Threaded Mode | Linear Mode

DSanduril

I have an SPX30 (V3.05 software) connected to a P70 (V3.07). A non-Raymarine chartplotter is connected providing navigation information over NMEA.

The AP works flawlessly in Auto mode and does a pretty good job in Wind mode (from an ST60 over ST through an ST-STng bridge). Where it fails completely is in Track mode. My old 3000+ wheel pilot used to handle this without problem 20 years ago.

The NMEA output shows I am sending ECRMB, ECRMC, ECAPB, GPVTG, GPGGA, GPGLL, and GPRMC sentences to the AP (and other units on the network). I don't have a Ray MFD, so can't grab diagnostics.

With the SPX30 here is my sequence:

Get the boat on track (XTE 0.00 or 0.01) and pointed on heading. AP heading output agrees with nav system heading within a degree or two. Display shows that the AP is receiving XTE, COG, SOG, BTW, DTW from NMEA (on rare occasions I have a hiccup here and I'l get the 'no nav data' warning, but that is pretty rare).

Change to track mode. The AP will change heading exactly 30 degrees. If the XTE sentence says -0.00 it will change to the right, if it says +0.00 it will change to the left - i.e. if the LH in Auto mode is 000 it will change to either LH 330 or LH 030. I can have zero XTE, but it changes course based on the sign of the XTE. This is with response from 1 to 3. If I set response higher than 3 (which is where I normally have it in Auto mode) the turns become larger.

Of course that gets it way off course so then it over corrects and turns 60 degrees the opposite direction. It takes about 15 seconds, but it pretty much always ends up at exactly 30 degrees the other side of the correct heading.

This continues forever, never settles out. The longest I've let it run is 30 minutes, it usually wears out my patience long before then.

Other than the response setting I haven't found any other adjustment I can make (such as limiting the LH deviation from the BTW to 5 or 10 degrees instead of 30).

Of course it doesn't do anything like in this in Auto or Vane, there it seems to run just fine. Any thoughts? Any additional data I can provide?

Thanks