New Plugin NmeaConvertor

[Schnapsy]

Thank you for your reply.

The indicated sentence ($ SDDBT, 0027.2, f, 0008.3, M, 004.5, F) is the one that directly "comes out" of my Furuno sounder and I recorded it with VDR at the same time as the $ IIDPT sentence.

I think your $ DDBT sentence is sent with a good checksum, which is not the case with mine.
Of course, in OpenCPN connections the checkbox for "Control checksums" is unmarked.

I will re-install NmeaConverter and redo recordings on the boat, this next WE.

[yollivie]

Hello,

I'm trying to install the plugin on my raspberry pi 3 with latest openplotter 0.17.1 stable. It looks like there is an issue with cmake as I got these error messages:

pi@openplotter:~/NmeaConverter_pi/build $ cmake ..
-- *** Staging to build nmeaconverter_pi ***
CMake Error at /usr/share/cmake-3.6/Modules/FindPackageHandleStandardArgs.cmake:148 (message):
Could NOT find wxWidgets (missing: wxWidgets_LIBRARIES
wxWidgets_INCLUDE_DIRS)
Call Stack (most recent call first):
/usr/share/cmake-3.6/Modules/FindPackageHandleStandardArgs.cmake:388 (_FPHSA_FAILURE_MESSAGE)
/usr/share/cmake-3.6/Modules/FindwxWidgets.cmake:917 (find_package_handle_standard_args)
cmake/PluginConfigure.cmake:54 (FIND_PACKAGE)
CMakeLists.txt:34 (INCLUDE)

anyone already encountered this? Is there something that I should do to add wxWidgets_libraries? (I'm new to linux)

Thanks a lot for your help
Yann

[rooiedirk]

For compiling you do need the 'develop' files for wxWidgets. To get all needed files you might ever need for compiling you can take the instructions for https://opencpn.org/wiki/dokuwiki/do...ompiling_linux and then special the get the required programs.

sudo apt-get install build-essential cmake gettext git-core gpsd gpsd-clients libgps-dev wx-common libwxgtk3.0-dev libglu1-mesa-dev libgtk2.0-dev wx3.0-headers libbz2-dev libtinyxml-dev libportaudio2 portaudio19-dev libcurl4-openssl-dev libexpat1-dev libcairo2-dev libarchive-dev liblzma-dev libexif-dev libelf-dev libsqlite3-dev

If you still get error(s) you might need an extra line in the CMakeList.txt like this:

set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -std=c++11")

Somewhere in the first few lines of the file.

succes

[yollivie]

Ok worked perfect. I had to add the extra line in cmakelist.txt and then copy libnmeaconverter_pi.so into /usr/lib/opencpn as was said in a previous post.

Thanks a lot, just have to play with it now!

[rgleason]

Dirk, I've compiled it for Windows and tested. It works.
Since you've changed the Version No. to 1.0 I assume this plugin is "Released" and can now move to the main download page and move the documentation to the main wiki. Thank you.

NOTE: We would like to have more sample test VDR files and examples! Please contribute here! Upload your VDR files and explain how you have used this plugin!

I have uploaded the executable for Windows here

[rooiedirk]

I did indeed send Pavel (Nohal) a message if he would be so kind to move it to release.
Dirk

[RhythmDoctor]

Hi all, I'm trying to use this nice plugin to calculate True Wind Angle and Speed using SOG from GPS instead of paddlewheel transducer. (We have very weak currents in the Chesapeake, so I prefer to use more reliable GPS.) So far the speed calculation looks good, but direction is generating a lot of "-1.#IND0000" error messages. This seems to especially happen when near DDW, almost as if there is a divide by zero error somewhere??? (Haven't been able to track the exact source yet, though.)

The equations I am using are here. These are identical to ones supplied in message #78 of this thread, but I'm using $GPVWR and $GPRMC sentences from my Raymarine ST60 wind instrument and Garmin GPS, and opting to pass through some of the VWR words from the source:

$XXVWT,$GPVWR1+180*acos((sqr($GPVWR3)+(sqr($GPRMC7)+sqr($GPVWR3)-2*$GPRMC7*$GPVWR3*cos($GPVWR1*3.14159265/180))-sqr($GPRMC7))/(2*sqrt(sqr($GPRMC7)+sqr($GPVWR3)-2*$GPRMC7*cos($GPVWR1*3.14159265/180))*$GPVWR3))/3.14159265,$GPVWR2,sqrt(sqr($GPRMC7)+SQR($GPVWR3)-2*$GPRMC7*$GPVWR3*cos($GPVWR1*3.14159265/180)),$GPVWR4,,,,
<return>

Note the return is important!!!

I am testing this by playing back couple of VDR scripts. When I am close-hauled, everything is fine. In an example where I am motoring through a turn where the wind passes behind me (or for an extended time close to DDW, I get the error messages.

Any suggestions of how to suppress these errors?

[transmitterdan]

Rhythm,

If you are using github commit and push your code and tell me where it is. Also, tell me where to get the VDR file (or just add it to github repository). I should be able to compile and find the source of the error.

[RhythmDoctor]

Originally Posted by transmitterdan

Rhythm,

If you are using github commit and push your code and tell me where it is. Also, tell me where to get the VDR file (or just add it to github repository). I should be able to compile and find the source of the error.

Thanks for your kind offer to debug for me. My C++ install is way out of date (WXwidgets version prior to O4.2), and since I'm not a true programmer I don't have an account on Github. So I'll just post the files here.

I used the compiled NMEAconverter Win32 .exe install package that was on Github as of yesterday (version 1.0):

https://github.com/RooieDirk/NmeaCon...-1.0-win32.exe

My formula was posted above. It's a variant on the one given in post #78, except using the $GPVWR sentences that my Raymarine system puts out. When you enter this into NMEAConverter, make sure to place an (invisible) return after the four commas, otherwise nothing works. (Took me an hour to stumble on that trick.)

I'll provide two VDR scripts - one that seems to work fine, and one that gives the INDETERMINATE error. Both are test files that I have appended .pdf to in order to work around the file size limitation. Do not open them in Reader, just rename them by deleting off the .pdf.

[RhythmDoctor]

Hi Dan, have you had a chance to look at the TWA calculation yet? Thanks!

[transmitterdan]

Originally Posted by RhythmDoctor

Hi Dan, have you had a chance to look at the TWA calculation yet? Thanks!

Yes, I was able to run it today. There seems to be a fatal bug in the wxEcEngine.cpp module. It failed for me with a wxTrap using either file. I hacked the wxEcEngine::evalf() function to get around that bug. Not sure if my hack is totally ok or not. There is a lot of weird stuff in that module as there usually is in any parser like that.

So I am not sure what you are trying to do. I cannot see any true wind on the Dashboard with either NMEA file. I don't understand what it means <return> in the equation. Do you literally have to enter the text <return> or must make sure the line ends with a carriage return?

[transmitterdan]

Ok, I was able to delete the text <return> and then it makes true wind on the Dashboard.

The C&D Canal file produces NaNs for true wind sometimes. I am looking into why.

[transmitterdan]

This string evaluates to NaN:

"113.00+180*acos((sqr(8.30)+(sqr(005.8)+sqr(8.30)-2*005.8*8.30*cos(113.00*3.14159265/180))-sqr(005.8))/(2*sqrt(sqr(005.8)+sqr(8.30)-2*005.8*cos(113.00*3.14159265/180))*8.30))/3.14159265"

[RhythmDoctor]

Originally Posted by transmitterdan

This string evaluates to NaN:

"113.00+180*acos((sqr(8.30)+(sqr(005.8)+sqr(8.30)-2*005.8*8.30*cos(113.00*3.14159265/180))-sqr(005.8))/(2*sqrt(sqr(005.8)+sqr(8.30)-2*005.8*cos(113.00*3.14159265/180))*8.30))/3.14159265"

Thanks, you're definitely onto something.

The number within the arc cosine function is >1, which is indeterminate. I didn't do the trigonometry, I just copy/pasted what was on post #78 and changed the variable names. But as I look at this, I see that we have sqr(8.30) term twice, which to me is a red flag for an error. If not, then it would likely have been entered as 2*sqr(8.3). So that's a place to start.

I'll try to re-do all the trigonometry myself, but it might be a couple days before I have time.

[rgleason]

It might be helpful to follow the derivation of the formula. I setup the problem and found a formula which I entered in the wiki asking for help to get it into nmesconveeter. Perhaps we screwed up at some point along the way. It is pretty much chronological with references.