel for dorade box, install baffle, heatshield, vent, and drain hole and prepare for installation on deck. Screw and plug the box (Done)
 
Heat reflective lining as this is a hot air exhaust 
Teak oil to protect it for now.
Drain holesPocket drilled holes
5. Secure dorade box to deck with 3M 5200 and pocket screws.
6. Install califont on bulkhead through cabinet. Because the califont's top mount is a screw pocket, I will need to mark its position on the bulkhead,  then use some plastic or tracing paper to set out the screw in the correct position (like a picture hook). The bottom screws are simply driven through holes in the chassis. DONE!!


Update: 4/8/21

The first guaranteed sunny day meant an early start before work ~ time to pull the deck plug and get this chimney installed!

Plug is gone, not far from the reducer to the deck

The stainless steel flex-tube comes by the meter!  You can see how it can conform to the offset

I used JB Weld High Temp (600F) epoxy putty above, below, and around the opening as a barrier

Cut to size with a hacksaw

Three layers of the high temp putty, in between the top and bottom deck laminate, around the base of the flue, and another layer blended up the flue. *The repaired gelcoat cracks/epoxy will vanish when I apply non-skid in the spring.

TIPS! 

~ Sand the SS pipe and the deck with 80 grit to get a good bond.
~ Warm the putty in the sun before trying to knead it and blend it.
~ Bring a cup of warm water and wet your fingers as you roll the putty into a long thin snake, and work it down into the gap and up the flue.  I continued to work it for a while until it was very smooth and uniform, and sealed well around the pipe and deck.

I also fitted the dorade box with 3M 5200 Fast Cure.  I taped it off around the bottom outside edge to keep the 5200 off the teak.  I set it temporarily in place and marked the spot with pencil marks.

Matching set!

I weighed the box down with a diving belt while it set up but LEARN FROM MY EXPERIENCE! - Make sure you have some shims or blocking to keep the dorade from sliding down the cabin top before the glue sets.  It is a substantial curve and the mass is both straight down and off at 45degrees...the box will slide if you don't lock it in place.

7. Replace water pump and accumulator (Done), plumb new buteline pipe to inlet of califont, then hot water pipe to galley and head.
8. Run gas line from LPG locker to stove and califont.  Make connections at both ends and then remove original hoses.
9. Varnish/ oil dorade boxes.
10. Modify and install trim from galley top to underside of califont to hide hoses.

Update 22/9/21

Dont get me started on brass fittings.   Did you know that there is a BSPP and a BSPT? Well the P stands for parallel and the T for tapered.  I had to find a fitting that was 1/2" BSPP to 3/8" BSPT. I am waiting for it to arrive to connect gas line to califont. 

Update 19/10/21

FINISHED!!

Plumbed and working, I just needed to build a cover over the hoses to tidy it all up.


If you're wondering, the satin finish comes from 6 coats of gun stock oil.