Own Ship COG Prediction Vector off

My own ship prediction vector is consistently about 5 degrees to port of the real world vector. This isn’t a heading issue, that is correct. I assume the prediction vector is derived solely from the GPS position integral. This is true no matter the heading or the current/wind (ie. if the real world vessel vector is 5 degrees to starboard of heading due to current, position vector would show aligned with the vessel heading). The GPS position is always correct, as is the SOG stator, so I’m perplexed as to how the vessel prediction vector can be off.

Hi Soar73,
I gather that you do not have an electronic compass installed and connected into your navigation data.
If not, then the predictor will show it’s best guess based on GPS “course” (the direction the boat is moving) instead of “heading” which is the direction your boat is pointing, regardless of which way it is moving. Think of your boat pointed north, while a current is pushing you to the west. Your heading is north, your course is west. If you don’t have a compass wired in, your display will show your predictor (and the boat icon) to the west, if you have a compass, the boat icon will point north, while the predictor will show west.
With that being said, so we are using the same terms, the course predictor will give it’s best guess where you are going to be in 30 minutes, or what ever you have that set at in Boat Properties.
The only thing I can figure if you are continuously 5 deg off in predictor, is that your boat is actually crabbing 5 degrees. You didn’t mention what sort of boat you have, and if it’s a twin engine, your tachometers could be off calibration to cause this, a bent prop on one side too, if it’s a sailboat, you may have a bent rudder, or bent keel, or something else causing the crabbing.
Have you tried any other navigation program and does it show the same thing?
or am I misinterpreting your problem.
My next recommendation would be to install an electronic compass, which is quite useful for navigation in low vis at slow speeds, as your heading is much more accurate. Typically GPS based course gets pretty wonky at slow speeds and especially if there is any current. Imagine your forward speed being 2 knots, in the fog, with a 2 knot current, your course predictor based on your speed over ground will be wildly inaccurate. I know a guy who put his boat on a jetty due to current and not having a compass.

Thank you for the reply, however this isn’t a heading vs course issue, it is the actual prediction vector. The only input needed for calculating the prediction vector should be the time derivative of the GPS position with some filtering factor multiplied by the user selected time length of the vector.

For instance, and I’ve done this, no current (lake), no wind, 10 minute prediction vector - set the boat up so it is going straight / steady on an AP heading and mark the lat/long of the prediction vector head and make no changes; the boat’s actual position 10 minutes later will miss that predicted lat/long (to the right of the mark) by almost exactly [sin(5 Deg) x s/6]. So to me, this has to be a software calc issue.

Ok, so you do have a compass feeding the autopilot. have you checked the deviation table in sensor settings? It’s almost certain not to be a software issue, since it works for everyone else, it’s almost certain to be something going on in your specific boat or your specific settings, or instruments. Have you tried to run a straight course using visual ranges and comparing your track line to the predictor?

Coastal Explorer does not calculate the COG, the GPS does that. The predictor line is simply showing where you’ll be after the selected time if the COG and SOG as reported by the GPS are maintained. Have you double checked that the GPS is reporting the COG you expect (you can see the actual data reported by the GPS in the Settings > Electronics pages)?