Seatalk voltages

[duncan99210]

OK, so this is probably buried somewhere in the manuals but I'm sure someone on here knows the answer off the top of their heads. What are the voltages expected in Seatalk cables? The autopilot and tridata instruments are refusing to wake up: I've got neglible volts (like 0.3) between the ground and one pin and about 3.5v between ground and the other pin. Before I start ripping things apart to get behind the chart table does anyone know if that's the correct sort of volts or should they be higher, indicating that there's a poor connection somewhere?

 

[prv]

Much higher. The power (red) cable should be at 12v compared to the negative (bare) one. The data (yellow) cable idles at 12v but is pulled down to 0v to signal, so on a busy bus a voltmeter might show a bit less than 12v on average.

Pete

 

[duncan99210]

Once again, the speed and knowledge of the folks on this forum amazes me. The answer is the one I thought I'd get. Now got to find out what's gone wrong at the chart plotter end, which is where the supply feed comes from.

 

[duncan99210]

Just to close this out, the fault lay in the negative lead feeding the autopilot and instruments where the crimped terminal had parted company with the cable.

 

[alan]

OK, I know this thread is "closed" but I thought I would just add a sentence in case someone is searching for info in the future!
I happen to have my S3 autopilot connected to an ST6002 control head and sitting on my desk at the moment (it has had an intermittent fault!!): as a very rough check I use a cheapo RMS multimeter to give me an idea if things are working and just for reference the Sealtalk red wire to ground should be the same as your power supply going to the Autopilot controller (S3 in my case) and the Seatalk yellow wire sits at around 10.5 volts and once every five seconds the voltage dips to around 9.5 volts for about 0.5 seconds. I presume that the lower voltage is when there is data being transmitted on the Seatalk bus which gives an average (lower) reading on the multimeter (one day I must buy an oscilloscope!!!). Anyway, I hope that may help someone in the future to diagnose correct operation or not of their Seatalk bus.

Alan.

 

[MattiR]

Much higher. The power (red) cable should be at 12v compared to the negative (bare) one. The data (yellow) cable idles at 12v but is pulled down to 0v to signal, so on a busy bus a voltmeter might show a bit less than 12v on average.

Pete

I'm having trouble trying to use a st60 wind display as repeater. The seatalk comes from raspberry pi 4 TxD pin. The sentences are correct, but the voltage from raspberry can only be 3.3V at highest. I would just like to have a confirmation. Does the data signal really have that high voltage?

 

[AngusMcDoon]

I'm having trouble trying to use a st60 wind display as repeater. The seatalk comes from raspberry pi 4 TxD pin. The sentences are correct, but the voltage from raspberry can only be 3.3V at highest. I would just like to have a confirmation. Does the data signal really have that high voltage?

Yes. See here for a circuit to create Seatalk voltages from lower digital voltage levels...

Thomas Knauf SeaTalk Technical Reference

 

[MattiR]

Yes. See here for a circuit to create Seatalk voltages from lower digital voltage levels...

Thomas Knauf SeaTalk Technical Reference

Thanks Angus! My desperation eases a bit. Thanks for the link!

 

[northwind]

I'm having trouble trying to use a st60 wind display as repeater. The seatalk comes from raspberry pi 4 TxD pin. The sentences are correct, but the voltage from raspberry can only be 3.3V at highest. I would just like to have a confirmation. Does the data signal really have that high voltage?

Also its a 9bit signal, what are you trying to feed it? Nmea?

 

[MattiR]

I have nmea0183 feed from pypilot.org's weather sensor. The raspi is reading the sentences and a systemcl service runs to pass the data forward. Seems I need to have the signal amplified some how. I'll try first to feed the data to my B&G Zeus2 plotter. Maybe the plotter could put out a signal the ST60 display understands.
9 bit? Hmm, not sure how to apply that to a typical sentence, say $IIMWV,320.61,R,6.58,N,A*30

 

[northwind]

I have nmea0183 feed from pypilot.org's weather sensor. The raspi is reading the sentences and a systemcl service runs to pass the data forward. Seems I need to have the signal amplified some how. I'll try first to feed the data to my B&G Zeus2 plotter. Maybe the plotter could put out a signal the ST60 display understands.
9 bit? Hmm, not sure how to apply that to a typical sentence, say $IIMWV,320.61,R,6.58,N,A*30

Seatalk is not nmea, the simplest method is to buy a converter, orbuild one of Angus's ones..
your plotter won't spit out seatalk, its a raymarine only protocol. It may spit out nmea 0183 which you could feed into a converter.

 

[MattiR]

Okay, thanks for the input. I was told at local RM service that the display understands nmea. Maybe he was wrong. The signal voltage at least seems to be too weak.

 

[northwind]

Okay, thanks for the input. I was told at local RM service that the display understands nmea. Maybe he was wrong. The signal voltage at least seems to be too weak.

Nmea is a messaging protocol, however it is generally transmitted using rs422. The output from your raspberry pi will be at 3.3v where as rs422 is 12v peak to peak.

 

[prv]

The ST60 Wind doesn’t accept NMEA data (of either type).

Even if it did, the proper voltage range for NMEA0183 is -6v to +6v (RS422), not the 0v to +3.3v TTL your Pi apparently outputs. And I imagine you would need to bit-bang the data since presumably the Pi UART doesn’t offer a mode with 9-bit bytes.

Pete

 

[LadyInBed]

10 secs of googling seatalk pin allocation produced
SeaTalk 3-Pin Press-Fit Socket Pinout

 

[GHA]

I'm having trouble trying to use a st60 wind display as repeater. The seatalk comes from raspberry pi 4 TxD pin. The sentences are correct, but the voltage from raspberry can only be 3.3V at highest. I would just like to have a confirmation. Does the data signal really have that high voltage?

Don't!! The Pi gpio pins have no protection , if they see a voltage much beyond 3.3v there's a real chance of blowing your serial pin and maybe more in the Pi. Ignoring NMEA/Seatalk, it's much safer to use a USB/Serial adapter, cost pennies anyway. And you don't lose the serial output for debugging etc.

 

[MattiR]

10 secs of googling seatalk pin allocation produced
SeaTalk 3-Pin Press-Fit Socket Pinout

Thanks, clear as a bell! But if you mean I didn't bother to look, you're wrong. I searched for "seatalk voltage", which did not produce that result. The pinout location I was familiar with, so I wasn't looking for that, although the page includes the voltages too. Well, it's always a matter of finding the right keywords

 

[MattiR]

The ST60 Wind doesn’t accept NMEA data (of either type).

Even if it did, the proper voltage range for NMEA0183 is -6v to +6v (RS422), not the 0v to +3.3v TTL your Pi apparently outputs. And I imagine you would need to bit-bang the data since presumably the Pi UART doesn’t offer a mode with 9-bit bytes.

Pete

Thanks Pete, looks like 9-bit serial has been supported for some time. Maybe too much work, but if my quarantine lasts long enough, then why not?

 

[GHA]

Thanks Pete, looks like 9-bit serial has been supported for some time. Maybe too much work, but if my quarantine lasts long enough, then why not?

also, there's a tweaked arduino software serial library knocking about which does 9 bit, not tried it but google should find it. Nano might be an easier way to go - whatever you do - be sure to share
And jlcpcb istr do the atmel chip surface mount so with time on your hands you could get a little board made up with opto isolation and the atmel chip premade for not much money. JLCPCB are great
If you've time on your hands and got a Pi running openplotter/signalk this sort of works with various monitoring and other bits plus some not finshed python for an ESP32 to send it as signalk over wifi to the Pi>>
boatybits/boatymonpy

 

[laika]

A few thoughts which may be irrelevant as it's maybe 7 years since I was mucking about with seatalk on posix systems.

• The ST60 wind doesn't support NMEA0183 but if you've been told your display does might it be the ST60 multi which *does*? That would save you a lot of trouble
• Apologies if MattiR gets this already but it's not just the data transmission that's different Between seatalk and NMEA0183, the application layer data looks completely different too. I think posters may have assumed everyone understands that, which may or may not be true.
• You can emulate 9 bit by mucking about with the parity bit IFF your uart / driver supports mark and space parity. Only one of my 3 USB to serial devices did when I was playing with this. Don't know about the pi's uart: a 10 second google suggests kernel patching is required but that may be out of date
• I suspect that "9 bit supported" mentioned above may refer to bit banging through gpio rather than uart
• Reading seatalk via Stupid Parity Tricks under Linux is one thing. For writing, some people seem to pat themselves on the back for a project well done without thinking "collision avoidance". Maybe for this project you're the only talker on the bus so you don't care. Otherwise offloading this onto a microcontroller such as one of AngusMcDoon's marvellous creations is a better plan