Canbus Electrics

[westernman]

I am no expert, i just listen to my Captain friends tales of woe. They hate the systems installed on their boats.
I think part of the problem is that all these superyachts are one offs. They all get a bespoke design. Very different to cars that go through rigorous testing. Anybody converting a small yacht would likely have less issues simply because a small yacht will be way less complex.
I have a .raymarine nmea2000 backbone in my boat. I just traced a power spike problem on a wind speed repeater to a faulty 24/12v dropper. It must have an internal short as it was super hot to the touch. It dragged the system voltage down. New dropper and we are back in business. It took a day to find the problem. It would be way simpler with hard wired system in my view. When canbus works it great. Can be hard to trace faults though

The key to any kind of data networking system is that the bus is properly terminated and for any shared media system (which CAN is and original ethernet was) stubs must be kept very short. Any extension to such systems have to go through a proper store and forward system.

The NMEA 2000 if properly installed is very reliable. However, many people have often tacked something onto the bus without doing a proper job. E.g. with a stub which is too long, with improper termination etc etc.

 

[Baggywrinkle]

I think it has been asked before, what s/w did you use for the dia, may have to do one like that (but it's going to be a bit larger I'm afraid with just one N0183 branch/device
Would be good to have such a graph for others (or me with memory failing as time passes!)

Unfortunately, it was done in 2021, and I sold all the HW with the boat .... from what I remember I had SW on the tablet from Yacht Devices or another source which received the YDNR data over WiFi and I could see the raw messages .... it was obvious that the Seatalk1 sub-net was being flooded due to the total amount of traffic on all the other busses - the router routes everything everywhere by default. A simple disable of all routing to ST solved the problem - so I knew it was the routing.

NMEA0183 was simply an RS232 to USB device and a terminal app. In post #38 above, the picture of the chart table has 2 x D-Sub 9-Pin connectors in the bottom right ... they were NMEA0183 Plotter Data and AIS data before I added NMEA2000 + WiFi.

This was the info in the handover document I gave to the new owner along with the wiring diagrams.

 

[vas]

thanks for the info, wrongly formulated question on my behalf:
curious on the software you used to produce the graph, was it actisense or a mix and match bag photoshoped together in a full thing?
Will probably never bother to do such a detailed diagram from scratch, would need a push so to speak

cheers
V

 

[geem]

The key to any kind of data networking system is that the bus is properly terminated and for any shared media system (which CAN is and original ethernet was) stubs must be kept very short. Any extension to such systems have to go through a proper store and forward system.

The NMEA 2000 if properly installed is very reliable. However, many people have often tacked something onto the bus without doing a proper job. E.g. with a stub which is too long, with improper termination etc etc.

I will give you an example. Fly by wire throttle/gears on a 72ft racer cruiser. To swap from one wheel to the next, there is a protocol. You set the lever to neutral and the light changes colour so you can then go to the other wheel and push the button and you now have control from the opposite wheel.
It is supposed to be impossible to start the engine with the levers in drive. Both levers at the wheels should be in neutral to start the engine. On one occasion, the engine was started, with both levers in neutral but the engine went to immediate full reverse throttle. Fortunately the boat was at anchor. The Captain was able to take control at one wheel and throttle the engine back. If this had happened at the dock it could have been a different story.
After the event it was impossible to replicate the event again and it hasn't happened since. Was was at fault? The engineers checked out the fly by wire throttles and could find no fault. Bloody scary for the Captain and crew at the time. A lot of loss of confidence in the system as a result. They never knew if it was going to happen again.

 

[westernman]

I will give you an example. Fly by wire throttle/gears on a 72ft racer cruiser. To swap from one wheel to the next, there is a protocol. You set the lever to neutral and the light changes colour so you can then go to the other wheel and push the button and you now have control from the opposite wheel.
It is supposed to be impossible to start the engine with the levers in drive. Both levers at the wheels should be in neutral to start the engine. On one occasion, the engine was started, with both levers in neutral but the engine went to immediate full reverse throttle. Fortunately the boat was at anchor. The Captain was able to take control at one wheel and throttle the engine back. If this had happened at the dock it could have been a different story.
After the event it was impossible to replicate the event again and it hasn't happened since. Was was at fault? The engineers checked out the fly by wire throttles and could find no fault. Bloody scary for the Captain and crew at the time. A lot of loss of confidence in the system as a result. They never knew if it was going to happen again.

Fly by wire means stuff has to be properly wired up.

A rats nest of wires and improper out of spec bus connections will create chaos.
Non standard, non professional standard installation wiring will eventually result in issues. Some of those can be seemingly random. Wiring in many boats is substandard. Particularly when stuff has been added after the boat left the yard.

A properly installed CAN bus system is more reliable than the wiring it replaces.

 

[Baggywrinkle]

I think these weird failure modes are way more likely with Cambus systems. That's why my Captain friends on Cambus based superyachts detest them so much. We have pretty tiny systems and it's apparent that it's not just me having issues. Can you imagine the potential failure issues on a 100 metre superyacht.

Where I work, the engine controller, gearbox controller, x-drive clutch, airbag gyro system, and braking system all communicate with each other in real time to provide torque vectoring to ensure the car goes in the direction intended, doesn't swap ends, and all the systems respond with millisecond accuracy - it all used to run over CAN but is now FlexRay. We have put millions of cars with these systems on the road.

When you are in total control as a design engineer of all the systems communicating with each other - and the messages they send - then it is very reliable. The bus itself is not the problem.

I suspect that what your friends are seeing are integration problems between systems that are "supposed" to be compatible rather than a failure of the underlying communication bus or the technology behind it.

 

[Baggywrinkle]

thanks for the info, wrongly formulated question on my behalf:
curious on the software you used to produce the graph, was it actisense or a mix and match bag photoshoped together in a full thing?
Will probably never bother to do such a detailed diagram from scratch, would need a push so to speak

cheers
V

lol ... I thought "dia." was diagnostics, not diagram.

You might laugh, but the diagram was done in Powerpoint.

 

[geem]

Where I work, the engine controller, gearbox controller, x-drive clutch, airbag gyro system, and braking system all communicate with each other in real time to provide torque vectoring to ensure the car goes in the direction intended, doesn't swap ends, and all the systems respond with millisecond accuracy - it all used to run over CAN but is now FlexRay. We have put millions of cars with these systems on the road.

When you are in total control as a design engineer of all the systems communicating with each other - and the messages they send - then it is very reliable. The bus itself is not the problem.

I suspect that what your friends are seeing are integration problems between systems that are "supposed" to be compatible rather than a failure of the underlying communication bus or the technology behind it.

Agreed. The problem with these superyachts is they are one offs. They are all different. They all have problems. They are all complex. Really hard for the crew to fix stuff when it goes wrong. Totally reliant on engineers flying all over the world to fix the problems they develop

 

[vas]

lol ... I thought "dia." was diagnostics, not diagram.

You might laugh, but the diagram was done in Powerpoint.

you are definitely not helping there I'm afraid