GPS based Odometer

[LennartG]

Dave, this is something I've never seen or heard of before. Fact is that they both use the same speed data, from the filtered SOG/COG setting.
Need more information like: is it lower/higher, how many satellites etc. Could it be that you have bad satellite coverage or bad GPS placement?
The Odometer has a filtering that dashboard does not use, e.g. it requires a certain number of satellites (4) and a HDOP level of 4 (not required by dashboard). These settings can be changed in the OpenCPN configuration file, e.g. increase HDOP to 6.
/LennartG

[LennartG]

Dave, some additional information:
Look at the GGA message in the connections debug window. Field 6 (following E or W) indicates the HDOP value. It should preferrably be well below 4 for an acxurate GPS positioning. Further information can be found searching Wikipedia for 'Dilution of precision'.
I suspect your value is well above 5, at least to begin with. A better GPS placement would most likely cure the problem.
/LennartG

[rgleason]

Originally Posted by LennartG

The Odometer has a filtering that dashboard does not use, e.g. it requires a certain number of satellites (4) and a HDOP level of 4 (not required by dashboard). These settings can be changed in the OpenCPN configuration file, e.g. increase HDOP to 6.
/LennartG

Lennart is this info in the manual? Thanks.

[LennartG]

Rick, the information is there although not as detailed but more in general terms.
/LennartG

[redog]

Originally Posted by LennartG

Dave, some additional information:
Look at the GGA message in the connections debug window. Field 6 (following E or W) indicates the HDOP value. It should preferrably be well below 4 for an acxurate GPS positioning. Further information can be found searching Wikipedia for 'Dilution of precision'.
I suspect your value is well above 5, at least to begin with. A better GPS placement would most likely cure the problem.
/LennartG

Hi LennartG

Field 6 is showing 1.

I have changed config file to as suggested.
HDOP=6

All GPS receiver's on roof with clear view of the sky.

Will trial tomorrow.
Cheers Dave

[LennartG]

Dave, sorry but I used the wrong reference page and gave you the wrong field number, theese are the correct fields in the GGA message:
a/ Field 6 (following the E/W) is the Quality indicator, For you it indicates a 1 meaning you have a fix. Should be 1 - 5 so that is good.
b/ Field 7 indicates the number of satellites, should be a minimum of 4.
c/ Field 8 is the HDOP (Horizontal dilution of precition). This should be as low as possible with a default maximum of 4 (you changed to 6)
Theese are the only settings that could cause a reading difference between dashboard and the odometer. The speed masurement is using the same method and data.

Your GPS placement sounds as good as it gets.
If any of these values becomes 'invalid' then the speed indicator will go to '0' and, if you are using the latest version, it will take a few seconds again before showing a speed. This delay is as the spped may, straight after a fix, show the wrong values. It shall however never show a value that is higher than the dasboard indicates.

Sorry again and please verify the fields listed above.
/LennartG

[rgleason]

Thanks Lennart, I do think this is useful for your manual. Hope we can get the manual updated in the next few days.
What I mean is your changes updated in the overall manual webpages.

Let us know when you done editing OK?

[LennartG]

Rick, just uploaded that (and more) to prod but it then refused me to take it further to the Plugin Manager, saying there is nothing to compare/upload. Thought I'd give it another shot tomorrow with a further minimal change. Probably as the beta and prods are alike.

/LennartG

[rgleason]

Lennart great you are updating the plugin code.
Here is an example for the error reporting. Expand each of the jobs.
Access is via the green checkmark then pick "Details" at opencpn/plugins.
https://github.com/OpenCPN/plugins/r...ite_focus=true
It is probably a bad url or format, maybe some whitespace at the end or something small.

BTW having same plugin in Beta and Prod is perfectly ok, and causes no problem.

Also the Manual is updated separately using asciidoc/antora and you push the changes up to your remote repository as a commit, or you use the "Edit this page" and make a PR and then accept (since you are the owner). Then when the entire plugin manual is updated these changes are collected and converted to html and uploaded to the io website.

[redog]

Originally Posted by LennartG

Dave, sorry but I used the wrong reference page and gave you the wrong field number, theese are the correct fields in the GGA message:
a/ Field 6 (following the E/W) is the Quality indicator, For you it indicates a 1 meaning you have a fix. Should be 1 - 5 so that is good.
b/ Field 7 indicates the number of satellites, should be a minimum of 4.
c/ Field 8 is the HDOP (Horizontal dilution of precition). This should be as low as possible with a default maximum of 4 (you changed to 6)
Theese are the only settings that could cause a reading difference between dashboard and the odometer. The speed masurement is using the same method and data.

Your GPS placement sounds as good as it gets.
If any of these values becomes 'invalid' then the speed indicator will go to '0' and, if you are using the latest version, it will take a few seconds again before showing a speed. This delay is as the spped may, straight after a fix, show the wrong values. It shall however never show a value that is higher than the dasboard indicates.

Sorry again and please verify the fields listed above.
/LennartG

Hi LennartG


Adjusted config as first suggested (which you have now advised incorrect HDOP =6) and had a strange result.


You can see the time lapsed and the resultant speed in screenshots.
This speed stayed steady for minutes before climbing by 1 knot increments, to the correct speed of 5 knots ish in some 9 minutes.



So now HDOP=4 again in config.
With this I have made no adjustments to standard.


From NMEA window,
Field

6=1

7=8

8=1.2



Will trial again tomorrow and give some screen shots.
Thanks Dave

[LennartG]

Dave, you got me really confused.
You are reporting 8 satellites and HDOP as 1.2, should be ok in all aspects.
I've been using the GPS Odometer for two years now and it has been an official plugin for one year. I have used it on Raspberry and Ubuntu and most OpenCPN users are on Windows but there has never been any reports in this direction. In fact there has not been a single functional problem reported, just one GUI problem with the size of the settings menu. Still there must be thousands of users on various platforms looking at the number of downloads.
Narrowing it down to the speedometer, the design is the same as on dashboard apart from number of digits and that STW has been removed.
Need to know more to find the cause like:
- Your platform.
- GPS Odometer version.
- GPS type used.
- Is connection option 'Filter NMEA Course and Speed data' enabled and if, which Filter period is used?
- The sentence 'GPRMC' contains the speed in the 7:th field (following E/W), is that the same as shown by the speedometer? Note that the filter may have some effect on this if the speed varies.
- A VDR recording could also put some light on this.
/LennartG

[LennartG]

Dave, I've been doing some extensive simulator testing and my conclusion is that you have an error with your GPS unit. The conclusion is based on:
1/ The error you experience settles down/diminishes after some 20 - 30 minutes (comment #30)
2/ The Odometer is showing a (mainly) far lower reading than the dashboard. This could be caused by the GPS occasionally dropping speed data values . Releases before 0.6.x.x (when SK support was added) could be fooled this way.
3/ The OpenCPN SOG/COG averaging is also a player here as releases before 0.6.x.x does not use that setting. Dashboard is honoring it while GPS Odometer speedometer earlier based its data on each reading of speed data.
4/ It is not uncommon that GPS units do not distinguish between 'Satellites in view' (GSV) and 'Satellites in use' (reported in the GGA sentence). It is important that the fix is from satellites with a good spread for a stable position and e.g. 4 satellites in close proximity does not generate a good fix (see also my Wikipedia reference in #32). Some GPS units has been found indicating proper fix, satellites in use and good HDOP values even if they have not fully settled. That could also affect the speed data in the RMC sentence. From release 0.6.x.x this should affect both dashboard and GPS Odometer alike.
5/ I have never seen theese errors on a good working GPS/GNSS unit being NMEA 0183 or NMEA 2000 and I have tested several. Nor has any such problems been reported before in spite of many users.
/LennartG

[redog]

Hi Lennart


Your responses read and understood, I will look at the setup.


Ubuntu 20.04.3
Opencpn 5.2.4
Odometer 0.6.4.0
Filter NMEA enabled and set to 5 seconds



O setup
Priority 9 from AIS unit

/dev/ttyS0
Accept only RMC,GGA,AIVDM


Priority 6 from backup computer with separate GPS

1.1.1.255 port 10110

Accept only RMC,GGA


Priority 5 from echo sounder with own internal GPS

/dev/ttyUSB0
Accept only RMC,GGA,SDDPT,SDDBT,SDHDG,SDMTW,SDVLW


Last test I turned the echo sounder off, Priority 5.
The Odometer remained functioning until I turned echo back on.
Speed then dropped to zero until the echo found a GPS fix, at which point speed resumed immediately to normal.
With your theory of bad GPS, this might suggest that priority 9 and 6 are both bad and that the odometer is running on echo priority 5.
However the odometer does not stop at echo turn off, but when echo is turned back on, maybe suggesting 9 and or 6 are still accepted.



Could the system be retaining bad data and reusing it at this point?


VDR of this attached.


Also with Local time offset set to +1000. Brisbane Australia.

I see a time on the odometer of my local time plus 10hrs displayed.
If I set to 0000 I see the correct time displayed.
Is this offset elsewhere as well?



Thanks Dave

[LennartG]

Dave, your system appears fully updated so the version problems could not be the cause. NMEA filtering is also as normal. So far so good.
This is just a frst quick reply, I will look into the VDR file later today/tomorrow.
The main reason why I suspect a faulty GPS is that it works ok after a while. There is nothing in the GPS Odometer that could have that effect.

Looking at you setup you have multiple soutces for the RMC and GGA sentences (AIS unit, the separate GPS, and the echo sounder). This is normally a big no-no. You can be assusred they do not send exactly the same information. Over time they will probably get closer as the, internally, warm up.
Although the odometer is using the filtered speed signal just like dashboard I've not tested with multiple sources and I'm not sure if this could affect the result. The GPS Odometer does not see the priority order, it is an internal OpenCPN function, so this could be a difference causing the reading difference. Also the effects with echo sounder on/off indicate the same thing.
Not sure about the time offset, I'm simply using the time from RMC and add/subtract an offset.

To investigate further, please test by disabling RMC and GGA from the echo sounder and e.g the AIS, enabling only the GPS unit. There should be no need to have multiple sources as you can easily turn the back on if required.

/LennartG

[redog]

Hi Lennart


"Looking at you setup you have multiple soutces for the RMC and GGA sentences (AIS unit, the separate GPS, and the echo sounder). This is normally a big no-no. You can be assusred they do not send exactly the same information."


I have set these multiple sources up for redundancy.
When I first started I was introducing a whole range of sentences with the 3 sources and yes I agree, they do not send exactly the same information.


Example:
priority 9 RMC,GGA
priority 6 GLL,RMB
O would now have slightly different lat/lon from each sentence, as each GPS is on opposite sides of the boat and the auto pilot was turning port/sboard quite a lot.



However by using the same sentences, all be it not exactly the same information, but lower priority rejected if higher is accepted I have better stability.

I am of the belief that priority 9 would be used and lower only comes into play if 9 fails in some way.
My auto pilot is now very stable in its operation.


What you are suggesting, turning off the 2 other sources, will I believe work, but is the opposite to what I have learnt here on the forum.
Without knowing the reason, I find the dashboard and "O" in general very stable with my setup, whilst odometer not so.


However having had O crash several times on me during night navigation , I can assure you I do not want to be trying to turn on other settings/sources while under duress.


I am home now so any more trials will have to wait, but I am confident that removing the 2 sources will result in a good outcome. (stable speed in odometer) I have tested this in my car with a cheap portable GPS.


Hoping you find something in my VDR,
thanks Dave