Trawler Forum

Trawler Forum (http://www.trawlerforum.com/forums/)
-   General Discussion (http://www.trawlerforum.com/forums/s3/)
-   -   NMEA 0183 wiring question (http://www.trawlerforum.com/forums/s3/nmea-0183-wiring-question-42705.html)

seattleboatguy 01-24-2019 12:47 PM

NMEA 0183 wiring question
 
I would like to get my new AIS talking to my existing chart plotter using a NMEA 0183 network, and I would like to make sure I'm connecting the 0183 network wires correctly. I took a couple screen shots from the AIS installation manual and from the chart plotter installation manual. I also made a sketch of how I "think" all this magic should be connected. My questions are:
  1. The chart plotter "tx" wire should connect to the ais "receive +" wire, right?
  2. The chart plotter "rx" wire should connect to the "ais transmit +" wire, right?
  3. The ais "transmit -" wire and the ais "receive -" wire don't connect to anything, right?
  4. What would you recommend for the fuse size on the +12v power wire?
Thanks very much.


http://www.seattleboatguy.com/tmp/ais/3.jpg

http://www.seattleboatguy.com/tmp/ais/1.jpg


http://www.seattleboatguy.com/tmp/ais/2.jpg

twistedtree 01-24-2019 01:25 PM

I've heard that some Garmin products aren't NMEA 0183 compliant, and this is an example. It's pretty lame, but that's a whole other topic.


You will probably need to ground "receive -" (also labeled "Z"). But that said, there is probably no communications from the chart plotter to the AIS anyway, so you could even leave out that whole direction of communications.


What really matters is data from the AIS to the chart plotter.


Contrary to what Garmin's instructions say, I would do as you have shown and NOT ground "transmit -" ("X"). That would have the potential to damage the transmitter in the EmTrak.

tiltrider1 01-24-2019 08:30 PM

AIS. Brown Tx(+) connects to Garmin Rx(+) Brown
AIS Blue Tx(-) connects to Garmin Rx(-) Black (Garmin calls this data ground)

No other data wires need be connected.

Gordon J 01-24-2019 08:38 PM

Don't forget to set the proper data rate in the chart plotter set up.

twistedtree 01-25-2019 04:38 AM

Quote:

Originally Posted by tiltrider1 (Post 734538)
AIS. Brown Tx(+) connects to Garmin Rx(+) Brown
AIS Blue Tx(-) connects to Garmin Rx(-) Black (Garmin calls this data ground)

No other data wires need be connected.


Connect the AIS Tx- to ground at your own risk. The AIS is trying to switch that between 0V and +5V, and you will have shorted it to ground. Depending on the design of the AIS output drivers, you could damage the device. Some are fine with it, but other can be damaged. Furuno warns about this in a number of their manuals.


I think it's really lame that Garmin doesn't build a correct NMEA 0183 interface. It needs to be two wire, differential, end of story. Otherwise it's not NMEA 0183.

diver dave 01-25-2019 07:47 AM

Quote:

Originally Posted by twistedtree (Post 734591)
Connect the AIS Tx- to ground at your own risk. The AIS is trying to switch that between 0V and +5V, and you will have shorted it to ground. Depending on the design of the AIS output drivers, you could damage the device. Some are fine with it, but other can be damaged. Furuno warns about this in a number of their manuals.


I think it's really lame that Garmin doesn't build a correct NMEA 0183 interface. It needs to be two wire, differential, end of story. Otherwise it's not NMEA 0183.


Twisted: I think your heatburn is that the Garmin uses power RET as the NMEA NEG, and does not use a fully isolated driver on transmit.
From when I last designed with NMEA 0183 (circa 1992), I did pay for the spec and had it in front of me. IIRC, only the receive side was speced to be a fully isolated, (floating from ground) circuit. We used a optocoupler on the RX side. The TX side was not speced to have isolation from earth/ground/power.
Garmin COULD use a seperate wire for the Tx NEG side, but makes you use the power NEG.
If you do a bunch of Furuno gear, you will find manytimes a floating power reference, so that they can accomodate a NEG or a POS power ground. There, you can't do a "willy-nilly Garmin" approach to use EARTH, but need to find the proper NEG reference to use for communications.
I'm only midway thru my first coffee; so inbound corrections are welcome. :D

ps: that first release of NMEA0183 had only 4800 baud rates speced. I hope later revs included the fast AIS baud, and maybe even address multiple receivers on the same line. Early designs did not guarantee receivers, past one!

twistedtree 01-25-2019 11:50 AM

Quote:

Originally Posted by diver dave (Post 734611)
Twisted: I think your heatburn is that the Garmin uses power RET as the NMEA NEG, and does not use a fully isolated driver on transmit.
From when I last designed with NMEA 0183 (circa 1992), I did pay for the spec and had it in front of me. IIRC, only the receive side was speced to be a fully isolated, (floating from ground) circuit. We used a optocoupler on the RX side. The TX side was not speced to have isolation from earth/ground/power.
Garmin COULD use a seperate wire for the Tx NEG side, but makes you use the power NEG.
If you do a bunch of Furuno gear, you will find manytimes a floating power reference, so that they can accomodate a NEG or a POS power ground. There, you can't do a "willy-nilly Garmin" approach to use EARTH, but need to find the proper NEG reference to use for communications.
I'm only midway thru my first coffee; so inbound corrections are welcome. :D

ps: that first release of NMEA0183 had only 4800 baud rates speced. I hope later revs included the fast AIS baud, and maybe even address multiple receivers on the same line. Early designs did not guarantee receivers, past one!


I believe the issue is two fold. First, as you point out, using a single signalling wire requires a common reference, typically the equipment ground. That's Garmin's foot note about having a common ground for a single wire system. Alternately, a floating/isolated receive side can be referenced to the common ground by grounding the Rx-. That mostly works for a 1 wire Tx talking to a 2-wire Rx.


But it will still operate in an undefined voltage range, so you are counting on the actual chip/circuit implementation having defined behavior in an undefined area of the spec. This is because a valid zero (or maybe it's a one, I can't recall the polarity) requires a negative voltage across the Rx pair, yet a single Tx wire only operates between 0-5v, i.e. it never goes negative.


The second problem is if the 2-wire transmitter is not isolated or sufficiently protected, you can blow it out by grounding the Tx-. 0183 calls for 0-5V signalling on the Tx- wire (Tx+ too). But if you ground it, you have shorted out the output of the chip/circuit which might damage it. I mentioned Furuno because they caution about this, and I guy I know there told me they see more failures due to this than anyone would care for. I expect lots of devices have current limiting in their drivers and would be unharmed, but if you luck is anything like mine, you would wire up the one device that isn't protected and blow out the port.

seattleboatguy 01-25-2019 12:07 PM

Quote:

Originally Posted by twistedtree (Post 734669)
I ... but if you luck is anything like mine, you would wire up the one device that isn't protected and blow out the port.

Speaking of blowing out ports, what would be a reasonable size for the fuse on the 0183 12v+ power wire?

diver dave 01-25-2019 01:17 PM

OK; since I designed with 0183, I now see the standard has changed for the transmitter. Notably, version 1.5 has a single ended config, and later revs, to at least 3.2, have a differential transmitter. The recent "spec" is to simply specify "RS-422" as the Tx device. Alright, this is quite enlightening, and, of course, the trend to higher data rates make this rather inevitable. Saying that, 422 can still accept on the order of 10V common mode before damage will occur. And, must withstand shorted + to -, and shorts to reference. BUT, if you do connect it to 12V, as one example, it may blow up. And, like you say, 422 is a differential source, with an expectation of twisted pair, and impedance control; and yes, Garmin violates that in a number of ways.

I might still hazard a guess that Furuno makes this a bit more interesting in that some (or even most) products float the chassis off of power NEG, in order to support POS earth installs. A wiring issue here could present 12 or 24V dc to the rs422 transmitter, which I can see can cause death to the system. Especially if the installer took a Garmin approach to use frame ground as RS422 "B" (NMEA ret).

twistedtree 01-25-2019 02:23 PM

Quote:

Originally Posted by seattleboatguy (Post 734671)
Speaking of blowing out ports, what would be a reasonable size for the fuse on the 0183 12v+ power wire?


There isn't a fuse for 0183 itself, like there is for NMEA 2000. Just power/fuses for individual devices. I presume the only thing you are adding is the AIS, and its instructions should tell you what size used to use.

gkesden 01-25-2019 02:29 PM

NMEA, like most electrical transmission standards, requires that both a "+" and a "-" be connected for data.

To avoid confusion between the "- for data" transmission and the "battery -", the data wires are now usually referred to as "A" and "B", e.g. "Tx-A", "Tx-B", "Rx-A", "Rx-B", rather than + and -.

So, you are going to want transmit-A(+) to receive-A(-) and transmit-B(-) to receive-B(-).

If all you see are "transmit" and "receive" wires on each device, or on either device, look for something like "data -".

Because NMEA-0183 is built on RS422 underpinnings, it uses differential voltage, the data - is different from the battery ground.

twistedtree 01-25-2019 02:33 PM

Quote:

Originally Posted by diver dave (Post 734691)
OK; since I designed with 0183, I now see the standard has changed for the transmitter. Notably, version 1.5 has a single ended config, and later revs, to at least 3.2, have a differential transmitter. The recent "spec" is to simply specify "RS-422" as the Tx device. Alright, this is quite enlightening, and, of course, the trend to higher data rates make this rather inevitable. Saying that, 422 can still accept on the order of 10V common mode before damage will occur. And, must withstand shorted + to -, and shorts to reference. BUT, if you do connect it to 12V, as one example, it may blow up. And, like you say, 422 is a differential source, with an expectation of twisted pair, and impedance control; and yes, Garmin violates that in a number of ways.

I might still hazard a guess that Furuno makes this a bit more interesting in that some (or even most) products float the chassis off of power NEG, in order to support POS earth installs. A wiring issue here could present 12 or 24V dc to the rs422 transmitter, which I can see can cause death to the system. Especially if the installer took a Garmin approach to use frame ground as RS422 "B" (NMEA ret).




Yes, I meant to mention the different revisions of 0183 but forgot by the time I had written all my garbage above. I don't recall when it changed from essentially RS232 to its present RS422, but it's been RS422 for all the time I have been messing with this stuff which is about 10 years.


And as a bit of a side note this is why connecting marine electronics to an RS-232 port on a computer can be very problematic. It almost always at least appears to work, but can be a real mess under the covers. I made such a connection a while back because the only computer adapter I could get in Juneau was RS232. It all seemed to work fine, but then I looked at the incoming data stream using Coastal Explorer's sentence viewer and for every few good sentences, there was a corrupted one coming in as well. CE was discarding them because they made no sense, but some portion of good messages were lost, and imagine what could happen if one of those scrambled messages actually looked like some other valid message and my boat took action on it.....

seattleboatguy 01-25-2019 02:36 PM

Quote:

Originally Posted by twistedtree (Post 734716)
There isn't a fuse for 0183 itself, like there is for NMEA 2000. Just power/fuses for individual devices. I presume the only thing you are adding is the AIS, and its instructions should tell you what size used to use.


This is a picture of the cable I was going to buy, which does have a fuse. Should I be looking to buy a different type of cable? If so, which cable would you recommend?


http://www.seattleboatguy.com/tmp/ais/4.jpg

mattkab 01-25-2019 03:28 PM

Save yourself a bunch of headache, and get a screw terminal block, something similar to this:
https://smile.amazon.com/gp/product/...e_o06__o00_s00


I used that when doing a similar project, connecting AIS to a chartplotter, and I bought it to make the initial connections easier. It worked so well, it became the permanent installation.

diver dave 01-25-2019 03:30 PM

Just for "completeness"; the 0183 didn't call for RS232. that standard requires a negative source for the "low" signal; to bring it below signal reference. The early 0183 standards simply called for "no current", not some negative voltage, like 232. IOW: don't turn on the optocoupler at the receiver.



So, therefore, I disagree with this association of protocols. https://www.actisense.com/difference...-of-nmea-0183/

One of the good things about 422/485 is to NOT require neg voltage sources; only +5V. And, early or late 0183 did NOT need neg V sources at either end.

seattle: didn't your garmin come with the combo power/data cable? Or, are you saying the existing power cable only does power connection, or??

gkesden 01-25-2019 03:33 PM

To my knowledge, NMEA-0183 has always been RS422-based. NMEA-0180 and NMEA-0182, the predecessor standards, were RS232-based.

That, of course, hasn't stopped folks from hooking NMEA-0183 devices to RS232 ports -- sometimes with success.

But, other than by observation over time, it is very hard to know what results one might obtain by doing that without knowing likely hard to find details of how the specific ports and devices in use internally implement their respective standard interfaces. So, I guess the advice about that is, if it isn't important, one can try and enjoy it if it works for as long as it works.

...but RS422-to-USB devices are pretty inexpensive on Amazon and are generally reliable solutions.

diver dave 01-25-2019 04:19 PM

here is a decent history of this horribly supported transport:

Boat Projects: Beginners guide to Nmea 2000, Nmea 0183, and bridging

gkesden 01-25-2019 05:59 PM

Quote:

Originally Posted by diver dave (Post 734741)
here is a decent history of this horribly supported transport:

Boat Projects: Beginners guide to Nmea 2000, Nmea 0183, and bridging

True that.

I just chased this down in the docs to see when the change happened. It looks like the 2.0 standard came out, and with it the switch to RS422, in 1992.

Jeff F 01-25-2019 08:51 PM

I recently installed a new Garmin 942sx and a Standard Horizon GX2200. I wanted to share AIS data from the VHF to the plotter. What should have been simple was definitely not.

I read the documentation from both, and it left me stratching my head a bit, but I got a terminal block and tried connecting, first the TX+ to RX+ only, then adding TX- to Garmin black. Nothing.

I double checked the configuration on both sides, read through some support stuff online. No new ideas. I then tried switching to the second Garmin RX+. Nothing. But after trying that the second input port disappeared from the MFD menus. Uh Oh. Tried disconnecting and resetting everything but couldn't get the original Garmin port 2 input menu back.



I talked with a couple of folks at the place I got the gear. Next step was to call Garmin support.

Before I did that I tried connecting my old Garmin 540(?) plotter to the AIS source. Made the one wire connection and had it. Worked perfectly. So I was pretty confident the issue was in the new MFD.

After an hour or so on the phone they had me connect tx+ to Garmin RX+ and tx- to the second Garmin RX+ and it works.

Sorry I don't remember the wire colours offhand, but can check if anyone is interested.

seattleboatguy 01-25-2019 08:57 PM

Quote:

Originally Posted by diver dave (Post 734734)
seattle: didn't your garmin come with the combo power/data cable? Or, are you saying the existing power cable only does power connection, or??


I assumed (ass-u-me) that, since I had never used a NMEA 0813 network up to now, that it would require a new cable to plug into an un-used port in the back of my chart plotter. But, when I read your post tonight, I decided to take another look. As I think you already suspected, the NMEA 0813 wiring was already attached to the same cable that I have (up to now) only used to bring power to the chart plotter. So, no new garmin cable required. Since the ais has a similar setup, where the power wires and NEMA 0813 wires are all integrated into the same cable, I'm good to go there as well. Thanks for (very diplomatically) pointing that out.


All times are GMT -5. The time now is 12:22 PM.

Powered by vBulletin® Version 3.8.8 Beta 4
Copyright ©2000 - 2019, vBulletin Solutions, Inc.
Search Engine Optimization by vBSEO 3.6.0
Copyright 2006 - 2012