Autopilot Rolls Into Turns Like A Fighter Pilot

[Dynon101]

I have been working to try to smooth out the autopilot and it is doing a really good job except...

It rolls into turns very aggressively like it wants to be a fighter pilot. In other words, when cruising along at a fairly slow cruise IAS I get to a 30 degree course change the autopilot will crank and bank into the turn instead of a smooth roll in. It does however seem to roll out of the turn smoothly with no complaints. When using Track mode it rolls in and out normally so it seems to be a function of NAV.

The Roll Gain was 1.0 and I tried 0.5 and it still did the same aggressive roll.

Is Roll Gain the parameter that causes the autopilot to roll into turns quickly?

The SkyView Autopilot In-Flight Tuning Guide does not seem to address this.

THANKS for your help!!!

 

[Dynon]

One thing you can try is limiting the roll angle artificially so it just won't use as much roll authority.

 

[Dynon101]

>One thing you can try is limiting the roll angle artificially

Are you saying that I should limit the amount of bank angle that the autopilot will fly (like 30 degrees of bank to 25 degrees of bank)?

What exactly does the Roll Gain do? I was thinking that it was the setting that makes the roll into and out of a turn aggressive.

On this last flight I did not think of trying the heading mode and spinning the heading bug left and right to see if the autopilot would produce the same aggressive roll rate.

 

[Marc_J._Zeitlin]

I would lower the ROLL SENSITIVITY, not the ROLL GAIN.

 

[Dynon101]

THANKS Marc...as always I greatly appreciate your advice.

What is the difference between gain and sensitivity?

 

[dougnlina]

I lowered the allowable bank angle and it made for a more comfortable autopilot experience, you might want to reduce it and see what you think...

 

[jnmeade]

Do you want the autopilot to fly a standard rate turn (3°/second) or just something comfortable? A standard rate turn rule of thumb is speed divided by 10 plus 7 (Instrument Flying Handbook).

 

[Marc_J._Zeitlin]

THANKS Marc...as always I greatly appreciate your advice.

What is the difference between gain and sensitivity?

See:

V16 definitions

After installing V16 software, in the auto pilot setup screen for roll control, I see new roll control points. CDI gain, CDI dot gain, XTE gain, and XTE dot gain. I can't find any documentation that explains what they and how to adjust. Thanks in advance for any help!

forum.flydynon.com

 

[RVDan]

I would also suggest setting the max bank angle to whatever is standard rate for your airplane. I started out with 30 degrees and foundit unacceptable. By changing the max roll all was good.

 

[Marc_J._Zeitlin]

I would also suggest setting the max bank angle to whatever is standard rate for your airplane. I started out with 30 degrees and foundit unacceptable.

Bank angle for standard rate turns is dependent upon TAS - it's not a single angle. See:

Standard rate turn - Wikipedia

en.wikipedia.org

 

[CanardMulti]

What Marc states is correct, and it goes a lot farther. One specific TAS + one specific bank angle = one specific turn radius, which produces a specific rate of turn. Size and weight of the aircraft does not matter, be it a 747, 757, King Air, Bonanza, or Cozy.

Ken

 

[kellym]

Bank angle for standard rate turns is dependent upon TAS - it's not a single angle. See:

Standard rate turn - Wikipedia

en.wikipedia.org

For your typical Vans aircraft 22-23 degrees works pretty well.

 

[RVDan]

No argumentthat standard rate varies, but we don’t have a choice to select that function for bank limiting. In my case, I limited to 17 degrees, which gives me standard rate at TAS of concern. Close enough for bank limiting. All of the points about standard rate calcs are pointless with respect to the OP original question.

 

[Marc_J._Zeitlin]

No argumentthat standard rate varies, but we don’t have a choice to select that function for bank limiting. In my case, I limited to 17 degrees, which gives me standard rate at TAS of concern. Close enough for bank limiting. All of the points about standard rate calcs are pointless with respect to the OP original question.

Which was why I told them to adjust the sensitivity as well as the gain. There's no reason that it should be acting as if it's in a dogfight. Roll rate being too fast will not be fixed by changing maximum bank angle - it'll just roll too quickly to a lower angle. Doesn't fix the problem.

 

[RVDan]

Which was why I told them to adjust the sensitivity as well as the gain. There's no reason that it should be acting as if it's in a dogfight. Roll rate being too fast will not be fixed by changing maximum bank angle - it'll just roll too quickly to a lower angle. Doesn't fix the problem.

True, but I found that limiting roll angle reduced the rate of roll. Since the change is smaller the needed rate to get the desired change is less. If it doesn’t work for you, I would reduce gain. I fly an RV which has pretty responsive roll and found roll gain and sensitivity defaults to be good.

 

[Dynon101]

Thanks everybody for your replies always grateful for the education and experience on this forum.

Originally the roll gain was 10 and on that flight I reduced it to 0.5 and did not notice a significant difference.

The airplane was only flying at 135 KIAS and normally it would fly at 190 KIAS and I've never noticed a rapid roll rate during normal flights. Because we were sightseeing and since gas is so expensive it was flying at a low altitude.

I do remember somewhere back when the change to revision 16 it added some different autopilot settings so perhaps I need to do something different with the lower speed settings.

I remember that gain is the amount of force that it rolls into a turn and the sensitivity is the amount of twitch that it has when finally at that heading track course so that's why I digested down the roll gain

 

[RV8JD]

To the OP: What software version are you running? By any chance is it v16.2.4 or something earlier? The reason I ask is that I ran v16.2.4 for 5 flights. I thought that the AP roll responsiveness was faster/sharper than I thought it had been when I was running v16.0.4 (I never upgraded to v16.0.8).

Because I had the Tach Time calculation issue addressed in another thread, yesterday I downgraded back to v16.0.8 and flew the airplane. The AP responsiveness was less and was as I remembered it to be prior to the v16.2.4 upgrade. I did not look to see if v16.2.4 changed an AP settings, but I might take a look if I get a chance.

 

[Dynon101]

>To the OP: What software version are you running? By any chance is it v16.2.4 or something earlier? The reason I ask is that I ran v16.2.4 for 5 flights. I thought that the AP roll responsiveness was faster/sharper than I thought it had been when I was running v16.0.4 (I never upgraded to v16.0.8).

Hmmmm...now that you mention it...This characteristic just started after I switched to v16.2.4. Prior to that it worked super smooth and as expected.

Just curious to know what if anything was changed with respect to the autopilot software with v16.2.4.

THANKS RV8JD for the advice...the suspense builds....

Dynon...any suggestions?!?!?

 

[RV8JD]

Hmmmm...now that you mention it...This characteristic just started after I switched to v16.2.4. Prior to that it worked super smooth and as expected.

I'll add that I did not like the sharper, faster, more abrupt AP roll-in characteristic in v16.2.4 when initiating a turn. So I'm curious as to what changed also.

And I agree, prior to v16.2.4 it worked very nicely.

 

[steve m]

Ah..... 16.2.4.....I updated the software just prior to my last trip (RV10). Previous to this version software I have been tweaking the AP settings and thought I was happy. On my last flight, I did experience a high roll rate in a turn. I was in cruise probably 160 knots and transitioning waypoints I thought the roll rate into the turn was too aggressive, especially for IFR. (I have max bank angle set to 22 deg)

Also, on same flight I was shooting coupled LPV approach and I got the bucking bronco ride while on the VNAV descent (fast oscillations pitch up pitch down, probably about 1 hz oscillations )

You guys have me wondering now if new software had anything to do with this. I have already tweaked my AP settings after this flight?

Here are my current settings (I am not sure why XTE gain and XTE Dot gain are at the values that are shown? I have never changed those??)

Steve