2021-11-17, 04:11 AM
I have the common Raymarine i50 speed, depth and i60 wind instruments. According to Raymarine I need about $100 worth of cabling and connectors to interface these instruments with a NMEA 2000 network. That seems much too bulky, complicated and expensive for a less than one meter long network. I came up with a simple and free method if you already use the Waveshare 2-CH CAN HAT.
First a bit of background on SeaTalk ng. The cables carry +12v on red wire. Black is return. White is (NMEA H) and blue is (NMEA L). They also have a yellow wire which is the old SeaTalk protocol. The instruments default to sending data on this wire unless they "see" the 60 ohm resistance of a connected SeaTalk ng or NMEA 2000 network. If the 60 ohms is present across white and blue wires then the instruments switch to sending NMEA 2000 on those wires. BTW, the 60 ohms is just the two 120 ohm backbone terminating resistors in parallel. Conveniently the Waveshare 2 channel CAN hat has a 120 ohm resistor built into each channel. If you jumper from H to H and L to L of the two channels you've just created a properly terminated NMEA 2000 backbone. I've sketched this up to help clarify.
I haven't tried it but I'm pretty sure that, for those who are using a single channel CAN hat, all you have to do is connect a 60 ohm resistor across the blue and white wires to get SeaTalk ng data flowing.
If you study the NMEA 2000 or CAN bus wiring specs it seems quite complicated with strict requirements for cable lengths, terminating resistors or where to inject power (only on the backbone never a spur) etc. All this is important for larger networks but a simple, short, low current network can ignore many of those requirements as long as there's 60 ohms across the L and H and you don't exceed the current carrying capacity of the red and black wires. In case of daisy chained instruments like the i50/60 you probably don't want to supply power on one end of the chain, at an instrument and then power the Rpi on the other end as it is unknown how much extra power you can safely pull through the instruments.
Simon
First a bit of background on SeaTalk ng. The cables carry +12v on red wire. Black is return. White is (NMEA H) and blue is (NMEA L). They also have a yellow wire which is the old SeaTalk protocol. The instruments default to sending data on this wire unless they "see" the 60 ohm resistance of a connected SeaTalk ng or NMEA 2000 network. If the 60 ohms is present across white and blue wires then the instruments switch to sending NMEA 2000 on those wires. BTW, the 60 ohms is just the two 120 ohm backbone terminating resistors in parallel. Conveniently the Waveshare 2 channel CAN hat has a 120 ohm resistor built into each channel. If you jumper from H to H and L to L of the two channels you've just created a properly terminated NMEA 2000 backbone. I've sketched this up to help clarify.
I haven't tried it but I'm pretty sure that, for those who are using a single channel CAN hat, all you have to do is connect a 60 ohm resistor across the blue and white wires to get SeaTalk ng data flowing.
If you study the NMEA 2000 or CAN bus wiring specs it seems quite complicated with strict requirements for cable lengths, terminating resistors or where to inject power (only on the backbone never a spur) etc. All this is important for larger networks but a simple, short, low current network can ignore many of those requirements as long as there's 60 ohms across the L and H and you don't exceed the current carrying capacity of the red and black wires. In case of daisy chained instruments like the i50/60 you probably don't want to supply power on one end of the chain, at an instrument and then power the Rpi on the other end as it is unknown how much extra power you can safely pull through the instruments.
Simon