Interesting high-altitude autopilot behavior

[Chaffee]

Up at 45000, full burner, I encountered some interesting autopilot behavior:

1) Coupled mode is fine until it hits the selected heading, where it oscillates, making corrections in roll back-and-forth for more than a minute before gradually settling into the heading.

2) This behavior does not exist for Heading Select, where the autopilot behaves predictably, steering to its heading accurately without oscillating in roll

3) Engaging BALT at 3 degrees nose down causes such a severe vertical oscillation, centered on the horizon, that the aircraft nearly departs. I have not tested to see the effects if left on, but I will. Structural failure seems likely. It's that violent.

1 and 3 are so extreme that, if the real aircraft were to behave in such a manner, there'd be an ops bulletin in the latter case (like, "don't engage BALT at high altitude within normal parameters") and likely a software update in the former.

Ship was at half fuel, 6 AMRAAMS, 2 Sidewinders.

Anyone else tried these things with the new FM? Thoughts?

I'll repeat this tomorrow and attach some vids if necessary

[oldcrusty]

As far as coupling to TCN or ACLS TC (when intercepting FB after the 'push') - yes, I've seen the back and forward chase. It eventually settles. It's a bug and I don't think it's been reported yet. 

I didn't see any issues engaging BALT hold at high alt.   I tried it today after reading your post.  I tipped the nose to 3' down, starting from 45.5k ft / M.88  and pushed the BALT button passing 45k.  The nose pulled up to around 3 or 4'  NU then settled level after 1 mild oscillation. Looked normal to me.  I was heavy though, 48.4k lbs  (c/l tank and JDAMs)

[KLaFaille]

22 hours ago, Chaffee said:

Anyone else tried these things with the new FM? Thoughts?

Yep. I have experienced the same behavior in all autopilot modes where the plane trims the roll automatically, including HSEL. When doing CAP, I used like orbiting up around 45k - 48k and found the plane oscillating on my first fight after the update. I've had it start oscillating a couple degrees either way and then increasing to nearly 40 degrees as it fails to capture the heading.

It reminds me of the wing-rocking autopilot bug on the A320 when MSFS first released.

[Chaffee]

On 3/28/2024 at 5:47 PM, oldcrusty said:

As far as coupling to TCN or ACLS TC (when intercepting FB after the 'push') - yes, I've seen the back and forward chase. It eventually settles. It's a bug and I don't think it's been reported yet. 

I didn't see any issues engaging BALT hold at high alt.   I tried it today after reading your post.  I tipped the nose to 3' down, starting from 45.5k ft / M.88  and pushed the BALT button passing 45k.  The nose pulled up to around 3 or 4'  NU then settled level after 1 mild oscillation. Looked normal to me.  I was heavy though, 48.4k lbs  (c/l tank and JDAMs)

Haven't had the time since I posted, but I'll see if I can replicate the BALT flip-out. I was pretty light in that AA loadout.

Thanks for the responses. Minor, but annoying.

I can "game" wing-rock by selecting HSEL on the heading of the CPL, then going back to CPL.

I'll make some vids and report it over in bugs.

Edited April 5 by Chaffee