New GPS issues - does the boat needs to move to in order to get the first fix?

[Dsanduril]

There's been a fair amount of RTFM'ing on this thread. At least the one I have seen only has this to say (in the entire manual) about the 'question mark' icon:

When the chartplotter loses satellite signals, the green bars disappear and a flashing question mark appears on the vehicle (boat) icon on the chart screen.

The PGNs say we have good satellite signal and position, so it appears that the Garmin programmers have added 'features' that haven't been included in the manual.

[sgel]

Originally Posted by mhsprang

Sgel,

The contents of PGN 129029 clearly show you have a valid fix.

Your new GPS however uses Pinning, a feature where a stationary GPS outputs fixed and stable values for LAT, LON, an SOG set to 0.0 and a COG marked as "invalid".

You can see that in the screen shot of the PGN contents you posted. Look at PGN 129026 (COG & SOG): the COG field (bytes 3 and 4)contain "FFFF" which is a special NMEA 2000 value indicating "data not available". Your plotter therefore displays a "?" because it cannot determine the direction of the boat icon.

Best,
Meindert, ShipModul

Thank you very much for decoding this, very interesting. regarding the question mark, this was not the only induction, it kept showing my old position couple of miles away, and also the two gmi10 displays did not showed any position.

Anyway thanks again for taking the time.

One of my thoughts right now, is to build some nmea2000 device that will read the data from the network and will resubmit it back in the right format

[Brian.D]

So let me get this straight... and I ask because I don't know.

The data stream coming from the GPS NMEA2K is in ASCII text format or is in Hexadecimal format? In NMEA0183 it is ASCII format and is preceded with a $ followed by a sentence structure in plain text. The data posted above was in hex format and appears to be encrypted unless one has the tools to decode the sentence. That seems rather weird that a company would used a format that is not readily usable to a trouble shooter.

Next, and again out of no knowledge... what is the purpose of having a visual display of a chart if the device displaying the chart cannot indicate your position unless you move? So if you go aground and need help you cannot give your location because the chartplotter does not know where you are anymore.

These two things concern me as a sailor who is relying on the ability to trouble shoot issues and identify exactly where I am.

Asking out of a lack of knowledge.

[tanglewood]

Originally Posted by mhsprang

How is that? Since the inception of NMEA 2000 it is allowed to mark fields to "not available" if the information/value of that field is not available. In this particular case the implementor of the firmware in that plotter has chosen to show a question mark if a GPS has no COG and possibly failed to mention that in the manual. Or maybe the OP hasn't read the manual
After all, we are men and we don't read manuals. Yet we complain that our wives didn't come with a manual, right?

Meindert, ShipModul

This is all arm-chair debugging, but his display device appears to show position only when 129026 shows valid COG. I agree that FFFF in 129026 means no COG, but it shouldn't also be interpreted as invalidating lat and lon coming from other PGNs.


There certainly seems to be a bug here somewhere, but we are just making semi-informed guesses about what that bug is, or at least how it manifests itself.


And I have seen at least one case before where one vendor interpreted an FFFF as invalidating more data in the PGN than the sender intended, so a case of incompatible use of the same PGN.

[mhsprang]

@Brian: NMEA 2000 is a binary format not readable by humans. The screen shots Sgel presented are a hexadecimal (readable) ASCII representation of the binary contents of the PGN's. You'd need a tool like an Actisense NGT-1 and its software to read the contents easily. What I did was decode the data manually using the NMEA 2000 specs and packed it into an special NMEA 0183 sentence which my multiplexers/gateways understand in order to retransmit the data on my office NMEA 2000 network and check it from there.


@Sgel: this is cleary a bug in the firmware of your plotter. Each and every field in an NMEA 2000 message may be marked invalid and if that's the case, no assumptions may be made about other fields that contain valid data. I cannot tell you why your plotter showed a different position as well without havin a minute or so of NMEA 2000 from your network, provided you had to tools to create such a log.


Meindert, ShipModul

[Brian.D]

Thanks, mhsprang. The fact that one needs any type of tool to help trouble shoot is ludicrous, and, going from decimal to hex to ascii is so inefficient. But, that is the way it was designed. JMUHO (u=uninformed)

Again, thanks.

[mhsprang]

Originally Posted by Brian.D

Thanks, mhsprang. The fact that one needs any type of tool to help trouble shoot is ludicrous, and, going from decimal to hex to ascii is so inefficient. But, that is the way it was designed. JMUHO (u=uninformed)

Again, thanks.

Hi Brian,
Of course, when instruments send, receive and process NMEA 2000 data, it is not converted to hex and ASCII, which is much more efficient than converting from and to NMEA 0183. That is inefficient.


That plotter only converted it to hex ASCII in order to make it readable for diagnostic purposes when required.

[sgel]

Originally Posted by mhsprang

@Brian: NMEA 2000 is a binary format not readable by humans. The screen shots Sgel presented are a hexadecimal (readable) ASCII representation of the binary contents of the PGN's. You'd need a tool like an Actisense NGT-1 and its software to read the contents easily. What I did was decode the data manually using the NMEA 2000 specs and packed it into an special NMEA 0183 sentence which my multiplexers/gateways understand in order to retransmit the data on my office NMEA 2000 network and check it from there.


@Sgel: this is cleary a bug in the firmware of your plotter. Each and every field in an NMEA 2000 message may be marked invalid and if that's the case, no assumptions may be made about other fields that contain valid data. I cannot tell you why your plotter showed a different position as well without havin a minute or so of NMEA 2000 from your network, provided you had to tools to create such a log.


Meindert, ShipModul

Thanks Meindert, you mention somewhere that the number of sats in the message was 0, maybe this is the reason why the message was marked as invalid? (perhaps the plotter filters by min number of sats or something to make sure the position is accurate?..), it will be interesting to compare to an actual message from a Garmin GPS. (if anyone can get one )
I still waiting for the reply from Garmin, perhaps they will have more insights.

[sgel]

Originally Posted by Brian.D

Thanks, mhsprang. The fact that one needs any type of tool to help trouble shoot is ludicrous, and, going from decimal to hex to ascii is so inefficient. But, that is the way it was designed. JMUHO (u=uninformed)

Again, thanks.

Btw, (and Meindert correct me if I am wrong) the NMEA 2000 is kind of a close network in the sense that is copyrighted and they (the NMEA organization) does not disclose the full spec (this is also affect the price of the components). There are some open source projects to reverse engineer the messages and for building your own custom devices/ debug. e.g: https://github.com/ttlappalainen/NMEA2000

[Brian.D]

Originally Posted by sgel

Btw, (and Meindert correct me if I am wrong) the NMEA 2000 is kind of a close network in the sense that is copyrighted and they (the NMEA organization) does not disclose the full spec (this is also affect the price of the components). There are some open source projects to reverse engineer the messages and for building your own custom devices/ debug. e.g: https://github.com/ttlappalainen/NMEA2000

From Wiki -
"The NMEA 2000 standard was defined by, and is controlled by, the US-based National Marine Electronics Association (NMEA). Although the NMEA divulges some information regarding the standard, it claims copyright over the standard and thus its full contents are not publicly available. For example, the NMEA publicizes which messages exist and which fields they contain, but they do not disclose how to interpret the values contained in those fields. However, enthusiasts are slowly making progress in discovering these PGN definitions.[1]"

[tanglewood]

Originally Posted by Brian.D

From Wiki -
"The NMEA 2000 standard was defined by, and is controlled by, the US-based National Marine Electronics Association (NMEA). Although the NMEA divulges some information regarding the standard, it claims copyright over the standard and thus its full contents are not publicly available. For example, the NMEA publicizes which messages exist and which fields they contain, but they do not disclose how to interpret the values contained in those fields. However, enthusiasts are slowly making progress in discovering these PGN definitions.[1]"

It’s really more than copyright which is the norm for standards. Signing up to get the N2K spec also entails a confidentiality agreement which prohibits you from disclosing anything about it. So once you know, you also can’t talk. I think that’s a big big part of why interoperability hasn’t converged, even after nearly 20 years. I’d say it’s a joke, but it’s really more of an embarrassment.

[mhsprang]

These standards are indeed copyrighted material. The people defining these standards make a living out of that, just as an autor of any book. I paid for the standards too, being a manufacturer of NMEA multiplexers and gateways. So have Simrad, Garmin, Raymarine and everyone else participating in this business. These standards do indeed improve interoperability but it is always possible for a software developer to oversee something. I made mistakes too when developing my products. It's just that my company counts 4 people whereas the Simrads and the likes are huge. So whenever a problem is reported, I can fix that almost immediately (I must because I am always "in the middle" linking different brands) whereas the large companies are only willing to guarantee interoperability bethwee their own devices. The NMEA on their part take a huge effort in making all brands work together (that is after all the goal of NMEA 2000) but at the same time leave too much open for own interpretation. I could develop an NMEA 2000 GPS receiver for instance that would not transmit one single GPS related NMEA 2000 message and still get it NMEA 2000 certified.

[continuouswave]

Originally Posted by Dr. D

Please read section 1107 in:
https://en.wikisource.org/wiki/The_A...ioning_Service...

Compare at:

https://insidegnss.com/wp-content/up...-SOLUTIONS.pdf

Read the opening two paragraphs. They give a good summary of the two fundamental methods of speed calculation in a GPS receiver. Note that one method "is simple to implement", which I suspect is why it is used in most marine GPS receivers.

The article proposes several other methods for obtaining speed, but the methods are more complex. The notion that a more complex method is possible is not a basis to assume that the $3 GPS chip in a marine GPS is going to use that method.

I am sure that aboard world-class racing sailboats, there are devices for speed determination from GPS that are more accurate than the usual simple methods.

[Stu Jackson]

My decade-old Garmin hh GPSMap76Cx will show the "?" mark when satellites are not present or being read. When they are, the ? turns to the boat triangle. If I'm not moving, it doesn't matter where the pointy ends, er, points. Point being that once a fix is obtained, the boat triangle appears. I doubt very much that Garmin has changed their output symbols.

[Dr. D]

continuouswave: Thanks for encouraging me. Now I am certain all GPS chipsets use Doppler for velocity measurements. I did find one reference that said old (maybe before 2000) chipsets did not have the computing power, but current ones do.