I've been using 4.0.0 since it came out, and loving it. Really stable version - great job! I've been busy changing jobs lately and spending what little time I have actually sailing. Thus I haven't been able to do the kind of "abuse testing" of OpenCPN
betas that I normally do.
But in my normal use of 4.0.0 on my boat, I have noticed one problem vs. 3.3.???? that I was using last season. In my setup (Win8.1 tablet) I have GPS
data coming in on COM3, and I also have autopilot
data (VTG,ECRMB,ECAPB) going out COM3 to my autopilot's NMEA
input. This allows the autopilot
to operate in "Track mode" and also allows me to display XTE, SOG, and other data on the Autopilot's control panel
This issue is with the Autopilot's display of SOG data. Within OpenCPN's dashboard plugin, SOG is very stable, predictable, and correlates well with my own visual observation (it goes up when the boat speeds up, goes down when the boat slows down, etc.). When running 3.2.2, the SOG displayed on the autopilot agrees well with the OpenCPN
dashboard. It doesn't agree exactly, but it's very close. The differences are easily explained by an apparent time-averaging algorithm, since the two values move in the same direction, just at slightly different rates. It was very reliable.
But when running 4.0.0 to send the NMEA
data to the autopilot, the values displayed on the autopilot are crazy and in wild disagreement. I can see the boat slow down from 5 knots to 4 knots (with dashboard as well as my own eyes), but the autopilot may continue going up to 5.7 for several seconds before starting to drop. And by the time it's dropping, a puff may have come up to speed up the boat, but the autopilot does not register it nearly fast enough. The values are so far off that SOG display on the autopilot is basically worthless.
And the issue is more than cosmetic. An incorrect SOG can cause the autopilot to go unstable, since its control logic automatically adjusts the rudder
sensitivity based on SOG. So if it thinks that the boat is going 2 knots but it's actually going 5, it could oversteer and cause growing oscillations in the heading.
So my question is whether you guys did any changes in the damping algorithms for 4.0.0 that would affect the NMEA sentences going out to the autopilot in a different way from the 3.3 betas. Because I'm definitely seeing a difference.