Large S-turns around nav course

[N707RW]

Hello,

I have a recently installed Dynon AP in an RV-7A. I am driving it with an Apollo GX-60 GPS over the serial RS-232 port. I am getting severe S-turning back and forth across selected nav courses. On the first couple of test flights, the s-turns happened each time I selected nav. Yesterday the air was very smooth and it happened on the fourth time I selected a nav course.

I know that there may be compatibility issues with the Apollo GPSs, but I believe that my GPS is outputting all the required data, as evidenced by the three successful GPS tracks. They are in the datalog (attached) at 012639-012831, 013524-013706, and 013812-013939.

In the datalog at 014222-014754 there is a good example of the s-turns. Initially the GPS was given a direct-to a point. Nav was engaged. The airplane turned from 346 to approx 300 heading and held that well until 014459 when another GPS track was selected (actually a GPS approach was loaded). Nav was not disengaged. The airplane turned to 045, then left to 300, then right to 045, then left to 304. At 014754 nav was disengaged because the s-turns were not stopping. The s-turns seem to go to a 45 degree angle and then reverse direction.

In bumpy air the problem is much worse. Sensitivity has been varied from a high of 15 to a low of 5 with better results at 5. Torque is set to 100% and there are no apparent roll slips.

The autopilot was installed with the help of beta tester Rich Wilson and he is submitting this report for me.

Thanks,

Paul Pajak

 

[rvmills]

Paul (and Rich),

Couple thoughts. You answered the first one when you mentioned the sesitivity settings. I'm at 7 (RV-6), and it performs mo betta there for me.

Q's for you: how far off-axis is the direct-to point you are selecting, and how fast are you turning to intercept?

I've noted that a pretty far off-heading direct-to results in what looks like an overshoot, and some S-turns to capture. I have my turn limits set at 2 deg/sec and 20 deg AOB, so its a pretty docile turn. I do know that the GPS draws the line for the direct-to when you execute it, and if the turn is pretty docile, you are actually flying somewhat downrange of that line, so the AP looks (to me) to be correcting back to the original course line that the GPS drew...and it seems to be a pretty aggressive intercept. If the point is more straight ahead, the intercept seems less aggressive, though I do feel that even in those cases it is a little aggressive (in correction angle).

One thing I've found to help (and it works for me at work too) is to do a second direct-to the same fix, and execute it just before the airplane gets to the original course line (actually just before the aircraft heading is parallel to the original direct-to course line). Seems to dampen the overshoot, though occasionally I've gotten a little bump in the ailerons as the AP sees the new course line and attempts to capture it.

You may be dealing with direct-to fixes that are nearly on heading and still getting the S-turns, which (IMHO) would be not-so-good performance, but just thought I'd toss this out.

If you find a fix that really dampens out the S-turns, let us knuckleheads out here know!!

Cheers,
Bob

 

[agrantz]

Paul.
Which version of the firmware are you using?
-Alan