Again, I believe this happens because stage A and B use the same point of origin, which is your position when you turn the Doppler to MAN/AUTO. If A is set to 45°, 275 miles, and B to 5°, 10 miles, the plane will NOT turn to 005 for 10 miles when reaching B, it will turn back towards a point that is 10 miles bearing 5° from where you were when you turned the DNS on, hence the big U-turn you experienced.
Just try this :
If you want to fly leg A 100 miles 180°, then leg B 100 miles 90° in an L pattern, DNS input should be
A : 180° 100 miles, B : 135° 140 miles.If you set up leg A 180° 100 miles and leg B 90° 140 miles on the DNS, upon reaching B, you'll find yourself turning back heading 50° something degrees to a point 140 miles East of your initial position.
Legs A and B are calculated from the same point of origin. They are shown on the map as forming one trajectory, but what appears on the map is NOT what the DNS calculates. That's the only plausible explanation I've found.
Pythagorus is my copilot
I'm only using DNS for long cross-oceans or cross-desert legs, the rest of the time I still stick to VOR to VOR. Sure, I'm not flying the airways, but at least I get to my destination every time.
Cheers
Nico