NMEA-0183 Interface Problems; a Real-world Example

Electrical and electronic topics for small boats
jimh
Posts: 11673
Joined: Fri Oct 09, 2015 12:25 pm
Location: Michigan, Lower Peninsula
Contact:

NMEA-0183 Interface Problems; a Real-world Example

Postby jimh » Fri Mar 15, 2019 6:50 pm

A boater recently posted on a popular marine electronics forum that he was seeking help making an interconnection between two devices that had NMEA-0183 interfaces. Each device provided the NMEA-0183 signals on loose wires that were part of a multi-conductor cable, with various different wire insulation colors distinguishing one conductor from another. The boater observed that in each device there was a blue wire and a brown wire, and one device had a green wire. The boater tried to establish NMEA-0183 serial data communication between the two devices by connecting the green wire to ground, connecting the blue wires together, and connecting the brown wires together. This plan was apparently based the boater's intuition about what the wire colors meant and how they should be inter-connected. The plan did not work. The boater made two very bad but rather classic errors;

First, the boater assumed that two different manufacturers used the same wire insulation colors for their signals; this would seem to be a reasonable assumption in most situations, as one would expect that some governing or regulatory body would have established a strict wire color code to identify electrical signals. We see this in many situations. For example, 120-Volt AC electrical wiring in residential buildings follows a strict color coding for wire insulation, and uses green as a ground, white as a neutral, and black (or blue or red) as a 120-Volt AC hot or line conductor. Unfortunately, in connecting recreational marine electronic devices, there is no strict regulatory agency enforcing simple rules like color codes for wire insulation to identify the signal--other than perhads red is a positive DC circuit and black is the corresponding negative side of the circuit, In boat electronics there is just NMEA, the National Marine Electronics Association. NMEA has not been effective in specifying wire insulation colors to be used in a standard manner in the NMEA-0183 protocol specification until very recently, and for decades manufacturers had little concern about compliance with any standard for wire insulation color, even when one was promulgated. It has been only very recently that compliance with wire insulation color standards for NMEA-0183 devices finally began to appear on some products, around 2017. Note that the standard began more than two decades earlier.

A second problem: even if our boater's assumption were correct, that is, all manufacturers used the same wire colors to identify signals, then interconnection of serial data ports from one device to another would NEVER occur by connecting the same signals--the same wire colors--togethers. In NMEA-0183 serial data communication for device to device, the OPPOSITE type signals are to each other, that is you connect a TALKER signal to a LISTENER signal. Connecting two signals of the same type together would result in no communication, e.g., connecting two LISTENERS together would not let the two devices communicate. You have to connect a TALKER to a LISTENER. So if there were standard colors for signals, you would NEVER connect similar wire colors together--except in the unusual case of connecting two or three LISTENERS in parallel to listen to one TALKER.

A third problem for everyone--including manufacturers--is the choice of wire colors that NMEA made. The NMEA-0183 standard proposed the following wire insulation colors be used:

TALKER A    = White
TALKER B = Brown
LISTENER B = Green
LISTENER A = Yellow


While these four colors are not particularly unusual, there is one big problem: where can you buy any four conductor cable with two twisted pairs that use white and brown for one pair and green and yellow for the second pair? Further, the NMEA-0183 specification specifies that these pairs are to be shielded and have a drain wire. There is no commonly made off-the-shelf cable available with wires insulation colors like this. There are two-pair shielded cables, but the pairs usually have insulation colors of red-black and green-white.

jimh
Posts: 11673
Joined: Fri Oct 09, 2015 12:25 pm
Location: Michigan, Lower Peninsula
Contact:

Re: NMEA-0183 Interface Problems; a Real-world Example

Postby jimh » Mon Mar 25, 2019 11:08 am

After a rather thorough search for an off-the-shelf source of a cable with two shielded twisted pairs with the recommended NMEA-0183 wire insulation colors (WHITE/BROWN and YELLOW/GREEN), the only source found was a German specialty wire manufacturing company. I contacted their North American office, which ironically was located only a few miles away from me, and inquired about obtaining a sample of the cable. Their local representative was extremely helpful, but found that there was no cable of that type in stock in any of their facilities on two continents, and to get any such cable would require ording the cable and waiting about ten weeks. Since I was not a "Tier-one" automotive company looking for a few miles of wire, I reluctantly informed the firm I was not in a position to place an order. Thus I have come to the unfortunate conclusion that while NMEA may specify certain wire insulation colors for identification of signals in the NMEA-0183 interface, there is no reasonable source for this cable, at least none I have found.

jimh
Posts: 11673
Joined: Fri Oct 09, 2015 12:25 pm
Location: Michigan, Lower Peninsula
Contact:

Re: NMEA-0183 Interface Problems; a Real-world Example

Postby jimh » Wed Mar 27, 2019 11:33 am

The topic of where a two-twisted-pair shielded cable with proper wire insulation color can be found has been moved to its own topic thread. See

NMEA-0183: Two-pair shielded cable with proper wire insulation
http://continuouswave.com/forum/viewtopic.php?f=9&t=5123