2022-09-28, 09:44 AM
Hi everybody. I am a French sailor who sails a 47-foot sailboat with my family on board.
I have been using pypilot almost every day for about 4 months now following a makeshift installation to replace my old non-functioning pilot. Pypilot has kept me sailing all this time in Scotland and Ireland, many thanks to Sean for his fine work which saved our trip.
To learn how to use Pypilot, I used the forums, wiki and the workbook, which I have started to translate into French here. In most cases I manage to set the PID gains correctly, even if I still find this system « abstract ». In general, everything works very well, however the adjustments are particularly difficult when the conditions are unstable, with little wind, a lot of swell and currents, and when I often have to change sails positions.
But among the questions I ask myself and for which I don’t find an answer on the internet, there is one that comes up a lot and that confuses me a lot: why, sometimes, the calculation of the course from the GPS (coordinates provided by the AIS transponder via SignalK/OpenPlotter) is wrong? Let me explain...
My last sail took 3 hours and I had to steer all the time. There was 20kt of wind, the swell that goes with it, and I was on a broad reach, port tack. I was sailing at about 5kt. I noted in my logbook: "I've been on a 130° course for 30 minutes, pypilot keeps showing me a 110° course. When I activate it to follow its 110° heading, it systematically turns left: 90, 80, 70, 60, 50, 40.... I ask for 130°, it displays 110° and turns again in the wrong direction: 100, 90, 80... port helm when it should be turning starboard. In the meantime, the compass is not working well... and you might say that this is the cause of my problem... Despite the swell, I run a calibration, so that it is less wrong, but after many tries, even if the roll and pitch seem to be good, the GPS heading estimation is still not good (even though the map software detects it well from the same GPS source) and the pilot persists in going the wrong way.
What I don't understand is why, sometimes, pypilot doesn't calculate a correct real heading from the GPS points, when it seems so simple, we're going fast enough to get an accurate value, and we've been sailing straight for 30 minutes? Why does it work so well, yet, most of the time? If the problem is the compass (which has an influence even with GPS navigation mode, if I understand correctly) and it is impossible to calibrate because of the swell, what solution is there to have a functional pilot for the rest of the trip? This is an important question, as I sometimes sail for several days in difficult conditions and it would be unthinkable to wait until the sea is flat to have a functional pilot. Thank you for your clarification and I apologise in advance if this topic overlaps with another.
I have been using pypilot almost every day for about 4 months now following a makeshift installation to replace my old non-functioning pilot. Pypilot has kept me sailing all this time in Scotland and Ireland, many thanks to Sean for his fine work which saved our trip.
To learn how to use Pypilot, I used the forums, wiki and the workbook, which I have started to translate into French here. In most cases I manage to set the PID gains correctly, even if I still find this system « abstract ». In general, everything works very well, however the adjustments are particularly difficult when the conditions are unstable, with little wind, a lot of swell and currents, and when I often have to change sails positions.
But among the questions I ask myself and for which I don’t find an answer on the internet, there is one that comes up a lot and that confuses me a lot: why, sometimes, the calculation of the course from the GPS (coordinates provided by the AIS transponder via SignalK/OpenPlotter) is wrong? Let me explain...
My last sail took 3 hours and I had to steer all the time. There was 20kt of wind, the swell that goes with it, and I was on a broad reach, port tack. I was sailing at about 5kt. I noted in my logbook: "I've been on a 130° course for 30 minutes, pypilot keeps showing me a 110° course. When I activate it to follow its 110° heading, it systematically turns left: 90, 80, 70, 60, 50, 40.... I ask for 130°, it displays 110° and turns again in the wrong direction: 100, 90, 80... port helm when it should be turning starboard. In the meantime, the compass is not working well... and you might say that this is the cause of my problem... Despite the swell, I run a calibration, so that it is less wrong, but after many tries, even if the roll and pitch seem to be good, the GPS heading estimation is still not good (even though the map software detects it well from the same GPS source) and the pilot persists in going the wrong way.
What I don't understand is why, sometimes, pypilot doesn't calculate a correct real heading from the GPS points, when it seems so simple, we're going fast enough to get an accurate value, and we've been sailing straight for 30 minutes? Why does it work so well, yet, most of the time? If the problem is the compass (which has an influence even with GPS navigation mode, if I understand correctly) and it is impossible to calibrate because of the swell, what solution is there to have a functional pilot for the rest of the trip? This is an important question, as I sometimes sail for several days in difficult conditions and it would be unthinkable to wait until the sea is flat to have a functional pilot. Thank you for your clarification and I apologise in advance if this topic overlaps with another.