Dynon D100 AutoPilot is self-engaging!

Well, the cause of my cylinder issues has been found and remedied. Now it's time to burn some more AvGas. After a great flight on New Years Day with plenty of loops and rolls, I put the RV-7 away for a couple of days. Today was a beautiful day with clear skies and no breeze at all. I had enough time to get a half hour of R&R in the logbook.

So here's where it gets interesting. I have a Dynon D100 EFIS and a D10A EMS, along with autopilot servos on Pitch and Roll. I do not have any other Dynon products - all autopilot functions are controlled by using the soft-keys on the D100.



During today's flight, I did a roll to the left. As I completed the roll, the A/P (pitch and roll) turned themselves on. I didn't recognize this for a few seconds because I was almost done with the maneuver and so stick inputs were minimal, but I could feel that I was overcoming the servos for control.

After disengaging the servos using the stick-mounted A/P disconnect, I tried a roll to the right - no problems noted. I tried a loop - no problems noted. I tried a roll to the left again - same problem occurred. The A/P picked up just as I leveled out of the roll.

I did this four times... Roll right - no problem. Loop - no problem. Roll left - A/P picks up at the end of the roll. This was extremely "repeatable", in that it happened at the exact same point of the left roll - every time!

Now, I've been flying with this exact setup for almost 500 hours. I've used the A/P coupled to the GPS for cross country flights for about 70 hours. I've never had any issues with the A/P before this.

Has anyone else experienced this before? I'm really surprised that the A/P (which is controlled by the soft keys on the D100) could somehow be made to pick up with no input from me.

Are you sure you aren't accidentally engaging the AP via the stick-mounted disconnect? Pressing & holding it will engage the AP.

If you are certain that is not happening, the other thing to look for would be a short in the A/P disconnect wire on the stick.

ChiefPilot said:

Are you sure you aren't accidentally engaging the AP via the stick-mounted disconnect? Pressing & holding it will engage the AP.

If you are certain that is not happening, the other thing to look for would be a short in the A/P disconnect wire on the stick.

Click to expand...

Grumpy,

Thinking the same thing as Brad. Either a consistent accidental activation of the disconnect button in left rolls, or the wire to that button is going to ground as you roll (or as you roll out). That's assuming you have a disconnect button on the stick. I believe you can disable that function of the button in the menus, and could do some ground troubleshooting steps to see if you can reproduce the coupling with and without that feature enabled. Good luck...same EFIS and AP here, so I'd be keen to hear what you find out.

Cheers,
Bob

I'd say chafed wire. AP disc. button needs to be held down for about 2 seconds for it to engage, which would be about where you would be near the end of the the roll. The chafe is somewhere near the bottom of the stick where it exits and is only on one side, such that left roll causes it to ground and right roll does not.
Of course this is all moot should you not have AP disc. or CWS on stick!

Gotta love the combined knowledge base available here on VAF! I hadn't considered the chafed wire on the control stick. I won't be able to get to my plane for a couple of days to see if that's the problem, but it sure makes sense.

or maybe it just thinks you need some help with that maneuver? hidden feature of the autopilot..

A few thoughts to add:

As Bob said, You can turn off "hold to engage" in the AP setup menu, so you could try turning this off and seeing if it fixes this behavior.

On the ground, you could engage the AP, then go full right and left stick and see if it disengages the AP. You don't hurt anything by overriding the servo. If you want to be tricky, give it left and right headings to hold and let it move the stick left and right itself and see if it disconnects itself!

Remember, the disconnect line goes to each servo, so it could theoretically be exposed anywhere along the run, and it could be G's (or lack thereof) that cause the issue.