Problem with Raymarine autopilot since upgrading to 5.6.2

[keoe]

I have successfully run OpenCPN in different versions for several years to send the active route to my Raymarine ST6001 autopilot to run it in track mode. When I some time ago upgraded OpenCPN to version 5.6.2, the autopilot started to ask for confirmation for a new route leg very often (like every 2-5th second) even when sailing on a straight leg.
I have installed 5.6.2 as an upgrade to the previous version (5.2) so the settings shall be the same as when it worked. The behaviour is the same for both old and new routes.
The following NMEA sentences is defined as transmit sentences in options for the connections: RMB,RMC,WPL,RTE,GLL,VTG,GSV,GGA,GPwpl,APB,XTE,ECRM B,ECRMC,ECAPB

Is there anyone that have an idea of whats wrong?

[houlejm]

Originally Posted by keoe

I have successfully run OpenCPN in different versions for several years to send the active route to my Raymarine ST6001 autopilot to run it in track mode. When I some time ago upgraded OpenCPN to version 5.6.2, the autopilot started to ask for confirmation for a new route leg very often (like every 2-5th second) even when sailing on a straight leg.
I have installed 5.6.2 as an upgrade to the previous version (5.2) so the settings shall be the same as when it worked. The behaviour is the same for both old and new routes.
The following NMEA sentences is defined as transmit sentences in options for the connections: RMB,RMC,WPL,RTE,GLL,VTG,GSV,GGA,GPwpl,APB,XTE,ECRM B,ECRMC,ECAPB

Is there anyone that have an idea of whats wrong?

I had the same problem with a Raymarine ST2000.

I worked around the problem by sending only RMB and RMC.
I think it has something with FS#2881


JM

[pa391as]

there is a bug in 5.6.2 which transposes WPTs
there is a word around in another thread

[pa391as]

you don’t send APB?

[rgleason]

What is the version of the autopilot_pi and the name of the plugin?

[houlejm]

Originally Posted by pa391as

you don’t send APB?

For my Pilot at least, All the required information is available in RMB.
The system started to work well when I removed APB. I think that because of the bug in RMB, APB and RMC send conflicting information to the AutoPilot which interprets this as a change of destination waypoint.

JM

[houlejm]

Originally Posted by rgleason

What is the version of the autopilot_pi and the name of the plugin?

No plugin used for that.

JM

[SamH]

Bug description and workaround here:
https://www.cruisersforum.com/forums...nt-265606.html

[keoe]

Thank you all for your replies!

Originally Posted by SamH

Bug description and workaround here:
https://www.cruisersforum.com/forums...nt-265606.html

I'm not familiar with the NMEAConverter plugin so maybe someone can assist me. I have add a new conversion with this string "$IIRMB,$ECRMB1,$ECRMB2,$ECRMB3,$ECRMB5,$ECRMB4,$E C RMB6,$ECRMB7,$ECRMB8,$ECRMB9,$ECRM?10,$ECRM?11,$EC RM?12,$ECRM?13,$ECRM?14" (see attachment), as I understood the description in the other thread. This causes my autopilot to steer really wiered in zig-zag pattern in a totally different direction compared to my route. What did I do wrong? I have also tried to only transmit the RMB and RMC messages in addition to this NMEA conversion but this doesn't work either.

Originally Posted by rgleason

What is the version of the autopilot_pi and the name of the plugin?

I not use the autopilot_pi for the moment.

[SamH]

Originally Posted by keoe

I'm not familiar with the NMEAConverter plugin so maybe someone can assist me. I have add a new conversion with this string "$IIRMB,$ECRMB1,$ECRMB2,$ECRMB3,$ECRMB5,$ECRMB4,$E C RMB6,$ECRMB7,$ECRMB8,$ECRMB9,$ECRM?10,$ECRM?11,$EC RM?12,$ECRM?13,$ECRM?14" (see attachment), as I understood the description in the other thread. This causes my autopilot to steer really wiered in zig-zag pattern in a totally different direction compared to my route. What did I do wrong? I have also tried to only transmit the RMB and RMC messages in addition to this NMEA conversion but this doesn't work either.

Did you exclude EERMB from output?
The conversion is creating a new sentence (IIRMB) from the buggy one(EERMB), and if you dont block the original EERMB from output your AP will probably behave like you describe, as it receives both simultaneously.

Post a VDRlog if this is not the case.

[keoe]

Originally Posted by SamH

Did you exclude EERMB from output?
The conversion is creating a new sentence (IIRMB) from the buggy one(EERMB), and if you dont block the original EERMB from output your AP will probably behave like you describe, as it receives both simultaneously.

Post a VDRlog if this is not the case.

Thank you Sam! I had missed to add IIRMB to the transmitted sentences.

[rgleason]

SamH excellent! Thank you. I will add this example to the manual.

[RhythmDoctor]

Hi all, I upgraded to 5.6.2 and can confirm that I had exactly the same problem with my own Raymarine ST4000 autopilot's track mode. (I wish I had looked here first.) I'll try the workarounds as a temporary fix, though it will be a week before I get back to my boat.

Has anyone tested version 5.6.2's MOB functionality with this? Does this bug prevent successful autopilot navigation back toward an emergency MOB waypoint?

[SamH]

Originally Posted by RhythmDoctor

Has anyone tested version 5.6.2's MOB functionality with this? Does this bug prevent successful autopilot navigation back toward an emergency MOB waypoint?

Yes, v5.6.2 will swap the orig/dest waypointIDs when setting a MOB-Wp. The workaround will also work in that setting.

$ECRMB,A,35.281,L,MAN OV,1.0 NM,5237.156,N,00307.879,E,37.755,80.000,-6.000,V,A*1
$IIRMB,A,35.281,L,1.0 NM,MAN OV,5237.156,N,00307.879,E,37.755,80.000,-6.000,V,A*15

[RhythmDoctor]

Originally Posted by SamH

Yes, v5.6.2 will swap the orig/dest waypointIDs when setting a MOB-Wp. The workaround will also work in that setting.

$ECRMB,A,35.281,L,MAN OV,1.0 NM,5237.156,N,00307.879,E,37.755,80.000,-6.000,V,A*1
$IIRMB,A,35.281,L,1.0 NM,MAN OV,5237.156,N,00307.879,E,37.755,80.000,-6.000,V,A*15

I hope that team realizes that the failure of MOB autopilot navigation is a serious bug that warrants a high priority for pushing out a bug fix.

I started to make the needed changes for this workaround, but I remembered that NmeaConverter_pi has, in my past experience, been glitchy with occasionally unpredictable results, and also that my NMEA sentence filters and TalkerID codes have been tweaked over many years to get them just right. I'm concerned that making changes to all these things will lead to many hours of tweaking to get it to work, instead of sailing. That's a Pandora's box that I don't want to open.

I hope a permanent fix to this issue is available soon.