Bob gave some information about the interconnection of his two marine electronic devices by NMEA-0183, and then said:
I don't even understand what the heck I just wrote...
To help you understand, I can offer some explanation.
UNDERSTANDING NMEA-0183 INTERFACESThe BasicsThe NMEA-0813 interface consists of electrical signals which are in two categories: TALKERS and LISTENERS. TALKERS talk, that is they send signals. LISTENERS listen, that is, they listen for signals.
Thanks to the work of NMEA and their NMEA-0183 standard and adherence to the standard by manufacturers, just about every facet and aspect of connecting these devices has already been prearranged for you and will allow this connection to work. All that remains for you to do is to perform the PHYSICAL connection of the signals, that is, you have to connect the electrical wires that carry the signals from one device to the electrical wires that will carry those signals into the other device. This is the simplest and easiest part of the entire process of making the devices communicate. Really, just about anyone can do this.
The key to interconnecting the signals is extraordinarily simple. The signals are not bi-directional. You have to connect them for a one way flow of the signal. To do this, you connect a TALKER to a LISTENER. What could be easier to grasp? Also, you have to understand that for a two-way conversation between devices, you have to make two sets of connections: one device's TALKERS to the second device's LISTENERS, then vice-versa for the other direction.
Of course, you need some way to identify the signals. The signals have names. The names are TALKER A and TALKER B for the talker signals, and LISTENER A and LISTENER B for the listener signals. Just to be clear, each single interface, that is each "port" or "comm port", will consist of both a TALKER A and a TALKER B and a LISTENER A and LISTENER B signals for that one port or comm port or interface. Some devices have more than one set of these interfaces or ports or comports. If a device has multiple NMEA-0183 interfaces, it usually identifies them in some manner, like referring to them by numbers, like "Port 1" or "Comm Port 1" or "Primary" or something similar. But keep in mind that each port or interface will have both A and B signals for their TALKER and LISTENER signals for that interface. That means there will be four wires at each interface or port or comm port. There is a simple rule for connecting them. You connect one interface's TALKER A to the other interface's LISTENER A, and the same for the B signals, first interface's TALKER B to the other interface's LISTENER B. This is such a simple rule that most people don't even have to write it down. You can remember something this straightforward.
That's all there is to it. If the manufacturers of marine electronics just followed the rules, we'd be done.
Decoding Signal NamesFor some reason--I have no idea why but it perhaps it is due to some sort of general lack of rigor in preparing technical documentation that occurs in many consumer electronic products--the signals on a NMEA-0183 interface are almost never clearly identified with TALKER and LISTENER, the names recommended in the standard. Instead of using the clear, intuitive, standard names of TALKER and LISTENER, the manufacturers give the signals all sorts of names, probably the names their engineers used as shorthand names when they designed the product. In my article,
Guide to NMEA-0183 Interconnections (that I mentioned earlier in the thread and which is a bit more concise in its writing), I give a table of translations so you can decode the non-standard or shorthand names used by the manufacturers into the standard names. These translations are not that hard, actually. They are as follows:
--if the manufacturer's name for a signal is "in" or "input", that signal is a LISTENER.
--if the manufacturer's name for a signal is "receive" or "receiver" or "rx" or "rcvr" or any form that looks like it is derived from "receive", that signal is a LISTENER.
--if the manufacturer's name for a signal is "out" or "output", that signal is a TALKER
--if the manufacturer's name for a signal is "transmit" or "transmitter" or "tx" or "xmit" or "xmtr", that signal is a TALKER
But the manufacturer's weren't done creating confusion. They decided they would polarize their signal names and give them positive and negative designators. This is completely baffling because these signals do not really have any sort of DC polarity to them at all. (To be technically accurate, the A signal is the logic true or non-inverted signal, and the B signal is the logic inverse or inverted signal of the A signal; sending electrical signals like this is called differential signaling.) They are not positive and negative like the terminals of a battery. (The signals are logical inversions of each other.) You have to unravel this further misidentification in the naming of the signals from the manufacturers, but the secret decoder key is really simple. Here it is:
--if the manufacturer's name for a signal is "positive" or "plus" or uses the plus-sign (+), that is the A signal
--if the manufacturer's name for a signal is "negative" or "minus" or uses the minus-sign (-), that is the B signal
Again, these translations are not really that hard. You may not have to write these down to remember them, either.
Three Rules for InterconnectionsNow that the signals are identified, all that needs to be done to interconnect those signals is as follows:
--(1) connect a TALKER to a LISTENER with the same A/B designation
That is our first rule. Maybe write it down to help remember it.
But the manufacturer's weren't done creating confusion. Instead of following the standard that has been in effect since 1992, which means for 24 years, the marine electronics industry again went off the main channel and into the weeds, and decided they would randomly OMIT some of the signals. This probably saved them $0.25 in costs per unit, but it created a billion dollars of confusion. But we can handle this problem, too. The omitted signal is always the B signal, so we have to make an allowance for some interconnections between devices to have the B signal missing. That means we need a little more complicated rule for how to handle a B signal when it has no mate to connect to. We need two rules because we treat the missing signals differently, depending if a TALKER or a LISTENER is missing. These two rules are probably the most complicated rules in the whole realm of this interface problem, but they are simple, too. Here are the second and third rules:
--(2) if there is a LISTENER B and no corresponding TALKER B to connect it to, then connect LISTENER B to the power common (that is, the black power wire circuit or the device's chassis circuit or the battery negative circuit);
--(3) if there is a TALKER B and no corresponding LISTENER B to connect it to, then insulate the TALKER B and don't connect it to anything.
So there are a total of three rules for interconnecting, and they are very simple. That is it. We have figured out how to interconnect any situation with NMEA-0183 interfacing. Now we go on to the actual problem of accomplishing the interconnection.
Making the Actual ConnectionsAt this point, we know that there are two kinds of signals, and the manufacturers hide them with a bunch of different names. By now we have figured out how to decode the manufacturer's names to get to the standard names. And we know how to interconnect the signals because we followed the three simple rules of interconnection given above. Now we want to interconnect the signals, that is, we want to find the wires that have these signals, and connect those wires together physically so they make an electrical connection.
How do we find the wires? You won't believe this, but those manufacturers decided to throw in more confusion. Instead of using a standard connector with a standard number of pins and a standard arrangement of the signals on those pins, most NMEA-0183 interfaces are provided on a wiring fan out with a bunch of individual wires coming out of a little multi-conductor cable.
In a situation where we are provided with a bunch of wires, the only sure way to distinguish one wire from another is by the color of the wire insulation. You may not believe this, but even though there is a published standard for the wire insulation color to be used for a particular NMEA-0183 signal, the manufacturers once again went off on their own and made up wire insulation color patterns for the signals as they pleased. They went so crazy that sometimes on the same product they would change the wire insulation color from production run to production run, so that their own literature about wire color coding was not even correct. There is absolutely no rhyme or reason to the colors that are used. There is no point in trying to discover a rule or pattern--there is no pattern or rule to this madness. Every piece of NMEA-0183 gear seems to use a completely new and unique choice of colors to identify the wires.
The burden of figuring out signal identity from wire insulation color falls to the installer. The installer has to find some documentation in the manufacturer's literature that identifies what wire insulation color is used for what signal. I cannot help you with this, because there is no consistency in the use of these colors. You will just have to pore over the literature and find the page or drawing that shows the colors. And, just to keep up the insanity, sometimes this information is only given on special addendum sheets or in special installation instructions, and is omitted entirely from the main documentation like the operating guide or instruction manual. Don't throw out any literature that comes with a marine electronic device because that little slip of paper may be the only place with the wire color codes.
This completely insane behavior about wire insulation color is why I NEVER tell people how to connect devices by telling them to connect a certain color to another color. It is a crazy way to give directions. It is like giving traffic directions by telling someone to turn left at the green light and to turn right at the red light. It makes no sense. So I never offer advice like that. I offer advice like I have given you above. I teach you to read the roadmap, not turn left at the red light. Some organizations, like the USCG, tried to collect all possible combinations of devices and describe the interconnections by wire color. That does not work. The first time I looked up data on their chart for two devices, I found it was wrong. That is another reason to avoid advice that comes to you with a recommendation to connect a certain color to a certain color: there is a good chance it is bad advice. Figure it out yourself; you are probably smarter than the guy giving you advice.
Working with Small WiresEventually, let me assume you find a wire with a certain color and want to connect it to another wire. The manufacturer's weren't done making this task hard. They usually provide these signals on very small wire, probably 22-AWG or 24-AWG. Most boaters are not very handy when it comes to working with small wire gauges because it is very unusual to find any electrical conductor on a boat smaller than 16-AWG or perhaps 18-AWG. That has occurred because those gauges are the recommended minimum gauge for any electrical conductor on a boat. But the marine electronics manufacturers did not let those recommendations get in their way; they just use 24-AWG for their NMEA-0183 signals.
Making physical and electrical connection of 22-AWG or 24-AWG conductors can be a bit difficult. This is one point where a boater might want to get some help from a good technician. These small wires cannot support themselves, and if you make a connection between two conductors with small wire like this, the wires must be held in place physically by some means other than their electrical interconnection, otherwise there is a good chance that, over time and with movement and vibration, the connections will fail.
There are many ways to make electrical connections between small wires. One popular way is to use a very small terminal strip with compression contacts. You can get these small terminal strips with compression contacts from almost any supplier. (More on this below in a follow-on posting.)
I have my own method of interconnecting NMEA-0183 using a five-pole connector arrangement. I wire every device the same way to the five-pole connector. Then I construct a motherboard with mating connectors that is pre-wired to interconnect the devices. My invention of this method employs a universal wiring arrangement for the device-to-five-pole connector, and a universal arrangement of the five-pole connector to the motherboard. With my method, interconnecting devices is just a matter of connecting each device to the five-pole connector in a standard way, then just plugging devices into the motherboard to interconnect them. I explain the method in detail in an illustrated article. See
Universal NMEA-0183 Interfacehttp://continuouswave.com/whaler/refere ... A0183.htmlIt took me a while to develop the Universal NMEA-0183 Interface. There is nothing like it from any of the manufacturers or from the NMEA organization. If they had come up with this 24-years ago, no one would have ever asked a question about how to interface NMEA-0183 devices.
Now, as for the notion that you will have to pay someone $440 to figure this out, I want to caution you. There are a lot of people employed by someone as a marine electronics installers who won't get this correct. There is a good chance you might pay a dealer $440 and when your radio and chart plotter are installed, they won't communicate via their NMEA-0183 interface.
Signal Rates: Fast TalkersOh, one last thing. When I said you only had to figure out the physical interconnection, I was leaving out something: the signal rates. Usually all NMEA-0183 signals are sent at the standard rate, but--as you might expect--there is an option for a different signal rate, a faster rate. The normal signal rate is called "4800". This works almost all the time, and, for once, the manufacturers usually have their gear pre-set to this rate and you never have to worry about it. The faster rate is called "38400" and is used only in special situations. The most common special situation is for the TALKER from an AIS receiver. If you have an AIS receiver it will normally be set up for the faster rate because it is possible that an AIS receiver could have so much data coming into it from its two radio receivers that all that data could not fit into the NMEA-0183 TALKER stream of data in the time available if the lower signal rate were used. Or, perhaps simpler to understand, the AIS receiver is a FAST TALKER. When you have a FAST TALKER you have to set the LISTENER to match its fast talking rate. You'll have to carefully read the operating instructions for the LISTENER device to find where the controls are for setting the signal rate.
There is one more consideration about NMEA-0183 and fast and slow TALKERS, but that is more of a graduate school level topic, so I will let it rest for now. I hope my little essay has been helpful to you and now you won't have to say:
I am writing this like I know what I am saying. Not true.
[Note to readers: I updated this article several times from the first posting, so I recommend re-reading it if you got the first-draft version--jimh]