[TG] [TG11] Integrating legacy Seatalk devices

[Tom - Raymarine - Moderator]

Hello Wdeertz,

Now we start to get into the realm of personal preference and experience.

Quote:Chuck (see http://forum.raymarine.com/showthread.php?tid=8143) confirmed in this posting that all SeaTalk autopilot devices should be to a separate St to StNG converter and St non-autopilots devices to a separate converter.

Are you now saying that all St devices should go through one converter?

My view is that you should have the fewest networks possible in the system, with the fewest connection points, and no duplication of network types. My reasons are:

• Not all data is bridged through the convertor in all circumstances, due to limitations in the ~30YO design of ST1. Examples include depth unit selection and magnetic variation. If you separate networks then more of the system will potentially use the wrong value.
• latency: every translation between networks involves reception, decoding, re-encoding and a wait for a transmission slot onto the other network: these delays can add up when you have multiple links to a slow network such as ST1 so that, for example, the wind or BTW data your pilot's steering to is out of date by the time it's received.
• ST1 'backing-off': ST1 devices are designed to 'back-off' and not transmit a particular data type (e.g. depth) if they see the same data already on the network. If you have multiple networks then you can find data being echoed back to the transmitting device, which then backs off and stops transmitting.
• It doesn't look like a risk in your network, but multiple connection points are a classic cause of data loops: if you have data going in a circle between different networks (e.g. STNG > ST1 > NMEA0183 > STNG) then you can get either backing-off (no data) or worse looped data (the same static value continuously cycling round - a problem if this is depth or GPS.)
  

Quote:Tom, ok I made it over to the boat and was able to get all devices to work when I connected everything inside the pilothouse using shorter cables. See attached schematic. It seems that there is too much resistance on the 15 foot cable going to the flybridge as you suggested. Can I add supplemental 12v power for the flybridge devices? If so which pins on the ST cable should be 12 +/-?

Unless your supply voltage is really marginal, that's not just normal voltage drop but a fault in that cable. Don't attempt to work around a cable fault by adding another power connection:

1. If the problem is simply high-resistance then you provide multiple paths for power that will be at different potentials (voltages) due to the differing cable runs, the high-resistance connection between them and the different loads on either side of that connection. Depending on the instantaneous load from one moment to the next on either side of the join you will get differing current flow, and therefore voltage. Constantly varying voltage is also called electrical noise. This goes double if one of the power points is not straight off the battery but a switchmode supply as in your course computer.
2. Fuse protection. How do you choose an appropriate fuse rating for each power point: the whole network, just the half on that side of the high-resistance join? Wrong fuse ratings are not cool.
3. What if the problem in the cable isn't simply a high-resistance connection but a developing/partial short that's pulling down the supply voltage (non-infinite resistance between +12 and 0V)? By adding another power supply you mask that problem without solving it, and - presumably - add a second fuse which will need to blow in the event that the, say, brittle break in your cable sheath which is allowing red and shield to start bleeding current across through the foil wrapper opens up a little more and allows a full short.
   

Because of point 1 alone, you should only ever have a single point of power in any combined ST1, STNG network. (the NMEA2000 spec says it's ok to have multiple points for power as long as each one is fully isolated: if you don't know whether your supplies are isolated, they aren't.) When I say point of power, I'm only referring to the power for the network itself, not power for high current devices (MFDs, pilots etc.) connected to that network - they still need their own power.

It's up to you where you choose to power - though it ought to be close to the centre of the system current load to minimise voltage drop to the ends - but make sure it's in one place only. If you pull your ST1 / STNG fuse then instrument/pilot displays, GPS antennas etc. should power down: if they don't then you have multiple power paths.

Regards,
Tom