No GPS connected, but $ECRMC shows GPS Position?

[Dos]

Hi!

I am just testing a bit the direct "Navigate to" waypoint option in OpenCPN5 when doing a right click on the chart. This in order to forward the navigation data to the autopilot connected in my network by configuring the according COM port in OpenCPN to forward this data.

Just doing this at home testing without any sensor connected to my notebook only learning the general settings etc.
For sure there is currently no GPS sensor connected to my notebook.

As expected, looking at the NMEA0183 data stream window in the OpenCPN connections tab, the following NMEA0183 sentences are shown as forwarded:

$ECRMC - Recommended Minimum Navigation Information
$ECRMB - Recommended Minimum Navigation Information
$ECAPB - Autopilot Sentence "B"
$ECXTE - Cross Track Error

Nevertheless the forwarded RMC sentences are showing GPS position data:
$ECRMC,102340,A,5227.738,N,00502.400,E,0.000,0.000 ,280419,1.550,E,A*04

I would have expected that this $ECRMC will not be sent in case of no GPS sensor connected or showing a "V" on position 2.)

Am I wrong? What position is forwarded from OpenCPN5 in that case? The last known one automatically saved in the program?

BR,
Dirk

[nohal]

The position of the boat. There always is a boat in OpenCPN. The boat can be positioned anywhere even without GPS connected by right-clicking the canvas and selecting "Move boat here" from the menu. The boat is not removed when there is no data received from GPS or any other sensor providing position.

[Dos]

Hm, thanks for info.

Means in case of a loss of the GPS signal (GPS sensor damage, network connection problem) the autopilot in waypoint steering (track) mode will work with a "rubbish" position reported as valid position from OpenCPN.

What could be the further result of this wrong position report to the autopilot steering? Is unexpected and dangerous heavy manouvering into a complete different position possible?

[nohal]

Originally Posted by Dos

What could be the further result of this wrong position report to the autopilot steering? Is unexpected and dangerous heavy manouvering into a complete different position possible?

Your boat will probably continue to go straight all the way to the edge of the Earth disk and then fall down. Or do something completely different, like start going in circles, as it will keep getting constant XTE value regardless of what it does to the rudder.
Normally mariners either hold watches or at least use the watchdog plugin to be alerted when the equipment fails.

[denverd0n]

Depending on the software you are running on your notebook, it is possible that it is getting an approximate position based on the wifi that you are connected to. Does the position look even remotely like where you are? If so, that is probably the explanation. If not, then it is probably just a default, or last known, position that is left over in OpenCPN.

[Dos]

Hi denverd0n,
Thanks for your response. I also thought about appr. GPS position by WLAN, but in my case the position is neither the last saved boat position nor the position from WLAN. It seems to be some default position programmed in OpenCPN by the developers.

I have watchdog also in use, but from the moment of the alert maybe the autopilot also made the very bad steering already. Much fun in according waves. Therefore I do not agree with nohal. As a lone sailor in practise I do not sit at the rudder all days 24h ready for immediate steering.

I think it would be better if OpenCPn would not forward any position in such case. Then the autopilot would run into error and would give an audio alert ("No data" in case of my Raymarine autopilot) instead of steering - in worst case - a gybe under full sails.

[nohal]

Originally Posted by Dos

Hi denverd0n,
Thanks for your response. I also thought about appr. GPS position by WLAN, but in my case the position is neither the last saved boat position nor the position from WLAN. It seems to be some default position programmed in OpenCPN by the developers.

I have watchdog also in use, but from the moment of the alert maybe the autopilot also made the very bad steering already. Much fun in according waves. Therefore I do not agree with nohal. As a lone sailor in practise I do not sit at the rudder all days 24h ready for immediate steering.

I think it would be better if OpenCPn would not forward any position in such case. Then the autopilot would run into error and would give an audio alert ("No data" in case of my Raymarine autopilot) instead of steering - in worst case - a gybe under full sails.

There is a trivial way for you to try instead of constructing theories about what would be better or worse: Disconnect the GPS. If you are not satisfied with what happens, tell us what it is and we will try to fix it.