linking ST2000+ autopilot, GPS & wind

F

fjweaver

Well-Known Member
Joined
16 May 2001
Messages
256
Location
Middlesex
Visit site
I have bought an ST2000+ autopilot which I intend to link to my Garmin GPS & also ST50 Wind instrument. The GPS already has a Clipper repeater taking NMEA output from it. From the autopilot manual I think I can link both the NMEA & SeaTalk outputs.
Firstly is there a problem taking the NMEA output from the GPS to both the repeater & the autopilot?
And secondly I’m not sure what cable I should be using, I think it should be shielded cable to guard against any interference (and where’s the best place to get it from?).
 
I think you will need a PC/Seatalk/NMEA Interface Box from Raytheon (or are they Raymarine now. Your ST stuff will be talking in Seatalk which is different to NMEA. I have Seatalk instruments, a Simrad tiller pilot and a Furuno GPS. It all wires into the different terminals and comes out fine. The box also allows connection to a PC/Laptop. Wasn't that cheap if I remember but its a nicely made bit of kit about 6" x 4" x 1.5". Got it from Greenham Regis.
 
Sorry, I should have added that I take a NMEA feed from the GPS to the Furuno Navtex (so I can run two different GPS displays at the same time and switch over to the Navtex messages at any time). It should be possible to take a NMEA feed from your GPS to the Autohelm without any fancy interface and you will be able to set it to steer to Waypoint. If you want to steer to wind, you will need the interface box.
 
Re: linking ST2000 autopilot, GPS & wind

It is against the specification to take a NMEA output to more than one input, but you can usually get away with it. NMEA is meant to be single talker (output) connected to a single listener (input).

The only way to see if it works is to connect it up and have a go. If the output is not strong enough to drive 2 inputs, you won't do any harm, it just won't work.

What you certainly cannot do is the other way round, ie connect two talkers to a single listener using the same input connectors. You need a multiplexer to do this, sometimes called a concentrator.
 
You\'ve got me worried.

I'm also in the process of enhancing my instrumentation. I asked Garmin about connecting a GPS to a tillerpilot, a radio and a Yeoman, and was told that it was OK. Their spec says no more than three.
I also asked Simrad about connecting a Navman wind indicator to a tillerpilot and was told that it was OK. What I did not say was that a GPS was also connected?
I will go back to Simrad on this point.
 
Re: You\'ve got me worried.

What you will need to do that is a multiplexer, like the one on this site...

<A target="_blank" HREF=http://marinecomputing.com/products/nmeaseatalk/>http://marinecomputing.com/products/nmeaseatalk/</A>

...but that price, ouch!

Some instrument display heads have a built in multiplexer capability, for example the Cetrek C-Net 2000 display. For the price of just a multiplexer you could almost get another display as well that incorporates one.

These are vey simple devices, it is a pity they cost so much, lack of demand I think. When I last wanted something like this (a NMEA 1.0 to 2.0 protocol converter) I made one with bits I had lying around, for next to nothing.
 
Re: You\'ve got me worried.

Hopefully this answers my original question - no need for any interfaces even for steering to wind

-------------------------------------------

Interfacing the ST2000+, ST60 and NMEA 0183

Question:-
I am considering a system with ST2000+ and ST60 wind system (and possible other ST60 instruments to be added in the future). My current GPS unit is Garmin and Can I connect it to the ST2000+ NMEA port and then have the GPS sent NMEA sentences sent over to the seatalk network by the ST2000+ (to be used by the ST60 Wind to calculate true wind values) . Is this how the ST2000+ works ?

Will the ST2000+ read and use the ST60 wind information it receives from SeaTalk to steer to wind ?

Answer:-
The ST2000+ will read and use the ST60 wind information it receives from SeaTalk to steer to wind and it will also recieve NMEA 0183 waypoint information from a GPS or loran. It will not pass along your GPS' SOG information to your wind instrument to be used in calculating true wind speed and direction. To accomplish this you would have to have your GPS feed its NMEA to our E85001 NMEA/Seatalk interface box, then tie the ST2000+ and ST60 wind to the E85001 via Seatalk to get all of the information shared in your system.

The ST2000+ only decodes and translates into SeaTalk waypoint steering information from a NMEA compatible device.
 
Ouch ...

Try these guys, "http://brookhouseonline.com/nmeamux.htm", down in NZ, but price seems better suited to most pockets if you need to combine 2+ NMEA0183 data streams. Haven't bought it yet, it's top of my list so far.

I am told by them (and pages on the web) that a single NMEA0183 output should be able to drive 2 or 3 inputs ok, but its worth noting that old NMEA devices operate in a different electrical manner to modern ones. Something to do with old devices driving signal relative to gound (thus requiring only one wirm) and newer ones (fully compliant with NMEA0183 standard) using a differential signal (requiring two wires).

My twopennyworth, Jeff.
 
Re: Ouch ...

Still Ouch but not so big a bruise at 119 US dollars, especially as I had not given this aspect a thought. Is the problem two simultaneous messages getting shuffled together?
My tillerpilot will not be operating in both GPS mode and wind mode at the same time. I know that I can turn the GPS interface off. If I can also easily turn off the NMEA output from the wind indicator, I have solved the problem. Could I install a switch in the wiring to block one of the NMEA outputs?
 
Re: Ouch ...

Essentially NMEA0183 seems to be a "one-way" standard: One Sender and One Listenner. Fortunately the electronics/chips which do the actual line driving at the senders end are normally powerful enough to drive multiple listenners. The problem comes when you have multiple Senders.

The NMEA protocol does not contain any hand-shaking or time sharing components - its is really just a form of Serial computer communications: 0ne way, 4800 baud, 1 stop bit, 8 data bits (but with top bit set to 0). So connecting two senders together to likely to result in NMEA sentences being sent at the same time, and the content lost, or worse garbled, confusing the Listenners. Not what you want with a DSC radio.

These "NMEA Combiners" are really quite simple boxes (from a computer point of view), as they generally seem to be 4 Listenner ports, and one RS422/NMEA Sender port. The box listens for NMEA sentences on each of the input ports, and buffers it up in memory, when the sentence is complete, and the Sender port is free, it outputs the buffered sentence. Adds a small delay, but speed isn't of the essence here, reliability is.

Some companies also make an NMEA splitter box: one listenner and 4 Sender ports. This is purely an eletrical box of tricks (after all it is only doing the same thing as twisting the right writes together), but provides "Opto-Electrical" isolation. By this they mean that the straight through electrical path from sender to listenner is broken by placing an LED and PhotoResistor between them. This make each piece of equipment eletrically independent of each other and reduces the impact of noise on either the singal lines, or the power supply lines. May also help if one device has a fault, isolating consequences of fault to one device.

As for the old NMEA devices - I'm not completely sure what the consequences of this are. Apparently it is a case of trying and seeing if it works, but I'd guess that pratically everying which is arround these days will conform to the proper/newer standard.

I am planning (money permitting), to install a Navtex 6+ later this year, and would like it to repeat the NMEA outputs from both my GPS and Instruments, so I'll be needing one of these combiners then. The NZ company were the cheapest I've found so far.

Does trhis help?
Jeff.
 
Re: Ouch ...

The tricky thing about "combiners" is what to do in the case of buffer overrun. In theory if you are trying to combine two NMEA streams it may not be possible to retransmit all the messages as there may not be the capacity for it on the line out. Therefore you get into the situation of having to discard messages that are either duplicate or out-of-date.

I would like to know if the commercial units handle this in an intelligent way.

Another thing I'd like to know - do any of the devices that have both a NMEA in and NMEA out actually duplicate the in on the out - if so it would make life simple. I intend to test that out once my electronics are back on the boat.
 
Yes, but...

Yes, this is a good point. There really isn't a lot which can be done - after all you really cannot make a quart fit a pint pot.

There are only two strategies to "manage" this eventuality:

1/ binning the oldest sentence, and
2/ prioritisation of input channels.

The combiners that I have read about so far all do #1, but only one so far (the expensive one mentioned earlier in this thread), does #2.

It seems that you can only evaulate your potential input bandwidth to the combiner, and keeper talker volumes down to below 480 bytes/s (approx). This is actually quite a lot of traffic in reality.

A GPS message must be about 20/30 characters, and each would be sent only once a second. The sentences from instruments are probably going to be shorter, and are sent on a similar frequency.

We are looking for the following infomation to be sent over the bus:

Position Lat/Log
SOG
COG
Distance to waypoint
Course to steer to waypoint
Wind speed/direction
Boat speed
Water depth
Battery Voltage

I must have missed some.. anyway,

So worst case load could be estimated at about 200 bytes/sec. (By the way, there is significant finger in the air waving going on here). Not having an actual electronic plotter/yeoman I've no idea if they generate data for the NMEA bus - but I do not see why they would.

So I don't think that there is a serious risk of the combiner having to drop sentences. However, in reality the "key" equipment would not be operating on the output of the combiner. The output from the GPR can be linked directly into the DSC radio - no potential data loss. Like wise into the electronic nav equipment. My B&G stuff takes input from the GPS, but only this, so again a direct connection is all that is required. The only reason I would need a combiner is to bring both the outputs from the B&G and the GPS together to feed into the Navtex 6+, so that is can display data from eitther, and can keep running log of everything. Some data loss here is not critical - it is not running anthing.

Jeff
 
Re: Ouch ... No, no, no

NMEA is a single sender standard, but not single receiver. How many receivers can be connected depends on the fan-out of the driver IC. Most of the ones drives used that I have seen, are also for RS442 which is multidrop and allows also for multiple senders. Most of these driver have a matching receiver chip and allow fan-out between 10 and 20.

The problem is that some devices do not comply with RS422 NMEA but use RS 232 chips which are indeed single on both ends, and they are not ground free, hence easily induce more noice.

With a real RS422 driver I wouldn't hesitate to try to drive a number of inputs, even an RS232, in the worst case it will stop working reliably due to noise, but you wouldn't really overload the sender to damage it.

Inversely you may with a badly matched setup produce more noice and disturb things like radios and Navtex.



<A target="_blank" HREF=http://www.taniwani.de>http://www.taniwani.de</A>
 
I only want one or the other, not both

I will have the GPS wired to the DSC and the tillerpilot and, if I get one, a Yeoman. The wind indicator will also be connected to the tillerpilot. But I will not want to receive messages from the GPS to the tillerpilot while on Sail to Wind mode. It seems to me that all I need is a switch to change between wind indicator and GPS input. Would a simple two way switch suffice or do I need some electronic gizmo?
 
Re: I only want one or the other, not both

I would guess that a simple 2-pole 2-way switch would do want you want: allow you to choose which data source to send top the tillerpilot.

Jeff.
 
No, no, no ... which bit, all?

Hmm - I thought that RS442 was simply a differential version of RS232 so giving benefits of greater transmission distances and "normal" speeds, or higher speeds over shorter distances. I can understand how a more robust physical layer can enable a single sender to talk at multiple listeners, but don't understand how this could be utilised as a sort of "bus" technology.

I thought that this problem (apart for the requirement for more bandwidth) has lead to the development of NMEA2000 (right name?) which is no longer a serial connection standard, but a bus based standard with more in common with Ethernet and the like?

You'll have to forgive me if I'm miss-understanding things, while I've been playing with computers for years, NMEA and boat electronics are still new to me.

Jeff.
 
Re: No, no, no ... which bit, all?

It is indeed all very confusing. To my understanding the real NMEA had defined the physical layer as RS422. A differntial signal with good common mode suppression.
RS422 is differential with an impedance of 120 Ohm and, say like a hose TV antenna would allow for a chain of matched receivers. NMEA explicitly allows that.
But, many cheapish GPS manufacturers have started doing NMEA protocol on RS232, which physically is an NRZ signal, relative to ground and common. With just RS232 receivers and senders, it'll probably be of to have a few receivers on one sender, but it will be limited. The problem that many people have with NEMA is that they end up mixing RS232 and RS422. Which kind of works, but voltage levels are differnt enough to cause some trouble aside of a complete impedance mismatch. Some RS422 receivers clamp at levels above 5V, and most RS232 drivers go out at some 7V to 9V.
Some of the comercial shipping instruments use RS485, which electrically is almost like RS422, except that the drivers are tri-state and leave the linefloating when not transmitting. This allows for multiple senders by some simple discipline as token passing or idle sensing with random retry timeout.
Often now the same drivers and receiver pairs are used ofr RS422 and RS485.
For example: LTC 488/489 as receiver and LTC 487 as driver. They quote a fan-out of 32.
NMEA2000 is yet another variant based on the car industry standard CAN, a cheapish mulltipoint network, that works on two and four wires (differnt speed), the two wire physics, very similar to RS485.
Anyway, on a boat it's all sucks, and a cheap optical fibre connector, like some high-end HiFi stuff provides would have been a much better choice.

<A target="_blank" HREF=http://www.taniwani.de>http://www.taniwani.de</A>
 
Re: No, no, no ... which bit, all?

Haven't come across RS485 before so can't comment. Not suprised that cheaper kit doesn't follow standard, something has to go to keep costs down.

This is pretty much in line with what I have read/heard - those I have spoken to or emailed say that you can reasonably expect one sender to talk to 2 or 3 listenners. But I do understand the issue of RS422 diff signal (using a +ve and -ve voltage providing to signal voltage) talking to an RS232 (using 0v and +5v (or more)). Their comments were that talking from '422 to '232 wasn't likely to work, but the other way round might if the ground input of the listenner end was independent of boat ground, so the '232 chip set saw a simple +ve voltage, or 0 (or less 8-) ) volts.

The company referenced at start of thread does a fanout box which might be sufficient to get over these problems, but at over 100 us$ it seems rather expensive for a couple of quids worth of components - should be possible to draw up simple circuit diagram to achieve this. I can say this (with smile on my face) as I don't have those skills.

I can see that this is one of those areas where if everyone stuck to the rules it would work, but thats not the way the world works. At least my GPS and B&G talk OK,1:1 connection - simple.

I suppose now I really need to find out how to identify what type of NMEA interface each of my devices has - then this will perhaps identify what future problems I'll be walking into.

Perhaps someone knows orf some simple gadets for conversion of '232 to '422 signals and back again? Again I can say this sound simple because I don't know whats involved.

Regards, Jeff.
 
Re: No, no, no ... which bit, all?

Boxes: You need to distinguish boxes that can combine multiple senders into a new combined output stream. They really pick out the data from the various incoming telegrams and compose a new list of output messages.

Passive buffers, to do level conversion or increase fan-out, need no digital logic and should be relatively cheap.

An easy way to convert RS232 optimally to RS422, is with an RS422 driver chip using some dividers and some feedback to generate a hysteresis.

For example: The LTC487 chip has four drivers and would allow you to make four separate RS422 outputs from one RS232 input. The drawing is simple enough to describe with words:

Each driver has one input and two outputs (an inverted and an uninverted output).
Take the RS232 signal and connect it to 3 resistors: R1 5.6k to ground, R2 10k to the driver input, R3 220k to the inverted output. Done. This gives a switching hysteresis of about 0.1 Volts, means with an input transitting from a positive RS232 level, say 6V to negative, the differntial output would switch when the input reaches -0.1V. This threshold can be increased and is approx. : 19k/R3



<A target="_blank" HREF=http://www.taniwani.de>http://www.taniwani.de</A>
 
Can I see some light ahead?

Yes, recognised difference between "Data combiners" and straight fan-out devices.

Have just dug up package description off rswww.com - as you say four drives in a simple 16 pin DIP - takes 12v as Vcc - makes life easy. Only question is that each driver has an "EN" line (actually a single package pin is divided and feeds two drivers) which I presume is an "Enable" line effectively telling the driver to be on or off (giving tri-state function). What should the input level for this be? My guess would be 0v/GND for off and 12v/Vcc for on, but should this be passed through a resistor or some sort? Just a guess though.

Regards, Jeff.
 
You must log in or register to reply here.

Share this page

Top